For solo consultants who want to make more and work less without hiring.
Jonathan: Hello, and welcome to Ditching Hourly. I'm Jonathan Stark. Today I'm joined by guest Denis Gontcharov. Denis, could you tell folks a little bit about who you are and what you do?
Denis: Yeah, sure. So I am actually a quite a recent graduate. I've been working for four years, mostly in the aluminum industry, so as a process engineer, mostly work in process control and data analytics.
Denis: But since two years I started freelancing, which means that I did data work for mostly corporates or startups, unfortunately, quite general. So without specific niche. And since then, I've actually closed this business again to become full-time employed. Mm-hmm. At an aluminum factory. I have done that for one year and have since last week actually quit this job to try a second time to become a soloist.
Denis: Hmm.
Jonathan: Cool. Very interesting. All right, so I already have a bunch of questions. What we're gonna talk about today is positioning statement for this freelance business. And we've gone back and forth over email a little bit. I thought it'd be easier to brainstorm on the phone like this. So that's what we're gonna do and you can get a sort of a peek into the process that I would use with say, a coaching student or something to kind of uncover a possible or maybe a couple of candidate positioning statements that then, you know, in this case Dennis would, would expose the market to and see if any of them clicked.
Jonathan: So we're gonna come up with what hypothetically would be an effective positioning statement, but you never really know until it does start to attract people who you want to attract, whether it actually works. Alright, so couple of questions on based, just based on your bio. Did you say process engineer?
Jonathan: Yes, that's correct. So explain what that is for someone who might not know. Sure. So
Denis: my first job was an aluminum smelter. And the aluminum smelting has a very complicated physical and chemical process involving lots of machines. So this process is steered by a big computer by PLCs and what my job entailed was writing the code, so the program that steered this process based on inputs like temperature
Jonathan: and so on.
Jonathan: Oh, fascinating. Okay. And you evidently went to school for that? Is that, was your degree kind
Denis: of? So, uh, my degree was in mythology, so I did study aluminum in school. So I looked at metals through a microscope, which is actually not a software education. I actually pivoted more into the field of software due to my first internship, which was really about data analytics and programming.
Denis: So it's not that I really went to school for it. Mm-hmm. Okay.
Jonathan: It's an interesting overlap, this sort of digital world interfacing with the supremely physical world. And absolutely. That's why I love it. Yeah, it's really cool. Okay, so what, okay, so that was, that was helpful process engineer and then the code to run aluminum.
Jonathan: The aluminum smelting machines process and with the background in microscope style metallurgy and, okay. Exactly. And then what, is that what you were doing most recently as your full-time job that you just left? No. So actually
Denis: after two years working as a process engineer, I decided to write the data science hype, which is in full swing in 2018.
Denis: And I actually switched away from the industry to join a consultant and do some freelancing. But uh, long story short, I focused exclusively on data, data science first, but data engineering second
Jonathan: eventually. All right. So what, again, for people who don't know, what's the difference between being a data scientist and being a data engineer?
Jonathan: Well, that's
Denis: a very important point, especially for this industry. The end goal is always data science, like building AI models with machine learning. The problem which occurred was that to build the models you need data and getting the data, especially in this industry, proved to be so difficult that an entire no discipline actually broke off the data science world to just build what we call data pipelines and data sources, so that the data scientists can actually get easy access to high quality data.
Denis: Mm-hmm. Yep.
Jonathan: Got it.
Denis: So it's like a supportive role for a data scientist in a large data team.
Jonathan: Yeah. I've, I've, it's unsurprisingly, this has been a, a problem in lots of different industries where they're like, oh, sweet, we can do machine learning except for we don't have any data. Like machine learning works now, but you don't have any data then There's nothing to train it on.
Jonathan: So, or, or you have data, but it's garbage.
Denis: It's a real pain. In fact, 80% of our projects that did at my last job actually involve data engineering, even though they were like AI projects.
Jonathan: So they're kind of putting the cart before the horse, it sounds like. Yeah. That might be interesting in for, for what we're trying to do here.
Jonathan: So could you tell me more about how common it would be for aluminum manufacturers to be thinking about instantiating an AI project? Is that super rare or is that like kind of, kind of a thing that's being discussed in the industry?
Denis: It's definitely a thing that discussed in the industry, but it's a bit like, uh, teenage sex.
Denis: You know, everyone talks about it. No one really does it, and the people that do it are pretty bad at it. That's how I always describe it.
Jonathan: Nice. Okay. What's important though is that there's interest. The, the teens are very interested in the subject, so that can be a hook to hang your hat on where they, you know, they're having this conversation maybe with their colleagues or their employees or their board or in their own heads.
Jonathan: And if you can join that conversation where it is, that could be a foot in the door to talk about, you know, to kind of work your way upstream and say like, oh yeah, AI could really revolutionize the industry. Like, yeah, let's talk about that. And it's like, okay, where are you getting your, what data do you have?
Jonathan: And they're like, I don't know. It's like, okay, let's talk about that. 'cause that's a precursor, right? So it's, it could be the kind of thing where you get your foot in the door, Based on an existing demand or an existing interest or desire in the mind of your ideal buyer. And then kind of break the news to them that it's not that simple and that there are other things to be considered and the, the one at least, I don't know if it's just one thing that you do.
Jonathan: I mean you seem to have a few different skills here. Uh, the one you seem to be focusing on is data engineer. Uh, but you know, that might not be, I think from our email thread that they don't really know what that is. They've never heard of it, so it's just gonna bounce off their brain. If somebody said, you know, you introduce yourself, oh, I'm a data engineer, and they're just like, great.
Jonathan: What kind of car? Like, there's just no follow-up question. Like, what's that? Yeah, absolutely. Alright, that's interesting. It, so speaking of ideal buyer, who do you think is the right person to decide to bring someone like you in to a aluminum manufacturing business?
Denis: So that was one of my main challenges. I can think of a company which could bring me in, but it's very hard for me to imagine like the actual employee at this company who I should target.
Denis: But maybe it would be helpful if I read you my L F P S, which I have so far. Mm-hmm. So what I said in my L F P Ss was that I help medium sized aluminum manufacturers that want to become more digital as the first sentence. Mm-hmm. And I specifically mentioned medium sized, because in my opinion, the bigger companies, which most of them are, they are mostly being helped between quotes by large consulting companies like your McKinsey or Deloitte.
Denis: Mm-hmm. So it's very hard to get into those companies as a soloist, I think. Mm-hmm.
Jonathan: Could be true. How would you define medium-sized by headcount? By annual revenue, by output.
Denis: So I think by output, like the production of aluminum tons, but also by a headcount, and I have this image in my mind because I operate from Germany and Berlin, we have a lot of what we call the middle stunt in, in Germany.
Denis: It's like a big, um, com, um, glomerate of medium sized businesses, which in fact is the biggest part of the German economy. Mm-hmm. Okay. So like your average factory,
Jonathan: so to speak, but for someone who's not. Yeah, good point. What's an average factory?
Denis: About 700 employees, I would say. Including the shift workers, but like your department of let's say engineers and it people would be like 100 people.
Denis: Mm-hmm.
Jonathan: That is a sweet spot for solo consultants, it seems. Uh, I've had personally and, and a lot of other students have success in that range because when they're smaller, they tend not to have the budget, and when they're bigger, they tend to have just hired or you, like you said, using McKinsey or Deloitte or someone.
Jonathan: Mm-hmm.
Denis: Oh, that's very reassuring to hear.
Jonathan: Yeah. And, and it's more likely that you can get access to the ideal buyer who is, if we worked our way down from the top mm-hmm. The, it's probably called a c e o or president. It's probably, it could be someone, you know, ideally that person that would make the decision to bring you in, but maybe it's not, they've got a lot of other things to worry about other than, than this kind of problem.
Jonathan: So do they typically have a C I O or do they not staff a position like that?
Denis: No, they do. It's important to note that in this industry it is mostly broken off into two parts, but we call it like the information technology and ot, which is the operational technology. Okay. And it's basically the carpet side of the business, is it?
Denis: And the concrete side of the business is the ot, like all the machines, the networks of the factory. That's mostly ot. I think I should be focusing it because, although I don't know, I'm not, not really sure.
Jonathan: Yeah, my gut instinct was ot. Mm-hmm. Mm-hmm. So, so who leads the, the, uh, what, what's the actual word? I, I like the carpet in cement.
Jonathan: Is that what you said? Yeah. Carpet in concrete. Carpet and concrete. And what, what's the o stand for again? Or Operational. Operational technology. Okay. And what's the, is it, is there like a v p of ot or is it a, a direct? Who's the highest person on that pyramid?
Denis: So there would be indeed some kind of, um, VP of ot.
Denis: He would not necessarily be a software person, could also be more of a production person.
