An experiment is a test to demonstrate a fact. In a startup context, more specifically at the initiation stage, it is meant to simulate the product, measure user behavior and get real feedback. If the first thing you plan for as soon as you observe a need and propose a solution is product development, you’re missing a big opportunity to first test the idea through various techniques that provide for small, practical and often costless changes but can have a major impact in initial validation, target definition and product feature selection. Here are 5 proven experiments you can conduct in this order.
Did You Check XYZ?
People are usually more willing to follow a recommendation if it comes from an outsider. After identifying your potential buyers, do not introduce yourself as the founder but simply say, did you check company X? I heard they do a good job at Y. This experiment can be conducted face to face or online. A friend of mine used this tactic. He found that sign ups were a lot higher when him and his team members introduced themselves as outsiders. More importantly, it showed them that people care for the solution and have a need to solve the problem.
What you need: A landing page showing off the product is a plus but not essential. You may start by simply observing user behavior upon introduction of the service. You can do this in forums or by attending meetings and social events. It is a powerful and yet simple experiment.
We’re Almost There!
Referred users, especially those who really need what you are offering, will want to create an account and start using the product. But you don’t have a solution yet! Test their intent by creating a longer than usual sign-up process. This is meant to increase the friction for two reasons: user filtering and gauging real interest. Those who go through the whole sign up process are more likely to not only use the product but also pay for it. Take the example of Buffer. They used a landing page and a series of registration steps to finally say, “Hello! You caught us before we’re ready.” In a blog post, Joel Gascoigne, Buffer co-founder said, “I used the landing page as an initial validation of whether they would go through a long sign up process for the product I pitched on the first page.”
What you need: First, keep your value proposition in mind and write down the different actions you want users to take for validation purposes. Second, design a landing page and implement the defined steps. You can always use non-scalable tools to simulate the registration process and gather feedback, however for a more realistic scenario, I suggest the use of a landing page in this case.
Sound Very Real
Following early stage validation from the first two experiments, incorporate customer feedback to come up with a video describing the product in details. As if it existed. Bill Gates announced Windows before even starting it. Dropbox showcased the product well before its completion.
What you need: With a strong early validation, don’t hesitate to invest in a video showcasing the product. Dropbox acquired thousands of users with their product video. Here is where you can learn more about their launch story.
Fake The Price
Nothing is more powerful than people pulling money out of their pocket for your service. After a long registration process, include a payment plan for testing purposes. Enable users to input their credit card information and pay. Upon payment, users read a message such as, “Hi! Thanks for your payment but guess what? We haven’t charged you. Instead, we’re giving you [discount] but we’re not quite ready yet! You’ll be the first to know when we are. Share the news with your friends here [social medial and email].” This a powerful experiment for validation and in building strong relationships with those who truly believe and need what you offer. Upon checkout and for testing purposes, Wufoo, acquired by Survey Monkey, showed customers a price but charged them less. This enabled them to test users’ sensitivity to the price in real time unlike surveys or interviews. The same experiment can be conducted without a product, that is, at the idea validation stage of the company.
What you need: Add a payment plan and gateway to the landing page you created for the second experiment.
Instead of delaying the service, you can perform it manually or by using non-scalable resources. If you cannot perform the tasks without automation, select a few of early users and do your best to simulate the experience. This stage is just a transition. It is meant for learning and validation. Users who take the time to go through the registration process then make a payment would not mind working with an OK product. The team at Doordash used their own vehicle, phone, Find My Friend app, Google Docs and other tools to simulate the food delivery process.
What you need: Whether it is a simple or complex offering, define the steps and find the tools or man power (you and your team) to simulate the experience. Founders at Doordash note that the simulation process helped them understand how their app algorithm should look like.
Idea or hypotheses validation can take many forms. Traditionally, it was about building the product. Lately, we turned to building smaller versions of the product (MVP). Today, it is about taking smaller simpler and earlier (than MVP) steps for early stage validation. This is done through experiments.
- Identify your potential buyers then recommend your solution as if it solved your problem too.
- Use a landing page and a longer than usual registration process to filter those who love your solution from those who kind of like it. End the form by a, “we’re coming soon”, message.
- After incorporate customer feedback from the first two stages, create a video showcasing the product and its features as if it exists.
- Pretend as if the product is ready for sale by allowing buyers to input and submit their payment information.
- Use non-scalable resources to simulate the product or service.
Can you think of another experiment? tell us about it below.