Business model validation


It’s obvious that you shouldn’t waste your time and money to build products that nobody wants to buy or use. But what do you do when your business model is based only on assumptions and educated guesses instead of solid facts? How do you minimize the risk of failure when you intend to sell something new which has never been sold before? How do you know that customers will buy it? These are ordinary issues for any innovative startup business model.

Experiments and primary data from the market are the best choice in this situation. Startups formulate a hypothesis related to their business model and validate or reject them in actual conditions. Literally you take an assumption and test it. If it’s validated, you test the next one until your business model is based on solid facts instead of just assumptions. Usually, it’s more expensive than checking already available data. Therefore, it’s highly recommended that you do your homework with an initial market research and only then proceed with hypothesis and experiments.


1. List all the assumptions

Finding the problem-solution-market fit is the most popular approach and it’s hard to argue with that. Basically, it means that you must find and confirm a problem which is worth solving. Once you have a confirmation from potential customers that this particular problem is a real headache for them, you must create an acceptable solution. Once you have a general approval from customers that the solution is suitable for them, then you ask them to pay for it. If you get paid the price acceptable for you, congratulations! You have found the problem-solution-market fit. But there might be much more important hypotheses which could help in developing your startup business model. Consider the following:

  1. Value proposition hypotheses
  • problems, jobs and wants hypothesis (is a problem or desired benefit strong enough?)
  • problem-solution fit hypothesis (does the solution fit the customer expectations?)
  • value proposition-market fit hypothesis (is the customer ready to pay a certain price?)
  • segment hypothesis (do we know who our early evangelists are?)
  1. Business model hypotheses
    • distribution and communication channels
  • costs and revenue
  • unfair competitive advantage
  • others
  1. Business growth hypotheses
  • business evolvement potential (how large could the new market be?)
  • the probability of initiating a viral effect (what are the chances of initiating a viral effect so that current customers bring new customers?)

2. Complete experiment cards

Not all startups use experiment cards as recommended by Eric Ries (2014). But those who do so agree that this simple tool made startup development much more structured, saved money, reduced risk, and helped to show additional possibilities.

It’s a two-part table which fits on one page. The experiment task explains which hypothesis you want to check, how exactly you plan to do it, and the next step if the hypothesis is confirmed. The same sheet contains the experiment result card which describes your findings based on facts from the experiment and obligates you to make a decision about what you will do next.

It’s recommended to have an experiment folder (whether it’s a folder on your computer or a real physical folder) where you can save printed experiment cards. Because there are many different types of hypotheses, testing a hypothesis only once won’t be sufficient unless you get extremely strong results to support your initial hypothesis. Usually, the same type of hypothesis must be checked a few times to evaluate different alternatives or pivot scenarios. For example, you may want to test if your proposed solution is accepted by target customers by checking which one of several different solutions fits the customers’ expectations best, which customer segment appreciates a particular solution the most, and so on. Therefore, you’ll have plenty of experiment cards! The more experiments you run, the more you learn about your market and business and the more chances your startup has to succeed. If you store all those experiment cards (tasks and results filed correctly) in one place, it will be much easier for you and your colleagues to follow up. You’ll have an archive of your own fact database. This is your knowledge base, your asset! When you are pitching your idea to a potential investor and get questions about why you are not doing this or that, it’s quite possible that you’ll already have an answer in your experiment folder. The more facts you have, the more professional and valuable your startup will look in the eyes of investors and possibly new partners. Here is what you should do with the experiment card (template No.1):