Jonathan: Yep. Managing people that like, like with wrenches, right? Yes. The fixed machines. Right. Okay. So maybe that's not a great fit. And then it is. We're all used to that. I suppose the people that are. Doing everything from making sure the wifi network works and the email accounts are set up and secure and all of that stuff.
Jonathan: Right,
Denis: exactly. It's basically a pyramid and the closer you get to the, uh, actual process, then it's ot. So it begins with like sensors, the network, the computers, P L C, that's all ot. But then once you go more into, let's say, database for planning or like an E R P system, which stands for enterprise resource Planning.
Denis: Mm-hmm. Invoicing, accounting. That's more
Jonathan: it. Yep. Okay. And then under it, so where's where if we're dialing our edge of sketch knobs? Yeah. The, the C I O or whoever's, whatever the job title is at the top of the IT pyramid, they're worried about a lot of stuff that isn't op Well, that's, that's not operational on the shop floor.
Jonathan: So, you know, like wifi and other things that you just listed as well. So like, is there someone on the carpet side of the, the room carpet side of the factory that. Specifically, maybe it's a director level or VP level that's specifically in charge of, Hmm, maybe this is too specific, but software that is used in the shop floor, in the operations side, or is that, like, does that kind of fall to a, a random person?
Jonathan: Like someone's just sort of picking up the slack there? Or is it the case that most of these medium-sized aluminum manufacturers really don't, don't handle their software at all? It's just something that the, uh, that the suppliers of the, the machines do. So, no,
Denis: it's definitely that they do handle it themselves.
Denis: They often buy from vendors, but they still need to configure it themselves. Mm-hmm. And to answer your question about who to contact, I think a good point would be to focus on the middle of the different data systems. There's basically four data layers in manufacturing, and each plant looks pretty much the same.
Denis: And the system in the middle, it's called the level three layer. Uh, the m e s, which stands for Manufacturing Execution System, is actually a place where both OT and IT work together. So it usually manages the system, but OT relies on the system to plan their operations. Mm-hmm.
Jonathan: Okay. And that's where you would want to interface?
Jonathan: What, what are the le what the levels on either side of that? Sure.
Denis: So let's begin. The first level is level zero, which is really like the sensor at your particular machine. Yep. Level one is like the logic that's, uh, the data that's coming from the machine used for steering. Yep. Level two is called scada.
Denis: That's basically a computer that operates a line, a number of machines, and then above level two you have the level three, which is the m e s, the management execution system, and that's responsible for basically tracking a work order through the through production. So it basically keeps track of what operations happened on which part and at what time.
Denis: Mm-hmm. The level above that, level four is called the E R P system, and that's really in the IT world stands for enterprise Resource Planning. Mm-hmm. That's basically everything through with, with transforming a customer order into a production order and creating an invoice, and then they have a level five in theory, which is called the clouds.
Denis: That's not very relevant for discussion.
Jonathan: Hmm. Okay. And the e r P also handles supply chain and all that?
Denis: Yeah. Well, I guess a part of the supply chain is start in the m a s system, but usually a part that's a difficulty about aluminum production. It's possible that you do one operation in one plant, then you have to ship the produced coil or aluminum bar on a train to a different plant somewhere in the world, do a second operation.
Denis: And keeping track of those coils is something that is still very difficult.
Jonathan: So the inventory is handled at layer three or layer four?
Denis: Layer four in general. But it, for example, it's, it goes one plant to the other plant. It'll have a, it'll leave a trail in the level three system, the m e s system of each of those plants as well.
Jonathan: Okay. And you feel like m e Ss is the, is the place where it's most likely that these, these two things intersect. So it's like the most likely place. So I wonder, I wonder, so in my mind that seems correct. I also wonder if the e r P level might be an interesting place for you to you to maybe not start the conversation, but at least play around a little bit in that space.
Jonathan: Because it's a little bit higher altitude. It's a little bit more in the IT world. Mm-hmm. And a little bit more c i o and unless, I don't know, is there, in a medium sized company like this, would there be anyone dedicated specifically to dealing with the E R P or is that
Denis: Yeah, there's usually one person.
Denis: So even, even at large company, there's usually one contact point for each system.
Jonathan: Okay. And do you, off the top of your head, do you know that these titles or, or examples of titles for those two levels? Three and four? I. It could
Denis: very well be that, um, they would have m e s or e R P in their job description, but probably not in their job title.
Denis: So I would aim at some kind of IT manager or or supply chain manager. Hmm. Maybe to answer your question, why I am, I try to aim more towards a lower levels. 'cause that's the level two and the level one. That's where all the process data, like temperature speeds, forces, and so on reside. Yep. Um, and I think those, that data is most useful for AI because it allows you to improve operations and that's just where all the money is.
Denis: If you can improve production, that's where all the big cost savings can be made. Right.
Jonathan: Yeah. That's where you would do your work. But I'm talking, but positioning is about attracting the buyers. So, It might be that there are buyers at that level, but I would tend to doubt it. It's not usually the case, but you could correct me if I'm wrong.
Jonathan: Mm-hmm. When I see, when I see manager in a job title, that's probably lower than, I mean, it might be a good place to start, but it would be better if it was like director or higher. Okay. It might just be different in, in your locale, or it might be different in this industry. I haven't talked to somebody in this space in a very long time, so it mm-hmm.
Jonathan: It could just be different, but you know, a manager doesn't, in the US at least, doesn't generally have the kind of budgetary control or autonomy that you would want. They're almost always gonna be having to not not be able to make the decision and then they have to make a case on your behalf to someone to approve the expenditure.
Jonathan: And I would rather have you talking directly to the person who can decide and allocate the funds.
Ah,
Denis: that's a good point. Yeah.
Jonathan: Yeah. So, so let's just say, who would an IT manager report to? Like if you think of a place that you just left, who would the IT manager report to? C i O or someone in the middle?
Jonathan: So that would be
Denis: the vp, like the VP of operations. 'cause I tend to focus more on the concrete side. Um, 'cause it, in the end of the day, they will not really benefit from my, uh, solution. Right. So maybe they would not want to buy it. Like if I'm managing email and wifi, I'm not really gonna be interested in buying a software that improves, let's say throughput or efficiency of the machine.
Jonathan: Think, right? Yeah, yeah, yeah. Operations might be a better way to go. So if so then who would the VP of operations report to that would be?
Denis: I think that's already pretty high. That could be the, the c e O could be the next level actually. The vp? Yeah. In our company where I actually left, he had the VP of operations and he reported directly to the C E O
Jonathan: C E O or c o
Denis: o, uh, c e o, chief Executive Officer.
Jonathan: Okay. There was no C O O I think.
Denis: No, there was, but um, that person was really on the IT side of things, not the OT side of things.
Jonathan: Really. That's, that's kind of surprised by that with such a physical business, but, okay. Alright. Then I am sold on the VP of operations as a potential ideal buyer because they've got close a access to the c e o.
Jonathan: What you do is so technical. The c e o is probably the right company for this is probably somewhere where the, the c e O has announced a vision for a. Let's just broadly speaking, say like modernizing the operations, digitizing the workflows, uh, for some very, very strong business reason. Oh, there's
Denis: actually, sorry to interrupt.
Denis: There is a very specific term in our industry, um, about exactly that. It's called Industry 4.0. Maybe you've heard of about it. Oh yeah, I have heard of it, yeah. Yeah. So the idea was it was actually invented 10 years ago to be a very quick revolution in terms of, but apparently it wasn't so quick because a decade later we're still almost nowhere.
Denis: Right. That's like everyone knows Industry 4.0. It's just that AI now became part of it. I
Jonathan: see. Yeah. 'cause I, I know that from the context of iot. Yeah, exactly. Okay, so, alright, so it's interesting because you have a perfect buzzword, but it's old. So it might, it might be maybe AI will revive it. It sounds like that's what you're saying.
Denis: It's still widely used. Everyone is still saying industry. Uh, 4.0. It's still on LinkedIn. It's all over.
Jonathan: Okay. And what, why would a c e O who would be the person to drive that vision and say like, okay, here's the vision we're gonna, we're gonna actually do this now. That maybe, uh, maybe the, maybe the thinking is like, AI actually makes this feasible.
Jonathan: Uh, so we're gonna finally make the leap into industry 4.0. What, what is the promise of doing such a thing or making such a, I'm sure a big investment of time, money, other resources. Mm-hmm. Retraining your employees. There might be retooling, uh, you might have to put sensors all over the place. Maybe you didn't even do that yet.
Jonathan: And what is the promise? I mean, it's gotta be increased profits, but is there something more specific than that? Yeah.
Denis: So one very. Big thing nowadays, which has been quite hyped up was called preventive maintenance. So to give you an idea, when our hot mill, which is one of our machines, uh, stands still for one hour, that costs us about $100,000.
