How to evaluate your startup idea.
We have worked with a number of founders with startup ideas. All needed help to turn them into a valuable business. We have refined the first few months into a consistent process (especially for eCommerce and marketplace businesses). It’s proven highly effective in assessing whether the idea will succeed.
1. Stay objective.
All founders fall foul of confirmation bias to some extent, but in this early stage you must be flexible when deciding if your idea is worthwhile. The worst situation is to end up investing a lot of time building a product for a weak idea. Once you move into a phase of execution it becomes harder to change direction. So, it’s far better to be open-minded now and amend the idea later if needed.
2. Use Lean Canvas to identify your assumptions.
Before we invest in any idea-stage companies, we ask founders to create a Lean Canvas (a business model canvas designed for startups). Below is a lean startup business model canvas when we were evaluating Lexoo as an online legal marketplace.
3. Identify assumptions.
Now you’ve filled out your lean startup canvas, find some friends (with business/startup experience) and draw the following chart:
Spend 15 minutes on each section of the canvas and ask yourself what assumptions you’ve made about the business. Write down all the assumptions and how you would test them on Post-it notes. Categorise them by how easy they are to test and how critical they are to the business succeeding. Below is an example of a few of Lexoo’s Post-it notes on the chart:
This exercise gives you an ordered list of assumptions and ways you can test them based on importance and ease. Assumptions on whether users want the service or product are the most important and can be tested by speaking to users to understand their needs. The hardest assumptions to test are usually to do with how the business will perform at scale (where market sizing and looking at parallels can help).
4. Test assumptions around the problem, customers, and existing solutions.
All startups are trying to solve a particular problem/aspiration. Some problems are clearly felt but for others, your potential users may never articulate them clearly. That’s because they didn’t know it was possible to improve on how they are currently doing things.
"Don’t ask your potential users what they want. It’s your job as a founder to figure that out based on your deep understanding of their unmet needs and your knowledge of what is possible".
However, through interviews and observations you should see if they’re struggling to solve their problem – or solve it in a way you can improve upon. We recommend interviewing 20-40 users of different types to carry out early validation around the problem, your customers and existing solutions.
It’s also a great idea to observe people as they try to solve these problems. You should also try to solve it yourself using existing solutions. The interview process gives you a sense of whether you are onto a good idea and validates your key assumptions. It also gives you a wealth of nuanced information that will inform your product solution.
"Sometimes I work with founders that are reluctant to invest any time in learning about their potential users before building their product. They usually have a hard time."
If you don’t see a strong need through this process, you should consider whether it is a good idea.
5. Test your unique value proposition and solution.
The next set of assumptions to test are whether your target market will respond to your unique value proposition and solution. How you test this depends on the product and business type. For consumer companies, we usually create a homepage with a call to action that requires commitment from the user.
For more complicated products (Saas or mobile), we have built prototypes and observed users interacting with them to understand if we are providing value. For hardware products, many companies run Kickstarter or Indiegogo campaigns to prove need (and get funding for the first production run). Dropbox famously used a video to explain what it did to prove its solution was what people wanted.
If you are building a homepage test, it’s important it clearly describes what you are offering – and it should be beautifully presented. Users' expectations of product design are high, so if your product doesn't look professional, they may not engage and you may come to the wrong conclusion.
With tools like Strikingly, or a few days from a designer and developer, you should have something that looks and feels professional. Below is an example of the homepage we used with Lexoo. It also featured a form for users to clarify the type of help they require.
6. Test marketing channels.
Great startups are created through great products and a relentless focus on growth. For consumer companies, paid marketing is likely to be part of the mix. Drive users to your service to get a baseline of cost and volumes you can expect from channels such as search, Facebook or Twitter.
We usually see CAC to be much higher in the beginning, but it should reduce rapidly over the 12 months through campaign and product optimisation. If it is extremely high, that’s a warning signal, as your unit economics might not work out.
Volume is also important as it helps you understand how quickly you need to invest in offline marketing channels and PR in the early stages. By driving users to your homepage test, you’ll see whether users engage and start to test some assumptions about your marketing strategies.
7a. Test value and revenue in an unscalable way through a concierge service.
Using a concierge-style service on top of a thin product is a great way to test the real value you want to provide. You can use manual effort to fill out the steps you hope to automate. Not only will you have happy customers, you’ll also get a deeper understanding of what you need to build to keep them happy.
Marketplaces and ecommerce lend themselves well to this technique, as do many startups that use intelligent systems to offer a personalised service. Lexoo did this by manually matching lawyers to jobs and sent lawyer profiles to customers via email. Over time it built more product to automate some of the process to get it to scale while keeping the same level of personal service.
"Use the concierge service not only to test you are offering value but also to have more face to face conversations with your users and deepen your understanding of their needs."
7b. Test value using prototypes.
For SaaS and more complicated products, you’ll need to build more product before you know if you’re providing value. In these cases we prefer testing that value through prototypes. It’s not as powerful as the concierge service but you can still learn a lot before committing to building something. Running design sprints is a great way to rapidly test a product or feature. It’s also a great way for the team and customers to provide input.
8. Keep learning more about your customers.
This last point is the underlying principle you should be doing while evaluating your idea. In the time you spend on this, you’ll continue to understand more about your users. Be sure to email every customer, give them phone numbers to call you, add live chat on the site and organise user tests (online and in person). It’s only through a deep understanding of your customers that you can move on to the next phase of building a product that resonates.