This was original posted on LinkedIn.
It’s been just over 6 months since I began working on 42 problems (and the Delegate/42ai product) with Hugh Hopkins. So far, it has been quite an experience!! So i thought it about time to share a little insight into our process of launching 42 as a venture. Here’s what I will talk about
The North star for this first phase is product market fit. Period. eat, sleep, repeat PMF. There is no point in thinking about anything else. It’s also worth pointing out that, as an entrepreneur, you will spend most of your time sifting through the one million reasons why you don’t currently have PMF. Wincing at every competitive development while trying to resist the temptation to try and do everything all at once. That’s entirely normal, I hope.
So let’s get into the specifics of Delegate/42ai… 150+ customer interviews, VC discussions, user tests and 2 MVPs later… Here’s how we took nothing and started something…
1. The problem - we had to start somewhere and there are a multitude of ways and occurrences that can inspire someone to take the leap. For me there were two reasons - I was always incredibly impressed by Hugh’s technical mind and ability to build, launch and commercialise impressive products and he was testing a V1 MVP as early as 2022. The opportunity to work with him was too good to pass up. The second was my first hand experience of the challenges of internal B2B company decision making… (a problem area one customer interview later described as nebulous - i was grateful for the feedback and to also learn a new word!). The two aligned and although risky, i hoped for the best and took the plunge.
Bottom Line: You gotta start somewhere…
2. Our hypotheses - Put another way, these are the few strands of hope I cling to each night as I try to convince myself that I have not made a terrible mistake… and why a massive tech company won’t just wipe us out sometime tomorrow. A few of ours were: (1) most of today’s fragmented tech platforms make the problem worse (2) someone is going to take a bazooka to the woefully inadequate traditional saas pricing model, so it might as well be us (3) AI has changed a lot but it is not where the unique differentiation resides for so many of the problems it will be a part of solving.
Bottom line: You will need some winds of change in your sails, because you will not get there on your own… You need to genuinely believe tomorrow can be different and place your key bets accordingly.
3. Customer Research Interviews - The simple way to put this is we never stop interviewing people outside our organisation. Ever. We continue to speak to people on a daily basis. Perspectives and feedback build over time and each conversation allows us to hone, reposition and refine the product vision and flesh out the details and impacts of the problem we are solving for. What inspired us to start this venture may not be what we end up building exactly. Adapting and repositioning (based on the depth of understanding we gain into your potential customers’ problems) is the the most enjoyable part of the job…
Bottom line: Success can only come from having a deeper understanding of the problem than anybody else. You can only find that outside of your organisation…
4. Prototyping - In parallel to our mission to understand our customers, the people who work in those companies and the problems we can solve for them - we knew we needed to test. Test a lot. Throughout the conversations we have had, Hugh has continually iterated on the MVP, taking realtime feedback and testing it out. This is critical for so many reasons. You need to truly test the feedback you receive by building it into your MVP asap, should you chose to take that path. We also knew we needed to understand the technical challenges that arise from these decisions, from building on AI and the viability of our direction. In many cases, it has never been done before, so we have encountered issues and I think i can safely say it’s safe to park as many of them as possible to get to PMF. Your sole goal is get something into the hands of your potential customers as quickly as humanly possible so you can begin to harness the power of their imagination and get real world feedback. How would their work change if they had this? How else could they use it? What other problems could it solve. Putting 42ai through 25+ user tests has helped crystallise the opportunity from the noise and seeing the excitement in peoples’ faces is another reason for us to keep going…
Bottom line: Get something in the hands of your potential customers as soon as humanly possible… park anything in the way of that goal.
—
As we build up to the launch of the 42ai Beta program we will be sharing more about the lessons we have learnt building on AI, in a space changing so rapidly and the lessons we have learnt. If you want to know more or would like to user-test our AI analyst, then please send over a message or connection request!
.svg)
.svg)