Denis: So if you can somehow prevent machine from breaking down or repairing it before it breaks down, then you can prevent such an cata, uh, catastrophic, uh, standstill and save a lot of money. So preventive maintenance is one of the promises of ai.
Jonathan: Okay. Did you call it a hot mill?
Denis: Yes, a hot mill. We have a hot mill and a cold
Jonathan: mill.
Jonathan: Nice. I I love the jargon. It looks just so great. Okay, so, okay, so that's a great promise. How would you, if, if A C E O was thinking about. Who would be tasked with, with, uh, okay, let, let me back up. I'm gonna paint a picture, right? Mm-hmm. So some manufacturer, a, B, c, aluminum is having in increasing number of hours of downtime.
Jonathan: Mm-hmm. And the c e o flips their lid. They finally have had enough. Why can't you keep the Hotmail running? And who do they yell at?
Denis: So that would be the VP of operations. 'cause at the end of the day, it's his or her job to have as many productive hours as possible.
Jonathan: Okay. That's what I figured. So this, this really sounds like the right person.
Jonathan: I, I think Okay. But we've just gotta trace it down now. So we've, we've gone up the organization to somebody who probably in this nightmare scenario, they're getting yelled at by the c e o and they come back and they say, all right, I've, I've, you know, bring me solutions vp and the VP comes back and says, I've got a solution.
Jonathan: How would you, let's say they magically found you and you came into this situation. Is there something that, is there something that you can do? To move this needle in the near term, or are you kind of way downstream from this? Like you can kind of help as a component of the thing like, like how much of a contribution could you make to helping this VP of operations calm the c e O down?
Denis: So I think it's more the latter. That was one of my fears. I think it's unrealistic to develop complete solution, meaning from data infrastructure all the way to the final model for one person. Yeah. So I think I'm more of a part
Jonathan: of solution. All right. So here's a, uh, maybe there's another way to ask the question.
Jonathan: What would they be considering. Like, what would they Google for what, who are the 400 pound gorillas in this space? If that doesn't translate, who are the obvious people that they would call? I mean, I guess we're, we've implied already that they're not, they can't afford McKinsey or Deloitte because that's why you picked this size business.
Jonathan: So let's rule them out. Mm-hmm. And what are the, what are the other things that they would consider? Is there anything off the shelf? Is there a bandaid? You know, is there a quick fix or, um, a sort of deferred maintenance, technical debt thing that they can do? Um mm-hmm. Like what are the different ways that they would try and placate the c e o?
Denis: Yeah. So my first job was actually at such a medium-sized manufacturer, and what they did is that instead of looking for outside resources, which are inexpensive, they also don't know that they're soloist. So it's either hiring like a Deloitte or McKinsey or hiring no one. What they would do is, Basically yell at the process engineer who is responsible for that particular machine and ask him or her to fix it.
Denis: It's actually a second good person to contact because their bonus directly depends on the performance of the machine.
Jonathan: What was that title
Denis: again? Process Engineer. So basically my
Jonathan: Oh, your old job. Yeah. Uh, I like, I I like this. This could be our Trojan horse into, all right, so, so the process engineer whose language you could surely speak.
Jonathan: Oh yeah, that's
Denis: actually the reason why I focus this.
Jonathan: Yep. So the process engineer. Alright. And the, and the pain that you help them with is you can help them with preventative maintenance. You can help them keep the machines running around the clock or whatever the. Whatever that, yeah,
Denis: I guess our first step would even be just to diagnose because what happens is that he gets yelled at and then has to look through a bunch of Excel files or databases to try to figure out what's wrong with the process.
Denis: Mm-hmm Basically sifting through lots of data and wasting lots of time 'cause they not, don't necessarily have the right skills in the right infrastructure at their disposal, do this work efficiently.
Jonathan: Interesting. Okay. Now the process engineer, I mean you could surely answer this because you were one, like what are we talking about salary wise and responsibility wise for someone just in a range.
Jonathan: You don't have to say what you made, but it's pretty high paid
Denis: professional. They're quite rare because they need an engineering education, uh, awesome experience. So it's one of the highest, highest state employees who is like not a manager.
Jonathan: Mm-hmm. So are we talking like six figures at least?
Denis: Yeah, six figures for sure in the US for sure.
Denis: Okay.
Jonathan: And. How many of the, if you went on LinkedIn right now and you Googled for process engineer in search for process engineer, you know, like these kinds of people, how many would you, do you think you'd find? How many do you think there are?
Denis: Uh, quite a lot because for each machine center you need one.
Denis: So for each department you need at least one. And I'm quite lucky in the sense that I try to always keep a good network of ex-colleagues and friends. So I know quite a bit of growth engineers myself personally. Mm-hmm. Mm-hmm.
Jonathan: Okay. I know I'm jumping all around here, but this is kinda the way it works.
Jonathan: How often, on average do you think they're worried about. What, what's this? Downtime, I guess. Downtime, right? Mm-hmm. Is this, is this a constant every week kind of thing that they are, their top priority is to mitigate downtime? Or is it just something that, you know, once or twice a year it explodes and they get screamed at and then they scramble and, and then keep their fingers crossed, do whatever they can with the Excel spreadsheets to hope it doesn't happen again.
Jonathan: Like is this, like frequency wise, how often does this happen?
Denis: Well, it's a very stressful job because downtime is perhaps not as frequent. It can happen a couple times a year, mostly not, but they have plenty of other problems. Like, for example, delays, production delays, um, mistakes in planning or just, uh, problems with quality.
Denis: Okay. But they're constantly stressed out.
Jonathan: What are the top three things? Stressing these people out? Delays
Denis: in production because of parts that don't arrive. Uh, mistakes that people make, things that break down, but also quality. Like if you produce aluminum, that my first job, for example. Mm-hmm. The process can go wrong in a lot of ways and the quality of your aluminum is not a constant.
Denis: It depends on things like impurities and so on. That's all something that they have to monitor.
Jonathan: Mm-hmm. Okay. So production delays, quality control, and was there a third
Denis: I guess? Yes, uh, um, safety. So one, the top priority of every plant always has to be safety of um, employees. Mm-hmm. So they take it very seriously.
Denis: So whenever someone like falls or sprains an ankle, It's actually a huge deal because they have to document it, they have to do interviews, they have to see what they can improve. Okay. So safety is
Jonathan: also very big point. Okay. So of those, so, so I've got four things here. Downtime, it's rare, but it happens.
Jonathan: Production delays, it's probably more common. Quality control is probably a constant, and then safety is a constant. So of those, uh, I'm guessing you can't help with safety, or is, is that not true?
Denis: Well, it's not something that I can necessarily help with, but it's, you have actually a safety engineer or even like a VP of safety at plants that just do that.
Denis: Okay.
Jonathan: All right. So I'm gonna cross that one off. The, the downtime production delays in quality controls seem like they're right at the alley of someone like you, right? Yes. Because data plays a crucial role there. Okay. So just taking some notes.
Jonathan: All right. Does this person, you, again, you'd know this since you were one, and I can't even over, I can't overstate how useful it is to have been one of these people. Um, do you have a budget that you control?
Denis: No, I don't because I was more of an, let's say I was hired as a process engineer, but they actually put me in front of a computer to program stuff, so I had no responsibility over, let's say people.
Denis: Mm-hmm. That was one job that was offered to me. I refused it because I wanted to be a software
Jonathan: guy. Mm-hmm. Okay. So, but generally speaking, do these folks have control over a budget
Denis: I should ask? I, I am not, I'm afraid to answer the question. I'm not sure. Mm-hmm. But I think they do.
Jonathan: Well, definitely the VP of O operations does, and that's who's yelling at them, so, so yeah, absolutely.
Jonathan: It would be nice if they did, because you could probably, I mean, I can remember, uh, my one corporate job that I had, a director level boss who. Could pretty much at will spend 5,000 bucks on whatever he felt like, like one, uh, once a week or once a month, some on some schedule. And it was like very clear, you know.
Jonathan: Wow. And, and he was like, what am I gonna spend it on? And he, you know, he was, he would have a lot of different potential things to spend it on. Anything from like a new giant rack-mounted hard drive or something, or tape backups. This was a long time ago. Uh, or, or whatever, whatever was the, the top of mind thing in the department that he was responsible for that had a chip in it Pretty much.
Jonathan: So everything from printers, you know, it's a different, different kind of person than a process engineer. But, uh, director level was, I think it was one level higher than, yeah, it was one level higher than manager, but he did have a certain amount of control. And then if he needed to, to get something approved, he, his direct boss.
Jonathan: Could approve things, you know, like I feel like I remember talking about 30 to $50,000 usually for hardware. But, but you know, that kind of thing was his responsibility and he, and if he wanted it, he got it. Okay.
Denis: It's actually, it made me think of it because there's a level in between the process engineer and the VP of operations.
