A CTO That writes more than code
As one of my new year’s resolutions, I’ve decided I want to start writing other things besides lines of code, so… here goes my first blog post.
So, here I am, Hadar Shpivak, 28, CTO & co-founder of Fixel (AI-Based solution that helps digital marketers create better retargeting campaigns)
I’ll open with a disclaimer: I’m not a serial entrepreneur after 2 exits and millions of dollars in my bank account and probably not the most seasoned techy out there. But over the past two years, I’ve learned a lot, mostly through my mistakes, but also from the guidance of good people along the way. The basis of what I’m writing about is how we started out — what would I have changed, and what would I repeat.
If just one person who reads this will gain a valuable takeaway from it, I’ve done my good deed of the day. And if anyone disagrees or has something to say about it — even better — let’s set a coffee and talk it out!
I’ve heard founders of amazing companies confess that they’re ”still figuring things as we go…”- I can easily relate. The things that seem trivial today, were mind-blowing only yesterday. Similarly, I believe that the troubles of today will seem so tomorrow.
How it all started
My brother and I started this round of “entrepreneurship” just as I finished my military service.
I was after 7 years of service in an IDF tech unit, working on command and control systems, and having established a Big Data team in the unit.
Etgar, my brother, came from the field of digital marketing, so naturally, that was the market we chose to focus on. (More on what it’s like working with your brother — on my the next post.)
When this journey began, we started off with the latest trending buzzwords. At the end of 2016, it was chatbots. We initially built a plugin, that was supposed to help small online retailers to interact with engaged users who haven’t made a purchase, by showing them messages on Facebook messenger. The idea was nice, but at the time, when this industry was still evolving, Facebook’s policy for these types of messages kept changing and ultimately was banned. We were forced to make a change in our product. Which led to:
Pivot #1: Same same, but different
With Facebook Messenger no longer an option, we had to find a different path. We tried off-site push notifications as a different output method to reach out, once the user has left the site. One of its limitations was that it required an action from the end user (opt-in), to receive these notifications, and that appeared to be a major obstacle.
Launching it publicly, we were able to get some initial traction, which continued to improve from week to week. But at some point — we realized it doesn’t take off, that we haven’t nailed that Product-Market Fit yet, and we had no actual proof that the market would benefit out of it.
So we went back to the drawing board. We sat down at my brother/co-founder’s living room that day and understood we have something good at hand, but it is still far from what we want — we need to change something. We started thinking about what were the positive feedbacks and what were the negatives ones that we got from those beta testers (can’t really call them even trial customers).
The two things that repeated were:
- How do you know if a user is engaged or not, if he/she didn’t take a strict action, like add-to-cart or something?
- Sounds cool, too bad it’s only a WordPress plugin at the moment
Pivot #2: Know when to try something new
Our pitch at that time was “yeah, it’s a Machine Learning algorithm” when at the time it was 2 queries with endless IF statements.
We decided to expand the idea of an engine, that answers a simple question — “how engaged is this user?”(1)
We set to build an algorithm that actually gives an indication of a user’s level of engagement. At some point — it started working. And we have seen a positive, significant correlation between the users we marked as engaged, to those who actually made a purchase. By drawing the line between the different users, we can tell the advertisers who are likely to convert, if even they didn’t take a strict action
We made quite a major pivot, from a limited Wordpress e-commerce plugin aimed at small sites, into a platform agnostic advertising solution, which helped us implement easily in various sites, regardless of their platform (2), even bigger ones. In the end, Fixel started to grow due to its simplicity, the more features we’ve decided to remove, the better it went. (We keep all the complex heavy lifting on our end, trying to keep low-touch solution)
I don’t think every startup necessarily must have a pivot, but breaking one of the premises, even though they looked so well placed, seems like part of the deal for almost anyone.
I heard a metric saying that MVP is 500 hours of code, no matter how long it takes. It’s a nice one, my initial metric will be, within approximately 3 months, have you managed to proof value, that is worth paying for.
In the previous pivots, we focused on creating a fancy website, CRM, pricing plans and payment methods, before we had something real in our hands. When Fixel started, we didn’t have a website, the pricing was “$99 sounds cool?”“Cool”. Heck, we didn’t have a name. It was a Word document that said “audience segmentation” we sent to my brother's colleagues in the industry. After we saw initial traction we started handling these issue.
My key takeaway on this one for SaaS products: Build something that works, and creates value. It might be ugly, the code might be inefficient, and you will do a lot of stuff manually at first. But try and generate value for someone. Let “how much will we charge”, “when to charge / when to refund”, and “Stripe or PayPal” be your next problems. (but really. immediately next — more on that in next posts)
