Mike Bifulco: This is tiny improvements.
I'm Mike by Fulco.
The first year of being a
technical startup co-founder.
Last year, I wrote a reflection on the
first three months of building Kraftwerk.
The article was largely a reflection
on the differences between building a
company from scratch in 2023, versus the
time I had done it last, which was 2016.
Plenty has changed since then.
And I think that article
is still worth a read.
Today, I want to reflect on the
first year of building craft work.
We've written many chapters
of our story in the last year.
And I think it's worth taking a
moment to reflect on what we've
accomplished and what we've learned.
You won't know some things until you try.
I've spent a good deal of my career.
Working in UX.
I've gotten to work with some incredibly
talented designers and UX researchers.
And I do believe that a good
design is a competitive advantage.
Even still, when building real features
for real people, you can't always predict
what will work and what won't designing.
Great software needs to be
both proactive and reactive.
You need to have a vision for what you
want to build, but you also need to
actively respond to feedback and data.
For example, the very first version
of Kraftwerk's estimate builder was
designed to be extremely self-serve.
We built a tool that would let customers
scope out the entire project they
had without needing to talk to us.
We thought this would be a great way to
let people get a feel for what we could do
for them, and to be extremely transparent
about what their project would cost.
In practice, we found that people
were often overwhelmed by the
number of choices they had to make.
Over and over.
We saw people on our site getting stuck
playing with the estimate builder.
They were weighing the cost of
one little change against another.
Many people that stuck in this
process and many never actually
submitted a request for painting.
On a hunch.
We ran a quick AB test to see if it
would be more productive to have a
conversation with one of our team members
first, and the results were clear.
Our customers have lots
of interesting questions.
Enterprise calculator just
wasn't answering all of them.
We've since reworked the estimate process
to be more of a guided conversation.
And we found that using it to create
a human connection with our customers
has been a great way to build trust and
to get the ball rolling on a project.
You can't do it alone.
While building my last company, I was
convinced that we could do everything
with an exceptionally small team.
We were just three people
from inception to acquisition.
I'm proud to say that we did the damn
thing, but it was really difficult.
You'll often hear founding
employees at startups.
Talk about how they're wearing many hats.
That's definitely true.
And it makes work a lot of fun.
It's also true that you
can't wear all the hats.
It's also also true that there
are exciting, beautiful, new
hats out there that you didn't
even know you need to wear.
From my first day as a working
professional, I've always
wanted to be surrounded by
people who are smarter than me.
Building craftwork has been no exception.
My co-founders are dedicated, talented
and damn good at what they do.
We've also hired some truly
amazing people to build our team.
A few things that I've learned
about building the right team.
Engineers tend to undervalue
the importance of understanding
the mechanics of the business.
Uh, our growth team at craftwork has been
instrumental in helping us understand how
to reach new customers and how to build
a profitable product that people want.
Building a brand is
like nurturing a garden.
It takes time patience and a lot of care.
has Kraftwerk's brand has really
grown into something special and
it's been a joy to watch it happen.
Give our Instagram a peek.
If you want to see what I mean.
At this stage in the startup journey,
hiring engineers is really challenging.
The perfect engineer for a startup is
someone who's comfortable with a lot
of ambiguity who can work independently
and who can communicate well.
In addition to being
technically competent.
It's a tall order and it can be a
challenge to find the right people.
When I have roles open on
my team hiring is the most
important thing I do day to day.
It's also a full-time job in itself.
It requires a fair deal of patients
to assemble the right team, as
well as a drive to actively and
creatively seek out the right people.
Just enough and nothing more.
I recently came across the
Japanese phrase, hold a Hodo,
which means just enough.
I've been trying to use it as a
reminder to keep things simple
and to overcomplicate things.
My team is tasked with building
the features that we need most
today with eyes on what we'll
need in six months and in a year.
And that can be pretty overwhelming.
It's easy to spiral out of control
with feature creep and over-engineering
finding ways to help each other do
just enough for now is a team remit.
In fact, just enough spills
over into a few areas.
For example, build versus buy.
Almost every feature our team needs
can be bought from a SAS vendor or
coach built by our product team.
Making this decision can be
nuanced and difficult and it's
something we revisit regularly.
Adopting tooling.
Engineers love to play with new tools.
There's always a new framework, language,
database, or library to poke around with.
We've been very deliberate about
what we adopt and we've been very
careful to avoid adopting things
just because they're new and shiny.
We have also adopted tools with good
intention, but found ourselves at
jettisoning them when they don't
work out for one reason or another.
Shipping new features, adding new
functionality for our customers
and teammates is always tempting,
but it's important to me.
But it's important to be
mindful of the complexity that
comes with each new feature.
Our engineering team is small and mighty.
And every new feature we build means
we have less time to support the
features that we've already built.
I used to say just
enough and nothing more.
And there's a funny irony in that phrase.
A year into this thing, it feels a
whole lot more accurate just to say.
Just enough.