Denis: Mm-hmm. And that's the department manager. Because engineers work in departments, for example, a cost house or an electrolysis department. And each department has a head. And this person for sure has a budget. I actually know a person like that personally. Mm-hmm. And he can just swipe, uh, swipe a cart and buy a project or a tool.
Jonathan: Yeah. Right. And they probably have some limit per month. Yes. So, okay, so, so it could be the department manager or the process manager. Somewhere around there. Maybe it's, it probably, I'm thinking less so it's VP of operations. Although if you, for example, had an opportunity to speak at a conference, you would want a whole bunch of VP of operations in the room, even though they might not be your ideal buyer initially.
Jonathan: But for working your way in, it seems like it, it could be true that operating in familiar territory, we've got a, still have a lot of connections, would be an easier way into the, into the building, so to speak.
Denis: That's interesting. I never thought before of separating who I contact to get a foot in the door and the person I'm actually helping or selling to.
Jonathan: Yeah. I mean, ideally you, it would be the same person, but sometimes it's just very, sometimes they're just. It's impossible or infeasible to get direct access to your buyer. So you need to go through some sort of internal recommend recommender, uh mm-hmm. Someone who, that, that was how, you know, with, with mobile, that's the way it worked with me.
Jonathan: I, I wrote a book for developers, other developers bought it, and then when the boss came rolling in and was like, Hey, we gotta do something about mobile, who should we call? Do you guys, do you guys and gals know how to do it? No. But this guy does. And they all pick up the book and say, we should call this guy.
Jonathan: So that's a riskier, longer term approach. It's not as direct and it, it can be problematic. It works. When it works, it works great. Um, because you're the name on everybody's lips when someone asks for a recommendation. Mm. Mm-hmm. But you could, in theory, well it sounds like through your network, you could find, you could pretty easily find out what is a department manager's sort of budgetary what, what's their discretionary budget?
Jonathan: And then design something around that, that it would be easy for them to swipe a credit card. So if it's 5,000 and you had some kind of productized service that you priced at, uh, oh, I don't know, 4,500 and it promised certain outcomes or certain, you know, when I say outcome, it could be clarity, it could be recommendations, it could be a roadmap, it could be an audit, feasibility audit.
Jonathan: There's a bunch of things you could probably do for that amount of money that would, that would be a great way to add some value, give them positive r o i, and then perhaps they say, you know what? We love these recommendations. Could you give us proposal to, to do it, to execute against this? Uh, or at least help us do it, like oversee, help, you know, be available in an advisory capacity as we, I don't know, deploy a sensor network or, or whatever.
Denis: I think that's a great idea indeed, because I think it's gonna be very hard for me, like without a book, without a, without a big reputation to sell a big consulting gig. So I think, uh, getting my foot in the door with some kind of product I service, like a workshop. I've done those before. So I've done some courses.
Denis: Cool. Okay. So
Jonathan: cool. Okay. So what, uh, I don't wanna get into the products y quite yet, but we could, so what are, what are some of the product life services you've considered so far? So you've done a workshop?
Denis: Yeah, so I've basically sold two programming workshops about by and r but those were, one was for students of a university, and the second one was basically, well also for students, honestly.
Denis: Mm-hmm. The second product I sold was basically for an old colleague who's also a freelancer, and he just needed a tool to visualize a bunch of data. Mm-hmm. And he just paid me like, ah, no, Dennis, he knows the data, so I'll just pay him an amount and he'll build it for
Jonathan: me. Oh, so it wasn't training? It was, it was,
Denis: honestly, it was a favor.
Denis: You know, I think the guy just really liked me. Mm-hmm.
Jonathan: Okay. Yeah. Alright, so let's see here. If the, let's just assume that both the process and engineer and the department manager, I think you called them mm-hmm. They are interested in min in decreasing downtime, decreasing production delays, and increasing, improving quality or, or improving quality control.
Jonathan: Mm-hmm. What can you do to help move those needles? What can you contribute to, to their situation that would keep that promise so to speak? It's like if you, if they come across you and you're like, I can help you decrease down time, increase, uh, decrease production delays and increase quality control, and they're like, well, that's okay.
Jonathan: How, and what would you, what would you say to the how.
Denis: Well, the first thing I would mention is that they probably underestimate the effort it's gonna take to obtain the data. So I would basically try to explain that it's not a data science problem, it's more of a data engineering problem, that before they start to talk about AI or maintenance, they really need to focus on their data infrastructure.
Jonathan: Okay. So that you started off a little bit down the road from where I was. So let's back up a little bit from there. So they say, uh, they say, oh, hey, we, we, someone recommended you, we saw your website. We understand that you all it says is that you can do those three things. You can improve those three metrics for medium-sized aluminum manufacturers, and that's us.
Jonathan: Mm-hmm. Or even more specifically for department managers of medium-sized aluminum manufacturers. Alright, great. How do you do that? And then just how, how do you des how do you, even if you weren't gonna contribute to it, like let's just, let's start here. What are the steps? Or maybe what are the questions that you would ask to start to, to qualify them as a lead or qualify their, their current situation?
Jonathan: Like maybe they do have great data. So what, like what are the questions that you would ask? So I
Denis: would try to get as much information I can about their current data infrastructure. So what kind of systems they have. I probably have know already what they have, because it's pretty similar in most plants.
Denis: Um, I would try to ask them if they have already had any experience with building AI projects, like data projects and what tools they use to currently monitor the process. But they may say like, oh, we use to do like, oh, the soft spy, or we have this data historian, and that gives me an idea of how mature they already are in their digital technologies.
Jonathan: Hmm. Let's, let's stay here. Let's, let's iterate on this. I'll be, I'll be you. This is gonna be hard because, uh, you, you might have to. Help me, but with my words. But then you be the client. Mm-hmm. Okay. And, and imagine in this process that, that you can probably, you can probably guess what they're using because most people do.
Jonathan: But let's just say, first of all, you definitely want to validate that. 'cause it could be an exception. But the other thing is you want to identify, you want to identify to them that, that you're just super familiar with the industry by asking incredibly smart questions, even though you probably know the answer already.
Jonathan: Mm-hmm. Mm-hmm. So, so they say, yes, we want, we, we'd be interested in talking to you about improving these three metrics. How do you do that? And you say, and you'd probably say something like, well, I can help people. What, uh, I help people implement industry 4.0 project or transformations with AI get started down the path to, to this transformation.
Jonathan: How would you, how would you put that.
Denis: So I think I would focus on the point that lots of the missed problems that happen, like delays the preventive, the, uh, machine breakdown mm-hmm. Is because people don't monitor their data. They don't keep an eye on it. So
Jonathan: to, so let's, let's do it that way then. So, so what would you ask me to find out whether or not we were organizationally we were doing a good job.
Denis: So, hold on. Are you still the client or are you now me? Oh yeah,
Jonathan: I'm now, I'm now I am the client. And you're you. Sorry. So how would you, how would you ask me to validate the maturity of the operations with regard to data? So,
Denis: well, let's focus on maybe the production delay. I would ask, well, why, like, how big is your delay?
Denis: Mm-hmm. And maybe they'll say, oh, well we are like, I. 50, I know it's behind, or we are 50 coils behind schedule. Mm-hmm. Or like two days behind schedule. But then at least I know that they have a number. Some companies don't even know how badly they are doing.
Jonathan: Okay. So let's say I said, uh, yeah, we're two days behind.
Denis: Then I would try to figure out why they think this happened, so. Mm-hmm. But that's a difficult question even for them to answer. It's could boil down to, oh yeah, well we had this problem, or the machine broke down, or, I'm basically trying to gauge how much they observe their own data.
Jonathan: Okay. So it's just someone looking.
Jonathan: Yeah. So just so, so just, let's just do it. I'll keep being the client. You asked me how long is your delay? I say, yeah, we're like two days behind. Then what would you ask without assuming nothing, just ask me a dumb question. So does
Denis: this happen often? I often two days
Jonathan: behind. Perfect. Exactly. So the answer is we're pretty consistently, at least a day behind.
Jonathan: Uh, right now it's a little bit worse than usual. Mm-hmm.
Denis: Well, then something must have happened. I guess
Jonathan: that's not a question though, so, so put it like, ask me the next question, right? Mm-hmm.
Denis: So can you repeat the, uh, what you said? Just Yeah.
Jonathan: It's like, did something change is kind of like the next question.
Jonathan: Did st. Mm-hmm. Did, yeah. Right. Did something change? Right. I don't know. The, the wizzy bob thing broke down. It, it's breaking down a little bit more. And then you would probably ask me something about, well, what are you doing for monitoring preventative maintenance and so forth. Yeah, exactly. Like, when did it
Denis: break down the last time?
Jonathan: Yep. Like, ah, I don't know. I think it's, I think it was, I don't know. It was maybe last two months ago, maybe. I don't know. Mm-hmm. This is, and then you say, where would you look to check? I.
