Leveraging AI for Creative Direction in Meta Ads → Depesh Mandalia
===
[00:00:00] What about scripting? The creative direction. One of the challenges I'm personally facing is I can no longer be a creative director. I just don't have the bandwidth. I'm solving other problems. So typically back in the days with clients, I would, get their problem, get their avatars, get the hooks.
I would put the hooks together. I'll work with a video editor and really just nail the creatives and even, interview the founder. How can AI really help with the creative process for videos you can just give UGC people or video editors or, a studio or, talent essentially all of those scripts or guidelines to really execute for your ads to scale up, especially in this day of rails and tick tocks and all Absolutely.
So one of the things that we've done is we've taken scripts of ads that we've run in the past and uploaded that to our AI. You can do this yourself. If you go into open AI, you can create your own [00:01:00] GPT. Now you can get let's say you've got five years worth of scripts that you've run from videos that perform really well.
Literally just take those, put them into one PDF or separate PDFs, upload those. Into your AI, what AI will then do is when you ask it a question, it will use that documentation as context. So if you say I've got this client, here's the avatar. Here's how we're targeting. Here's what we're doing.
And we want to create a video script. For them, it will go and look at all these past video scripts that you've created, that you've authorized, that these are good scripts and it will create a new one based on that. So that's what we're doing right now. So we can create YouTube scripts, Tik TOK scripts, video ad scripts for Facebook, and my AI can do that because we've uploaded all of our content.
So it knows what good looks like. So no longer is it going off into its vast expanse of billions of data. It's picked up off the internet. Okay. It's looking specifically at what's worked for us and then trying to interpret that to create something new. And when we upload frameworks, the [00:02:00] VSLs, video sales letters, frameworks for just all kinds of creatives.
So that's us telling the AI, the things that we want it to focus on and then it will fill in the blanks. And I guess you can also throw in competitors, scripts, so it has even more context as to what it's going up against. Beyond ChatGPT, what AI platforms excite you the most? Good question.
So right now, Descript. I absolutely love Descript. If you haven't used it, it's a video editor with AI capability. So for example, when we create some video nowadays before I'd have to take it to a video editor to get it edited and maybe take out the arms and the Rs and add some B roll, which is additional video and add some formatting and stuff like that, we can do it within 10 minutes and my virtual assistant is doing it now.
So she's able to use this AI tool.
Mhm.
So Facebook essentially is helping you at [00:03:00] scale, distribute a sales copy or sales letter, more or less. And you're essentially filtering out those people who would interact. So what'd you do with the data? What's, what are you looking for in the data Facebook provides you in regards to the segments of audience are really engaging?
What are you really looking at? One of the core metrics to look at is when you look at click through rate, you can look at CTR all click through rate all, which means it takes into account all the clicks. And then you can compare that to the actual button CTR. And the analogy is similar to how you look at an email open rate and you look at an email click through rate.
So it's the same thing. Like what you're looking at with the ideas, how many people actually interacted with the ad overall, which is good. Cause they open the email effectively. And how many actually clicked on the call to action? Cause that's what you care about. But that ratio. We'll give you an idea of even if your ad was good and it got them to take notice and they click the read more, how many actually didn't click the [00:04:00] call to action because there's something missing on the ad, maybe the targeting was wrong, maybe the copy was wrong.
That's a really interesting metric to look at whether your ad was effective. Essentially everyone wants to click. And if you don't get the click, there's maybe something missing the ad. So that gives you a good clue as well. The other thing to look at is when you get that click into your website, You can fire a view content pixel.
So VC for short, you can compare that to your outbound click. That ratio tells you how many people actually landed on the page and fired that pixel. There's a few reasons why that pixel won't fire. Number one is your page load is too slow. Number two, they landed. And they didn't like it and then they bounced back.
So there's different clues to give you an idea of where people are going. The third one I'd look at for e commerce is the click through rate into the cart. So from those that clicked from the ad to your landing page, how many actually added to cart? I don't necessarily when I'm testing care about how many people bought, because the problem is [00:05:00] an ad can only get them to your site.
Then it becomes a site conversion problem. And I think that's where people get a bit lost, which is my ads not selling my ads the problem. Sometimes your ads actually getting people to your site, your landing page is the problem. Yeah. And that's a whole nother conversation with regard to landing page optimization.
Absolutely. Okay. So looking at just the placement. I get it with Facebook. I get the sales copy. How should like e commerce brands listening to this or brand operators listening to this approach Instagram, as a placement, as a channel with advertising or with prospecting that first bit, do they pay more attention into the video or the image?
How would you approach it?
In this episode, we explore how AI can revolutionize creative direction in Meta Ads with expert insights from Depesh Mandalia. Discover how AI tools can help streamline the video script creation process, optimize ad effectiveness, and analyze crucial data metrics to improve campaign [00:06:00] performance. Don't miss out on these valuable strategies – listen to the full episode at https://2xecommerce.com/podcast/ep424/
