Very often product teams tend to approach their MVP thinking linearly! What you want to do is try and validate your core Value Proposition with your Target Audience, as quickly and cheaply as possible! This approach is illustrated in the video with a summary of the conversation I had with a Startup Founder recently.
If you would rather read than watch the video, the transcript of the video is below :-)
Without going into too much detail about what they were doing, at a high level this is what they were looking at. In the industry that they were working in, a multi-billion dollar industry, they were looking at developing an AI based image processing system, which would be able to detect fraud and reduce pilferage. This system would essentially consume data from transaction systems and supplement it with high resolution images to be able to detect this fraud. Their teams had worked on several projects involving image analytics and developing algorithms to detect fraud. Over that period, they had developed deep expertise in this space. They were now looking at productizing this idea. The conversation then moved on to how they planned to approach this. Their thought was to first develop a transaction system which would generate the data, and build the fraud detection layer on top of that. My first question to them was how long it would take to develop all of this. The founder went on to tell me that it would take about three to four weeks to develop the fraud detection system, and about six months to be able to develop the transaction processing system. One thing I have learned over the years is that when a tech person tells you that it takes about six months to develop something, they really have no clue. They're just giving you a number. Of course i didn't tell him that! 😃 I went on to ask two follow-up questions:
How confident he was about developing a transaction system in a period of six months, that was as robust and mature as something that their customers had been using for several years.
How confident he was about convincing the customers to shift from existing systems to a new system which had not been tried in the market.
The answer to both these questions were not a very confident YES! So I went on to ask a third follow-up question about the need to develop this transaction system at all. Why not look at developing an interface which could consume a flat file or a data feed from this transaction system? And, do your magic with the fraud detection on top of that on that data. When I asked this question, I could see a light bulb💡 go off in the founder's mind. The approach of not developing the transaction system, offered four major advantages:
Your time to market was reducing significantly. Instead of taking six months of effort, in an area where you did not have a lot of skills, you would be able to go to market within a period of three to four weeks. That too working in a domain where you have a lot of expertise - image processing and fraud detection.
This approach allows you to focus on delivering the core value proposition that you are betting on! By getting customers to try this out, you would get immediate feedback on, how big a problem fraud really was for these customers.
Reducing the friction in customer adoption. When you ask customers to replace systems that they've been using for several years with a new untested system it's a hard sell. Whereas, if you can tell them that you continue to use those systems but try this layer on top, that's a much easier sell. I.e. You are reducing that friction to trying.
By not developing the transaction system you are no longer going head-to-head with existing players in the market. Rather you are coming in as a value-added partner who would work with existing players in the market. Instead of taking them on as competitors, you are becoming partners which would up several go-to-market opportunities.
One of the biggest mistakes product teams make while trying to develop a system or planning a platform migration is to try and do everything in one go, in a linear fashion. But what you want to be doing is taking it chunk by chunk and working on the most critical pieces, one at a time. In the last couple of videos we had explored the beachhead market strategy. Trying to build this transaction system and then building the fraud detection on top of that, is like the Allied troops trying to land their troops in the heartland of Germany as part of World War II. When you're working on your Minimum Viable Product, what you really want to do is to flesh out the end-to-end solution think about the user experience, and then strip it out and focus on the most valuable pieces that are required to prove that your value proposition is indeed valuable to the customers.
Take as non linear an approach as possible to deliver an end to end solution for the customer!
If you're working on a Minimum Viable Product or planning a Go to Market ...
Comments