Denis: Right. It is. How can you know the exact time it,
Jonathan: it broke down. Yeah. Is there some way that you could check and be like, well, we've got it in Excel document, but we don't really keeping that up to date.
Jonathan: I could ask the, the shop manager or something he might remember. Yes.
Denis: They have to dig it up basically.
Jonathan: Right, right. And so you, so you could, you could by asking these, like the smallest possible next question. It's kinda like troubleshooting a software bug, you know? Is the computer Yeah. Like a doctor's visit.
Jonathan: Yes. Is the computer plugged in? Okay, we've got that established. Is it turned on? Okay, we've got that established. Is your wifi enabled? Okay. We've got that established just like the dumbest, dumbest questions. But if you skip over them, then you, you get lost. You get, you find yourself like stuck. So, so if you just kept asking me dumb questions and you, it started to reveal that to me and to you.
Jonathan: That are that, that our data is inaccessible if it exists at all, and that it would be very difficult to dig it up. You could probably talk to someone and, and, and have that epiphany moment. Right?
Denis: Yeah. And actually it's also the process often leaves clues as to what's going to happen. So often post-fact, you can actually see at your parameters of values and could have known that something like bad was going to happen.
Denis: Mm-hmm. But because people weren't looking, it happened without noticing.
Jonathan: Right. So, okay, so if you got to this point where we are now, and you were like, okay, their records are perhaps not that great, perhaps not that up to date, and they're definitely not looking at them, they don't have quick access to them, then I would say something like, do you think it would be useful?
Jonathan: Do you think it would decrease downtime if you had, I. Real time, instant access to certain parameters that, I don't know, maybe you could suggest, like, do you think that that would be beneficial? Yeah,
Denis: and I think just, sorry, uh, to think of a story, uh, to get more concrete maybe. Mm-hmm. We had once this problem in our process where once sensor broke down, uh, and no one really paid attention, but the sensor basically told, Hey, the feeder to this particular, um, oven is not working.
Denis: And so the oven wasn't fed for a number of, I think for two days, which caused like a major problem down the line. If they had just kept an eye on this one sensor, they would've known that there's a problem. It would've prevented the, the issue,
Jonathan: right? Yeah. I mean it's, it's, it's obvious when you put it out.
Jonathan: So when you, when you put it like that, but there's some reason that they aren't looking at it. There's too much work to do. Alright. So why don't you hire more people to kind of have someone who's in charge of just manually updating this stuff and sending a report out daily to everyone? Well, they do
Denis: this like an intern, which was my very first like pre-job, so to speak.
Denis: Mm-hmm. So, and I actually helped to install a system that basically, and some mail the daily reports of some very important KPIs.
Jonathan: Mm-hmm. Oh, that'd be another great question. Do you know what your most important KPIs are? You know, if you had time and energy? Mm-hmm. And, and processing power to track KPIs more closely, maybe closer to real time, or at least daily?
Jonathan: Do what ones would you look at?
Denis: Yeah. There's actually one major one which is very important for each plant. It's called O e E. It stands for operation Equipment Efficiency. Mm-hmm. And in a nutshell, it basically denotes how quickly you produce parts of good quality, which potential the time. Okay. It's like a number between zero and 100%.
Denis: So zero being you produce nothing at all and 100 produce only good parts as fast as possible with no downtime.
Jonathan: Mm-hmm. Okay, great. So you could ask, I mean, it would, would literally every department manager or process engineer be aware of this metric?
Denis: No. Uh, some plants even don't calculate it.
Jonathan: Oh, interesting.
Jonathan: So you could say something like, uh, are you, are you calculating your, whatever you said it was? Yeah, I
Denis: could, for example, I would say, Hey, what's your o e e? And they would say, well, we don't know. And I'm like, oh, I can build it for you because you
Jonathan: have to know. So So you're jumping too far. Yeah. You gotta, you gotta like lead the horse to water though.
Jonathan: So they, you know, they could say, we don't know, or we don't know what that is. Hopefully they would say, we don't know. Like, that would be great to know, but we don't know it. And then you, let's say that was the answer. That would be great to know, but we, we don't know it, we don't keep track of it. And then, then you could go down the path of why aren't they, so how just you just say, how come you're not.
Jonathan: How come you're not tracking that and find out all the obstacles, right?
Denis: Mm-hmm. And I guess they would say something like, yeah, well it's just hard to get the data. It's all those different sources. Mm-hmm. It has to be joined and we don't have anyone. But basically it's it's job to do it, but they're not doing it.
Denis: It'll be something like
Jonathan: that, I think. Hmm. Okay. Yeah, that's probably, that sounds pretty classic. Alright, well, whose responsibility would it be to fix this issue? So like, it is not doing it. You could, you, you need it. Is this something that you could take responsibility for? If, if you had someone to do it, you know what I mean?
Jonathan: Like, so that's what I would, I
Denis: was ask them, right? Yeah. Mm-hmm. Right. So the ideal person to do this job would be the process engineer, because he has the knowledge. Unfortunately, he is actually not, it's not really an office job. He's part-time in the office, but he's also very often on the concrete side of the building because when something breaks down, he has to go there.
Denis: So ideally this person has a lot of ideas that he wants to implement. It doesn't have the time. Mm-hmm.
Jonathan: Okay. And so that's where you start to, so like that point, it'd be interesting to be having this meeting with the department manager and the process engineer in the same call. Mm-hmm. And have the process engineer chime in about, you know, in, in answer to some of these questions.
Jonathan: Like, how come you're not tracking it? And the processing was like, it's just, there's just too much to do. I gotta run over to the, the floor and like look at stuff manually and dig through Excel files. And he said, well, if there was a solution to that, would it be useful? I mean, it, it's, I don't wanna put words in your mouth, but it seems like it'd be super useful if you could track this number mm-hmm.
Jonathan: On a daily basis, right? Mm-hmm. And you see if they nod and then they're gonna say something like, or you could prompt it, what would that entail? Like how would we even do that? And then you could say, not pitch, but you could say, well, step one would be we need to find out if the data exists at all. It doesn't always, maybe it does here, maybe you know that already.
Jonathan: But the data, first of all, doesn't always exist. So we have to validate that it exists. We have to validate that it is correct. We have to validate that it is in a format that we can consume in some kind of automated fashion. And, and then the next layer up from that is synthesizing the data. Maybe AI comes in at that point, uh mm-hmm.
Jonathan: Maybe it's later, but then you could just quickly lay out like three to six steps in the data pipeline, or whatever you call this from. Mm-hmm. Let's say from the sensors to a dashboard, like all, like, what are the six stages that you'd need to see, kind of have like a, a report card or a health checkup for each one of these steps and be like, well, how do you, how would you do if, if I asked you, do you have all the data you need that, you know, on a scale of one to five, would you say yes or no?
Jonathan: You know, on a scale of one to five, would you say you, you probably are capturing all the data you need and they could rate themselves there? Like, do you think it's clean and complete and accurate? They could rate themselves there and then whatever comes after that, you know, you could maybe, like I said, have three to six stages where they could kind of mm-hmm, mm-hmm.
Jonathan: Do a self checkup with you right there and say like, well, you know, and what's probably gonna happen, at least in other industries, what happens is a good buyer for a product I service, which is what we're talking about here, a good buyer will not know the answers to these questions and not even know or have the time to get the answers, and they're gonna feel stupid that they can't.
Jonathan: So you could say something like, and, and maybe this was the whole premise of the phone call in the first place, so you're not springing it on them, but it's like, well, this is what my, my audit, you know, my $4,500 audit does. Like I will come in and fix all of these things. But what you've done is you've walked the slowly, carefully into this place where they're forced to admit to themselves that they've been dropping the ball.
Jonathan: And that things really could be better. And they know it. And then, Hey, here's a, and, and this is not coercive or manipulative, it's all true. You're just asking them questions and they're just forced to admit to, uh, sort of third party that things are not in order. Which I would just, I would think for a manufacturing company in Germany would not be the kind of thing they'd be proud of, so, right.
Jonathan: So it's like, well, absolutely, you can rate me a check or you can swipe a credit card for 4,500 bucks and I can start on Monday and here's how it would work. And you just say, mm-hmm. I would meet with these people. I would do this, I would need access to this. And then at the end, you know, B B B B B, these steps, it'll probably take X number of out, roughly x number of hours across the course of three weeks.
Jonathan: I would need time with these of your employees. I'd need about this much time from you folks. I'd need this kind of access. And then at the end, I. I'll have a roadmap for how you can get from this place where no one's looking, you know, no one can hear the fire alarm going off. Hmm. You know, you, it's like having no batteries in the fire alarm.
