The lean startup on steroids: Poly-startup
You hear the same story over and over: people try to validate a product or service with the lean canvas and it fails. They try several other ideas and finally after a long period of time, they find something that works. Ain’t nobody got time for that.
How can we find product-market fit faster?
The current lean startup movement often focuses on validating one or a limited set of ideas at a time and this slows down the process. We could try to run more experiments in parallel (eg Poly startup) and thus find product-market fit in a shorter time frame.
Let’s take a look at the different steps:
- Coming up with ideas
How can you come up with hundreds of ideas? You don’t. Start with an audience you want to target and scrape relevant websites. For example, you could scrape the blogs of influencers or company documents. You filter the relevant content, based on several criteria (for example: sentiment analysis). On several websites users are able to upvote relevant comments and this is also a great filter. Let’s say you have 1000 ideas at this point.
You can read a lot of scientific articles about mining ideas from text. If you don’t want to release your inner data analyst, you can use this tool to get started.
- Selecting ideas
These ideas are collected in a database. The same idea could result in several MVPs and this needs to be elaborated (for example: ‘need cheaper transport’ could result in a public transport idea, an idea with taxis and so on). Next, you filter ideas based on how populated the market is (avoiding overly-competitive markets) and market size. This could be partly automated but is the most labor-intensive part of the process. Then, you do a quick technical feasibility check to see what ideas can evolve into MVPs. Let’s say you have 250 ideas after this step.
You could run automated surveys (targeting people with services such as Pollfish and Google surveys) or talk to people (or hire a call center) to validate your ideas even further. Let’s say you have 50 ideas at this point.
- Building Minimum Viable Products
The next step is to translate your ideas in requirements.
You can then hire developers to build your MVPs. This would cost you a few thousand dollars if your MVPs are really basic, or use tools like Zapier and Typeform to create MVPs. These tools were not available when the lean startup methodology emerged.
This is more complex for hardware but 3D printers and VR/AR projections could help validate your ideas.
- Getting feedback and select the ones that work
You buy ads on relevant (niche) sites and use other tactics to get attention for your MVPs and drive traffic to your site.
You know how many subscriptions you have sold, if people have invited friends and so on. Now you have metrics about these ideas, you could come up with a formula to rank your MVPs and pick the winners.
Finally you can select the most relevant ideas and build them into products. After a few weeks you are able to select the most promising concept. The time to find product-market fit has shrunk dramatically. On the other hand, this approach is probably more expensive and will likely result in incremental ideas, but it is a good start.
Follow me on Twitter
Learn more about creating a product.
