You know that a new idea has gained dominance when it becomes practically a cliché. That is what I’ve seen happen with the “Lean Startup” philosophy of viewing your startup as a scientific experiment in search of a business model. This started as a new idea, then became popularized, and now people view it as common sense and practically a cliché. And I definitely am one of those people who regularly advises our portfolio companies at ffVC to think creatively about how to experiment on their operating assumptions.
The challenge is: what experiments can you run to test your hypotheses? I’ve looked all over, but to my surprise, I have not been able to find a good collection of appropriate experiments online. So, I’ve gathered below some of the experiments that our portfolio companies and other startups we know have found most informative. If you have further ideas, please add them in the comments. To structure your experiments, I suggest use the Javelin Experiment Board.
Step I: Explore the problem and the market. Seek first to understand, then to be understood.
- Blog publicly about what you’re doing. This helps you to get qualitative, personal feedback first, which you need to inform quantitative feedback you gather later. Also, include face-to-face interviews with customers; read these quick tips for effective customer interviews. I’ve been writing for years about investing best practices as a way of learning about the investors who are the clients base for several of my past companies.
- Ask open questions on Quora and other online discussion tools about, “How do people solve this problem….” For example, “What CRM tools are used by venture capital and private equity funds?”. This will surface both competitors and customers. Make sure to look on AngelList and Crunchbase for companies in your space.
- Run surveys, playing with monetary and non-monetary incentives. Try evaluating response to an incentive like “$100 off our product when launched”; if that discount off a non-existent product creates a response, it’s further validation of customer demand.
- Collect pre-orders. Crowdfunding tools like Indiegogo* have made this dramatically easier than in the past, e.g., see this Indiegogo campaign for Interaxon Muse*. Even if your product is not a fit for crowdfunding, ask potential clients for a letter stating that they will pay you $X if you can build a product which does Y.
- Run test ads, using Google Adwords, Yahoo!, Bing, etc., which take viewers to a page soliciting email signups and possibly pre-orders. Test which ads are most effective, e.g., see this example of a grammar change that boosted ROI by 20%. See how Tim Ferriss titled his book based on Adwords conversion rates. Don’t just collect emails, but collect data in the form of a mini-survey. I suggest check out QuickMVP, an all-in-one tool for building launch pages, driving traffic through Google Adwords, and analyzing customer demand.
Step II: Explore the solution.
- Test multiple iterations of your site. Here are some Smashing Magazine examples. Launchrock is a great site for building launch pages and analyzing user data . Try experimenting with different A/B testing campaigns using Optimizely.
- Talk with real users of your beta product. Here are some ideas on getting beta users free. One of the reasons why crowdfunding is so powerful is that it provides not just capital but initial users; Interaxon learned a lot from the hundreds of people who contributed to their crowdfunding campaign.
Step III: Market.
- Analyze site usage. Testing what words get the most hits can give you insight into the target market. See Beginner’s Guide To Web Data Analysis: Ten Steps To Love & Success. Katya Constantine, CEO, Digishopgirl Media, suggests that you make sure to fully leverage the advanced features of Google Analytics such as goal tracking, demographic and interest segmentation, and cohort analytics.”
- Analyze which marketing campaigns get the most traction with Copromote*, Bottlenose*, and other social media analysis tools. Just as you need to understand your end users, it’s also important to understand the behavior of the influencers who touch your end users.
- Test referral programs with monetary and non-monetary incentives. A famous example: Dropbox’s referral program had a two-sided incentive for sharing: the person who signs up for Dropbox through a referral link gets more space than through a normal sign up, and the referrer gets additional space as well. See some of the best referral program examples for startup user acquisition and growth.
I have found some helpful resources on this topic below; please share any others you’re aware of in the comments:
– 4 Steps to Test Any Business Idea in 5 Minutes
– Lean Experiment Techniques
– Tactics For Testing Your Minimum Viable Product
– Five Pitfalls Of Running Lean Startup Experiments
– Why Lean Startup Experiments are Hard to Design
* ffVC portfolio company
A shorter version of this posted in Entrepreneur.com