Jonathan: It's like, well, we've got fire alarms. Yeah. The batteries in them. You know, so it's not doing anything and I can get you from, from not knowing what to do about it, to knowing exactly what to do about it and having sort of level of effort, perhaps even with a budget and a timeline for what that project might look like.
Jonathan: I mean, that's, that's the output of a good roadmap or whatever you call it. Mm-hmm.
Denis: I really like the approach of asking questions to let them get there. 'cause the problem, if I would just arrive there and say, Hey, your data is bad, they would be probably offended and Yeah,
Jonathan: totally. Right. You need, you need to just ask questions like a doctor.
Jonathan: A doctor's not gonna be like, oof, you need to stop eating as much. Right. You're eating way too much. Obviously. You, you just don't, that approach, it's not, uh, it's not that useful. It gets people to get defensive. Mm-hmm. Uh, especially if the CEO's on the call too. Right? So like, you don't wanna be pointing fingers, making assumptions.
Jonathan: Oh, yeah. You want to ask, you want to ask the smallest step, the smallest next step question possible. What I was calling stupid questions before. It's like, it's like, it is gonna sound like a dumb question, but you know. Mm-hmm. Are you guys tracking o e e? And they'll be like, yeah, not really. I mean, we collect the data, but then we never aggregate it.
Jonathan: And even if we did, no one ever looks at it. So, so, so yes and no. Mm-hmm. And then you can just say, okay, well, It, would it be a priority to, I mean, if you, if it was easy, would you look at it, you know, if you're getting alerts and stuff like that. Like what if it, what if it jumped in your face when it was hitting the red zone or whatever, and like, oh, that would be amazing.
Jonathan: And it's like, okay, well, you know, anyway, we already did that part. But, but it's real, I think it's really important to phrase everything in this kind of disarming, innocent question kind of way. That might be a better word for it. Just like innocent questions. I mean, I know it's an American thing, but I always point to this old, old TV show called Colombo, where this guy Colom, Peter Faulk was this Colombo guy and he would just, he was a detective and he would just a, he was just such a disarming, goofy character.
Jonathan: And he would say like, I, I just don't understand. I can't understand how you, can you help me understand why would, why would somebody put an eye hook in the ceiling right here? I just don't understand it. You know, and you just sort of like, not trick, but like, I mean, we're talking about criminals here, so you just kind of trick them into, or back them into a corner, really.
Jonathan: In, in, in a sense, in a model like this, you're backing the client into the corner of, of really facing reality so that they can, you know, then decide potentially to do something about it if there's an easy solution sitting right in front of them. Mm.
Denis: Mm-hmm. Yeah, I think that role fits me quite well because I'm actually not that experienced, so I can play it pretty naturally at
Jonathan: Oh, good.
Jonathan: Yeah. All right, so if we, now that was as a whole, that whole conversation was a sales conversation, but it helps me to understand what you might want to say to position such an offering. So if the pain here is the, the metrics we listed, downtime, production delays, quality control, and the dream is a hundred percent uptime, no production delays, uh, 10 outta 10 on the quality control Or like an o e e of a hundred or whatever you said.
Jonathan: Yes. Right. It's like then that's the dream. And then, and then the fix is you cannot get from where you are to where you want to be without, and then you, you tell me what you would call it, but like a real time, like data pipeline. I think time series data plays into this somehow. Like you need data. You need data.
Jonathan: Yeah. I would call
Denis: it a proper data infrastructure that unites both time series data and like relational data. Mm-hmm. Okay. In one place.
Jonathan: Right. So you, you need a, a unified access or some kind of, however you decide to say it. It's funny you mentioned unified
Denis: because the, there has been a solution in industry, uh, being hyped up as the unified namespace Hmm.
Denis: As the central location where every data, all data basic collected, you have to imagine basically we have all different systems, level one, level four. They all are very different. So the problem as we try to solve was basically by combining, by basically connecting all this data to one central database, which is called Unified Namespace, that basically organizes all this data in a neat structure, which is easy to understand.
Denis: Mm-hmm. Or human. Mm-hmm.
Jonathan: Okay. Yeah. Perfect. So that's the solution. But it, that might not be, that might not be the solution you provide. So there's a, there, but maybe it's discussed and now we're kind of talking sales piece. I
Denis: think still be, it'll be, for example, it can be built incrementally. So like, for example, build a small, unified namespace small database.
Denis: Right. Connect just the systems that they need for their particular use case. Mm-hmm. And it can grow.
Jonathan: Perfect. Yeah, that's perfect. So for the product I service, product I service would be this sort of diagnostic with recommendations and then the, the probable next step. If they're, if you still wanna work with them and they, they, there's still a good fit would be to help them in, in some capacity, implement this unified namespace, like an M V P minimum viable product, like the, the 80 20 rule.
Jonathan: You're just gonna build the top three things or whatever, the smallest version of this, a phase one. And you could, you could give them a, a three option proposal of three different ways that you could work with them to make that a reality, to get them to that phase. And then once that's over, then you know, move on to some other, perhaps phase two or different relationship, or you move on to a different client.
Jonathan: Mm-hmm. But, but so from a positioning standpoint, if we go all the way back,
Jonathan: Is really the title is department manager. Is there there, is there no name of department?
Denis: Yes. For example, I worked at the electrolysis, so electrolysis department and, oh sorry. It's called electro department manager. So it would be like the electrolysis manager. Actually, lemme Google this personal LinkedIn just now maybe.
Denis: Yeah,
Jonathan: check that out.
Denis: So his name is Chef de Electro in French. So, no, it's actually really like the chef of the electrolysis department.
Jonathan: Oh, chief, is it Chief?
Denis: Yeah. Well it's in French, so tough to say. I, yes, it's like Chef, it's like head. Head
Jonathan: would be the translation. Okay. Head of the electrolysis department.
Jonathan: And is that, are there lots of different, like do process engineers report in different departments or would all of 'em be in electrolysis? Sorry,
Denis: the name is Reduction and Gas Treatment Centers Manager. So I guess he's responsible for both the reduction department and the gas treatment department, and he's like the manager of that department.
Denis: Hmm,
Jonathan: interesting. Okay. So this, so sometimes it is the case that the, that there were lots of different titles for the kind of thing that we're looking for. So I don't, it might be a little bit difficult to come up with one or even a handful of them, especially if we're like dealing with language translations on top of it.
Jonathan: But the idea is that, that whoever process engineers reporting to is probably your ideal buyer for now. Mm-hmm. Mm-hmm. And, and you could perhaps network to them through process engineers or, or just through colleagues, past colleagues and, and start asking these kind of innocent questions about mm-hmm.
Jonathan: Mm-hmm. O e e, you know, like just, I'm doing a survey, I'm surveying the, the space. And for, for aluminum manufacturers of your size, I wanna find out how people would sort of self-report their O E E I almost said compliance, their o e e awareness and get a sense, like on a scale of one to 10, how often do you check your, I don't know how to say it right, but like, how aware of it, how aware of your o e E measurement are you, and how accurate do you believe it is?
Jonathan: And how important is that to your day-to-day decision making? So something like that. Just a quick little survey and maybe do that. Do a LinkedIn survey and share it with just people in your network process, engineers, anybody that works, really anybody that works in one of these types of mm-hmm. Target businesses.
Jonathan: And find out the answers to those questions. It'd be really interesting if, even if you got only 20 replies, it'd be really interesting because that might be, The, the words to choose to, to begin the conversation. 'cause if that's a key metric and not even everybody's tracking it, and that would be the first thing that you would probably do, or I'm putting words in your mouth, is that the first thing you would probably do is try and make that number more accurate and available.
Jonathan: Yeah,
Denis: absolutely. Because basically this number denotes how much money that the plant makes.
Jonathan: Cool. Hello? Got that. Made my eyebrows go up. So yeah, you could potentially sort of be an evangelist for the this target market to get better at tracking that number more accurately and quickly in real time. And then find all the objections.
Jonathan: So surface all the objections. Like, like if you didn't rate yourself a 10 outta 10, like why not? Or anybody below like a, a a seven. Like, oh, you know, do you have an initiative in place to improve this number? Or is it no big deal? Or, you know, what, what are the objections? How come you haven't. How come you're not working on it?
Jonathan: If you're not, and if you can find out those objections, then you'd be able to address 'em in your marketing and be like, oh, you know, process, process engineer's too busy. Well, you know, you, you could hire someone to just, you could hire a consultant to come in. It's like, well, McKenzie's too expensive. It's like, well, what if it was only 4,500 bucks and it was only gonna take two weeks?
Jonathan: I was like, well, that's pretty interesting. Yeah.
Denis: One question about that number. Um, what, in the case, if I'm wrong, and for example, if maybe not for some reason, not our biggest concern, or maybe they already have it, would it make sense to, if I was to send a bunch of emails into asking for interviews, or should I include like, maybe other metrics or other problems, or should I really focus on maybe only one metric?
