This post wasn’t planned, but I received many questions & comments following my last post. On Friday, I heard a pitch from a startup. I thought they would help me illustrate the themes I referred to in my previous post and another important theme I didn’t mention – focus! As you read, try to think about the thought process and taking this case study to a higher level.
In essence, the company is a Q&A platform similar to Yahoo Answers, Quora or JustAnswer.com. They have a sophisticated algorithm that mines and classifies questions, then forwards them to the relevant experts for answers. The company has been in beta for the last 9 months and has a talented team that includes AI and NLP experts, and a CEO who started a number of companies, some were more successful than others. The team is backed by a strong technical advisory board.
As they made their pitch, I noted 4 key assumptions:
- The need – Consumers are not satisfied with current Q&A solutions in terms of time and/or quality.
- The technology – They developed an algorithm that can accurately classify questions and match them to the right experts for a fast and accurate answer.
- GTM – Viral – The company claims that the service has a viral element and people would share with friends, leading to a >1 virality multiple.
- Monetization – Lead generation of products or services (e.g. books) sold by these experts.
Given these assumptions, I expected the team to validate at least 2 of the assumptions and present preliminary results in their presentation:
- Benchmarks with regard to the time and accuracy of current solutions. How big is the problem? A time for answer of 5 minutes with a 80% accuracy rate is very different than an answer time of 20 hours with 50% accuracy rate. As an investor, I want to make sure the team validates the problem and its magnitude.
- Stats measuring current accuracy. Post a question on the platform and across some alternative solutions and compare time, accuracy, or even basic classification.
- The company’s entire GTM depends on its viral affect and yet not a single stat on that.
- How many affiliation links were clicked per answer (then add conversion rate and test your revenue assumptions). Is conversion rate high enough to justify a lead-gen based model vs. existing business models ?
The reason companies go beta is to test assumptions, get early benchmarks or key performance indicators, and improve KPIs as they move forward.
Nine months of work and learning and not a single data point on their pitch (~25 slides). Instead the company (with 1K users) discussed the use of funds to expand globally and enable other monetization methods. Before replicating growth into other countries you want to make sure you have the right formula in a single country. Instead of applying other monetization methods, focus on one and make sure you get this right (or change to another one if this one fails).
In summary, let’s step back for a second to convey two themes:
- Think through your major assumptions and make sure you prove/dis-prove them early on. Proving some of the assumptions reduces risk.
- Make sure you focus on what’s important. As a small team you can only address a small number of issues: A single market segment, a single monetization method, one or two distribution channels etc. Make sure you get them right before you take a stab at additional sectors, distribution channels etc. Loosing focus would reduce your learning curve and would be a warning sign to investors.