Can One Developer Change a Roadmap?

Here at Epistema we put our money where our mouth is and actually use our own SaaS product, MENT for all of our decision making processes. Post the launch of the beta product near the end of 2018, our VP R&D, Zohar, asked the whole team a question: “What should be our highest priority challenge post the product launch?” The discussion took up rather quickly with strong participation from all. Four alternative answers were suggested to the question, by our product manager (Asaf), by the VP R&D himself (Zohar), by the UX/UI designer (Israela) and by one of the developers (and our first employee), Yair.

The first three answers were naturally focused on product values and features that should be next on our roadmap. They had to do with ways to increase user engagement, with completing the discussion journey and with the way we should enable users to take a final decision based on the discussion. The unusual answer was the fourth one, by our developer Yair, who claimed our emphasis should be on testing the product.

The Discussion Phase

The discussion went on for 4 days with 25 comments of all kinds by most of our internal users. A debate ensued on whether “Testing” is a legitimate answer to the question. Some argued that testing is an ongoing challenge and cannot be at any point “the highest priority” for Epistema. Another claim was that not all business questions should be open for a discussion to the whole company, e.g. questions regarding clients should be open just to sales and marketing, and questions regarding strategy should be open just to the management level. However, given our vision and philosophy, we insist on radical transparency, and we allow all our employees to join any discussion they would like. Furthermore, we even allow our board members to take part in all of our most sensitive internal discussions on MENT, and be aware of all of our internal debates and disagreements (and there are many…).

Discussion on Ment

Who Would Have Known?

The most interesting result of this discussion was that Yair’s answer came up as the top answer, i.e. as the highest score for its level of confidence, based on our proprietary Bayesian scoring model. MENT also allows us to analyse the reasons behind the support to this answer, and we saw two interesting aspects to this answer coming up as the best one. Firstly, many developers supported the answer with further reasons and secondly the whole of the R&D group, apart from one, voted in favor of this answer.

These analytics and the ability to follow the reasoning behind the support convinced our CEO that the developers are trying to signal that the pressure pre-launch had made us more fragile on the performance and stability front. The developers were arguing almost unanimously that we should invest time in testing the system. Further analysis on our system proved there is no “group think” amongst the developers. Their unanimous view in this discussion was unique and unusual. Given that analysis, the management team decided to change the course of the product development and put more emphasis on testing for several weeks.

 

a single point of view by Yair but actually represented the majority of developers’ views

Follow the Breadcrumbs Back

Running the discussion on MENT had two main advantages: people could join the discussion at any time of their convenience, from their desktop, from their mobile, or via our Slack bot. The management meeting had the results of the discussion and the differing views at the outset, and it helped focusing the discussion and taking the most trusted decisions. The cumulative view presented in MENT, proved that the answer on the need for testing the product was not a single point of view by Yair but actually represented the majority of developers’ views. Furthermore, MENT allows us to follow the line of reasoning, unlike other platforms that register just the vote without giving any reasons.

Share