Jonathan: All of this is hypothesis. This is, this is us brainstorming and. It's just ideas to start conversations and see if anything clicks. Sometimes they're like, sometimes not most of the time, but also not, it's not super rare. You might immediately get like, oh yeah, yeah, yeah. That would be amazing. You know, you, we like, we nailed it in the brainstorming session.
Jonathan: Mm-hmm. Most of the time it's close enough that they're like, well, no, not really, but the thing that we really care about is this. Okay. You know, so then you would use this to uncover what they really care about. So there's a million tactics. We could do that to find that out, but I kind of wanna zoom back up unless you have like a quick question.
Jonathan: No,
Denis: that's fine. It's just that, it's also new to me because I'm very technical, right? So I'm used to knowing everything upfront with certainty, and it's basically stepping out of my comfort zone to try to work with hypothesis, right? Yeah. Trying to
Jonathan: experiment. Yeah. This is exploratory, right? It's like you're, you're.
Jonathan: That's, that's a great point actually. I was just talking to someone with someone about that. I think it might've been Rochelle on the Business of Authority, and we were talking about how this sort of, it, it can be uncomfortable. Yeah. It was that episode we, we were talking about the difference between exploration and expedition.
Jonathan: So when you know where you want to go, right, you have the answer and you just need mm-hmm. Need to get there and like, there's gonna be a million decisions, but you know where you're going. That's the feeling that you're used to. Mm-hmm. Mm-hmm. And yeah, but you can't, it's too many assumptions in that when you're working with a brand new client, way too many, you need to go through an exploratory phase first where you find out the specifics of their, their weird idiosyncrasies and nuances and, and desires and drivers and hopes and dreams and fears and all of that stuff.
Jonathan: And you can't, at this stage, maybe, maybe in five years when you've done. I had a hundred clients like this, you could just build a product and be like, look, here's, here's the state of the industry, because I'm in a position to know, and here's what everyone's doing wrong. Here's what everyone needs to do better.
Jonathan: And like at that point, that's a different game. You know, you can do a cold read, just immediately skip over all of these steps, but right now it's more exploratory when you first meet someone. So it would be, and it can be uncomfortable as you pointed out, but I, it's very, very effective.
Denis: I'm glad to hear it's a viable part because I had this, I guess, false impression that I was supposed to know everything if I'm hired as an expert.
Denis: But that's not the case. I just think,
Jonathan: well, you have to find out. So you can be the expert at what you do, but you're not the expert of their business and you're not the expert of their organization. There's no way for you to know that you can make guesses, but you can't know it. Mm-hmm. You don't know who the troublemaker employees are.
Jonathan: You might find out that someone's sabotaging the data. I mean, you just don't know. There's just no way to know from the outside. You can make assumptions and use those to ask questions that kind of lead the, the client down this path. But it's way too soon to prescribe anything. Always think like in a sales interview every single time, I'm like a doctor.
Jonathan: Try and be like a doctor first. Do no harm, do not assume that you're going to give them a triple bypass just from looking at them, because a lot of people need triple bypasses and you know, you're a person so you probably need one. So it's off-putting and it, it's often wrong. Things are much more nuanced than that.
Jonathan: So, yeah. So in this sales interview, you'd just be very, or even any conversations leading up to creating the product I service. Whenever you're talking to people, just ask these sort of innocent, maybe slightly naive questions in a self-deprecating way. It's like, ah, I know this is gonna sound like a dumb question, but I, I don't wanna make any bad assumptions.
Jonathan: And, you know, how would I know and what it's like inside your organization. So, you know, with that context. And then you, you could just ask some, you know, questions that are probably, that you might already know the answers to, but you wanna make sure you do. Yeah, that sounds great. That who does? Cool. So, okay.
Jonathan: So like, our reason for being here though is positioning. So if we, if we just thought of a hypothetical positioning statement, could be something like, I help process, I, I wish we had a clearer title. We, let's just use VP operations for the time being. I help VP operations for medium, medium-sized aluminum manufacturers increase their o e E score.
Jonathan: I mean, that's not great, but that's the shape. That's a shape. Mm-hmm. Another one could be, I don't like this as much. It's pretty, it feels kind of tropey or trite, but like, I hope medium-sized manufacturers make the leap to industry 4.0. That's, that sounds so bad to me compared to the other one. Yeah, no, I wouldn't
Denis: do that one.
Denis: Honestly. That's,
Jonathan: yeah. It's too same. I'm sure there's a thousand people with that same positioning. You could say something like, you could pick one of the three big things like downtime. You could or, but no, I don't like that one. Production delays is probably much more common. You could say, you know, oh, like I help process engineers or their bosses or, or people.
Jonathan: People who apply process engineers, decreased production delays at medium-sized aluminum manufacturers. It's like, huh? That's pretty specific. Mm-hmm. You know, so the, these are kind of like the first half of an L F P or, you know, it's kind like the, the, the core value proposition of an L F P Ss. We haven't really said what your title is, and I kind of don't care about your title and I almost feel like in email you said that they don't know what data engineers are.
Jonathan: Yeah. It's, they almost honest.
Denis: I don't even care about my own title. I just wanna solve the problem with data in the industry, whether I'm a data scientist or engineer, an IOT engineer, to me, suite not important. I don't want a job.
Jonathan: Yeah. Yeah. I don't think the disciplines is like super important here. You could just, you could say something like data consultant manufacturing, data consultant, or something like, I don't know if I would say software consultant, data professional or software.
Jonathan: Mm. I would say consultant or advisor. Consultant. One of the two. I think data
Denis: consultant is
Jonathan: pretty common. Yeah. It's fine for the discipline to be common. It doesn't matter either. You want to go, you really want to go left field with your. Discipline. Like so weird that people ask you questions about it.
Jonathan: Like remarkably weird. Like, uh, one of my students, Corey Quinn, he just named himself a cloud economist, which just forces people to be like, what is that? Yeah. Or you know, if you're like a dog lawyer, what is like, what is that? So it's almost like a joke. It's like a icebreaker. Mm-hmm. But if you're not gonna do that, then the discipline doesn't really matter that much because it's gonna come out in conversation.
Jonathan: But people's brains wanna label someone with a job title. So you wanna do one that's accurate, but kind of vague. So data consultants good because then you can, that can kind of roar shark test itself into whatever it is they think they need.
Denis: Yeah, I like it one too. Yeah. I actually first omitted it. I just said I help medium sized element manufacturer, but maybe it's better to have it some title after all.
Jonathan: Well, the L F P S includes the four main components of a positioning statement, but you don't, you don't necessarily use them all in every situation. There's different situations where you'd want to use them. So, you know, if someone, if you met someone at a party and they say, what do you do? You could say, I'm a data consultant.
Jonathan: You could say, I'm a data consultant for medium-sized aluminum manufacturers. You could say, I help aluminum manufacturers make more money, or I help aluminum manufacturers increase their o e E score. It's different context. You're gonna use it in different, different ways. But the, the nice thing about having an L F P S is that it covers all your bases for like any kind of question you might get.
Jonathan: So if you said, right, yeah. So like on your business card it might say data consultant, and then like the main part of the card says, I help, you know, process engineers improve their O E E scores.
Denis: That's funny story because I didn't realize it. I'm not supposed to read the whole thing. So when someone asked me like the other day, Hey, what do you do?
Denis: I was like, I'm gonna practice my L F P S, and I just read those two very long sentences and they were like, what?
Jonathan: Yeah, yeah, yeah, yeah. I call that the cocktail party answer. You need to have a cocktail party answer version that's like five or six words, or like very, you know, like a small number of syllables so it doesn't feel like you're giving somebody a lecture.
Jonathan: Yeah, that makes sense. Um, okay, so we'll just go target market for now. Although I would love it if we could switch to ideal buyer.
Denis: I, I think I may focus actually on aluminum smelters instead of manufacturers. Ooh. Because it's a very vast industry and aluminum I is just way too big. Uh, I just realized.
Jonathan: Love it. I like that a lot better. And, and are there huge aluminum smelters, or are they all basically medium sized businesses?
Denis: It's a bit of mix. I think there are some, well most of 'em are big companies, but honestly the plants, they, they companies became big because of takeovers. So even though it's one company, They still have very different cultures.
Jonathan: I love it. Alright, I'm a data consultant who helps aluminum smelters and then pick, uh, we've, we've picked a bunch of potential expensive problems. Eliminate downtime, decrease downtime, decrease production delays, improve quality control, improve your O E E score, improve their O e E score. You could really test any of those and see which ones resonate.
Jonathan: Yeah. And then your unique difference is, this can be very hard to determine if you don't know who your competitors are. So maybe you do know who your competitors are, or maybe you just know that your background is super unique.
Denis: It's actually the both 'cause. To give you an example, I had this email today from a very general data consultancy agency, and they just didn't, they emailed me personally to improve my process with ai.
Denis: It just didn't speak me because those too general, it's not about my business. So I think what makes me different, different is that I have this both education and the work experience in this vertical.
Jonathan: Yep. So
Denis: I think I try to focus on that. Yeah. But what I had up to now was, unlike general data professionals, I rely on my methodological education and aluminum industry experience then to design a simple but robust software solution that automates repetitive data operations.
Jonathan: Yeah. Yeah. It went on for too long. But that the beginning part was good. It was, it was solid. So because you, you, you flipped into pitch mode. Yeah. It's like, just tell me what's different. What's different. And the difference is that that unlike, unlike
Denis: general data professionals, I rely on my methodological education and aluminum industry experience.
Jonathan: Yeah. You could just say, I have. Yeah, I have a background in metallurgical engineering and I've worked inside the industry or however you put it, but yeah, just keep it simple and yeah, that's solid. I think this is, again, hypothetically speaking, I think the data consultant piece, maybe this is an improvement you could make there, but it probably doesn't matter that much.
Jonathan: Mm-hmm. The aluminum smelters I absolutely love might not be great, but I love it.
Denis: So do you mean it's finally the company, or should I look for a person at such a smelter, the context I, or to
Jonathan: focus ultimately? It would be great if you could find your ideal buyer. Mm-hmm. Yeah. But today I don't know that honestly.
Jonathan: Yeah. Today you don't know. So, so we're, we're a little bit, so in a sense it's, it's pretty specific aluminum smelters. I mean, I've never even heard those two words together before.
Denis: There are not that many, honestly, in the world, like 30, maybe 50. I think.
Jonathan: All right. Well this is a, that's a great starting market to start out with.
Jonathan: If you could laser it down to a particular person inside of the organization. I think process engineers is a smidge too low. It's probably who they report to or the VP of operations. Mm-hmm. So, you know, that could be VP ops, could be the target. But I think for now, aluminum FE is where I would start. And yeah, the expensive problem, you want to pick the most expensive one that's, that the most exp combination of expensive and urgent.
Jonathan: So downtime thing, unless it just happened to them, which is probably, it could be a bad time. Mm-hmm. It could be a, it could be a very, like a, not a great
Denis: fun time. So one problem we have specifically at Smelters is that they have. Often hundreds of machines in parallel in series, like 260 cells. Oh man. And also the, um, quite often one of those breaks down, I think it happens almost monthly.
Denis: Mm. And that's a very big problem. 'cause then the whole wine shuts down. People have to wake up and go there
Jonathan: and fix it. So, so now check this out. By focusing down on smelters, specifically, downtime became, instead of once or twice a year, maybe to once a month probably. Oh yeah, for sure. Yeah. Yeah. So that, so I, I really like that.
Jonathan: So I'm a data consultant who helps aluminum smelters decrease downtime. So your cocktail party answer could be like, oh, I help aluminum smelters decrease downtime, or, or keep the machines running actually really like it because that's
Denis: exactly, it's, it sounds very, very real.
Jonathan: Yeah. It's like at that perfect level where it's not too specific, but it's specific enough that it sounds, doesn't sound like bss.
Jonathan: It doesn't sound like, oh, but we, I solve hard problems for smart people. And it's like, what? You read the whole website and you still don't know what they do. Yeah, exactly. No, I like it a lot. Cool. Great. All right, so sounds like this has been helpful. Yeah,
Denis: absolutely. I mean, thanks a lot. I think it's much more specific.
Denis: Great. Yeah, I mean, the only question I may have is like my homework, like how would I go to validate this?
Jonathan: Right. So I, there's a couple of ways to do it. It depends on how public you want to be about it. One common way is to update your LinkedIn headline to just say, I help aluminum smelters decrease downtime.
Jonathan: Mm-hmm. And, and maybe you could say, maybe you could phrase the decreased downtime in a way that's a little bit more in their jargon. You know, like keep the concrete side of the business running. You know, something like that. Mm-hmm. Running smoothly. I, I don't know. It's, but, but essentially the message that you wanna get across is that you can help them decrease downtime and then I would reach out to people in my network.
Jonathan: It's interesting 'cause you've got so many connections already, but, so, okay. But basically I would update my headline like that and I would update my about section to make sure that it kind of aligns with that. And it's not too pitchy. It's like, it's more about the expensive problems that you solve.
Jonathan: Mm-hmm. And then I would start connecting with people in the space like crazy. So, you know, just connect, connect, connect, connect, connect, connect with people. And then when someone accepts your connection, they're gonna see your headline, you know, it's like your name, and then it's gonna say, decrease downtime.
Jonathan: And you could ask them a, you could have, uh, you take a second to like research their profile, see if they post anything and look for a question to ask them something that you're curious about. Maybe you've got three to six kind of stock questions that you'd like to kind of survey this group. And maybe you just say, maybe you've got a survey ready to go and you say, Hey, you know, thanks for connecting.
Jonathan: I've got a 32nd survey that I'm, I'm trying to survey. The aluminum smelting space and I'd, you know, would you have 30 seconds to share your expertise? Just check a couple boxes. And you know, if they say, oh, I'll take a look at it, send it over. And then maybe they give you the answers and then maybe you have some follow up questions about the answers.
Jonathan: Just DMing them on LinkedIn. Mm-hmm. Mm-hmm. And eventually get to a place where you say something like, you know, what's, what's your biggest headache? Where you guys are your personal biggest headache, where you guys are, and maybe they say something like human resources type of thing. Like, oh, management is, management stuff is the biggest problem.
Jonathan: And you're like, well, specifically with regard to operations, you know, like, like keeping the machines. No, well I don't, I don't wanna put words in their mouth. So it's like, is it, what's more important to you guys? Like avoiding downtime or improving quality control? Something like that. Mm-hmm. So you can just start to find out what is gonna attract people's attention when they see it.
Jonathan: Like, so if you're walking in the drugstore and you've got a migraine and you see a. A jar of pills that says migraine, you know, quick migraine relief, it's gonna jump off the shelf at you visually. Mm-hmm. So you wanna find the thing that's gonna ha the, the expensive problem, the migraine headache that's gonna have people jump off, you know, have you jump off the shelf at them.
Jonathan: So they start to reach out to you and say, Hey, I noticed that uh, you help people with, you know, increase, decreasing, whatever the expensive problem is. How do you do that? Or maybe they look at your activity stream at your feed and LinkedIn and they say, oh, this, this guy's some kind of a data consultant and helps people with this.
Jonathan: Their data pipeline and dashboard, whatever, executive dashboards based on sensor data. Who knows? Like whatever you wanna find out how they think of it. And then slot yourself into the places where it's appropriate to start the conversation. And once you've got a feel for that, then it would be like, I guess in, in parallel, you would also be sort of iterating on what the product I service is the roadmap thing.
Jonathan: Like what is that, what is the description of it? How would you do it? But also what is, what would a sales page look like and what's the appropriate language to put there for these people that you're having conversations with on LinkedIn? Would the results
Denis: of this survey be something interesting from my daily
Jonathan: user?
Jonathan: Sure. I would think so. Yeah. And you could even have people, like, it could be a lead magnet. So you could, you could post a survey, like a type form or Google forms or something and just dmm the link to people or say, you know, ask them first. It would be okay if I, I give you a link to a survey. I'm trying to survey the industry.
Jonathan: I'm sure some number of people's gonna say, sure. And then say, and in the form it's like, if you'd like to see the results of the survey, enter your email address here. Hmm mm-hmm. That's a good idea. Yeah. So that, I mean, that's just one way, that's like a, it feels pretty appropriate for where you're at to start to kickstart your mailing list and, and the survey is gonna give you tons of ideas for the mailing list almost.
Jonathan: Almost. For sure. For sure. Cool. Well this has been fun. Is there anything else you wanted to cover? No, I think
Denis: I didn't know how to get started. I mean, I've learned a lot. I had some very big misconceptions and I think they're all mid dispel. So thanks a Tom Johnson. I'm really, really grateful for this.
Denis: Cool. No, I appreciate
Jonathan: super helpful. I appreciate you taking the time, sharing your expertise with that space. It's almost virtually completely new to me. And yeah, and I imagine that there're gonna be people listening who really get a lot out of it. Yeah, I hope so too. So where could people go to find out more about you, maybe get in contact with you, maybe introduce you to some product, process, engineer bosses, something like that?
Denis: Sure. So I think there's two places. The first one is LinkedIn, which is my, uh, first name and last name then is with one N And then gon Sharaf, that's G O N D C H A R O V. And also my personal website, which would be OV EU from European Union.
Jonathan: Perfect. Thanks so much, Dennis. Thank you. Alright folks, that's it for this week.
Jonathan: I'm Jonathan Stark, and I hope you join me again next time for Ditching Hourly. Bye.