  1. Declare expected outcomes of the experiment. You should clearly know the goal of your experiment, how you’ll measure results, and what criteria are being verified for each hypothesis.
  2. Clearly defined time or conditions for how long you will be running the experiment. Decide how long you will run the experiment (for example, until you conduct 50 customer interviews, drive 500 targeted visitors to your landing page, or get the first 100 people to sign up). It’s very important to set a maximum time limit for how long you will run the experiment. It might be that you’ll never get 100 people to sign up if your value proposition or call to action is terrible. If you don’t get your expected result in a set time frame, change something and run another experiment!
  3. Seek realistic results instead of chasing precision. You need to accept the fact that you will probably never have all needed information you really need with perfect precision. But you’ll need to make a decision in any case. So it’s better to seek to measure results that would reflect a realistic situation.
  4. Measure actions, not words or opinions. Always remember that saying is not the same as paying. This rule applies to any other action you want a potential customer to make. So don’t just ask “if you’d like to sign up to get product updates.” Instead, create a signup form, drive traffic to it, and check to see if people are really signing up.
  5. Always have a control group. This is probably one of the most important rules to run effective experiments. If you don’t have a control group (unaffected by your experiment), you’ll never be sure if your experiment results are valuable. You can only draw justified conclusions when you have a control group and their results are different from the experiment group.

3. Create the MVP that fits best for your experiment

Don’t overload your MVP (Minimal Viable Product) with unnecessary features and details. That’s the main rule to follow. You are seeking validated learning. Therefore, run planned experiments with MVPs as close to the real product as possible, yet make them easy and inexpensive to build. Selecting the right type of MVP depends upon the reason you are creating the MVP. Generally there are two categories of MVPs:

Low fidelity prototypes usually are used to:

  • Better understand the problem and related issues
  • Check how important the problem is for customers
  • Get confirmation if the problem is worth solving
  • Understand what kind of solution would be most welcomed

High fidelity prototypes help to:

  • Determine how much customers are willing to pay for the solution
  • Find early adopters and evangelists (passionate and enthusiastic customers who will spread the word about the product)
  • Optimize various aspects of marketing strategy (value proposition, call to action, communication channels, etc.)
  • Identify most potential viral growth techniques


While choosing the type of MVP for your particular hypothesis verification, you should consider:

  • What is the biggest risk you have right now and how you could check it?
  • How much time do you have to build this MVP and get results from it?
  • How much money do you have at this stage? What amount would be smart to use? Don’t plan anything fancy for your first tests!
  • What makes the most sense in your case? Which of the hypothesis validation strategies would bring your startup to the next level?


Types of MVPs

Low fidelity MVPHigh fidelity MVP
·      Customer interview

·      User story

·      Paper prototypes

·      Digital prototypes

·      Blogs and forums

·      Explainer video

·      Landing page and A/B testing

·      Audience building

·      Micro survey

·      Ad campaigns

·      Fake doors

·      Wireframe

·      Single featured

·      Physical product prototype

·      Concierge

·      Wizard of Oz

·      Piecemeal

·      Crowdfunding


  • Customer interviews are designed to collect information about the problem customers have and the solution to the problem you intend to provide. It is recommended that you run unscripted interviews which open the possibility of capturing uninfluenced and maybe even unexpected customer feedback, insights, and suggestions. The interviews should try to explore customer needs rather than sell your product. This can be done by listing the problems you intend to solve and then asking what customers think about them and how they would rank each problem (from the most painful down to not essential). Keep in mind that if you started developing your startup business idea based on assumptions and secondary data, customer interviews should be one of your first choices for problem verification.
  • A user story is something more than just describing what your product is and does. It is a short story about a user, what problem he has, how he uses (or will use) your product, and how he captures (or will capture) value from it. A user story usually describes a process or steps taken by a user and captures the value of the product. Products already in the market usually have many user stories, but to create an MVP you can start from a single story and modify it according to your findings. If it is your MVP, you should also employ a testing mechanism (for example, invite users to signup after the user story is told, do a customer interview of the user, or even offer a pre-order). Startups mostly use user stories to identify which features of the product should be built.
  • Blogs and forums are a great way for checking your ideas within the target market using minimal efforts. To set up a blog on WordPress or Joomla costs little except for the cost of hosting services. Having your own blog gives you a two-way communication with your potential customers, which might be extremely valuable in the process of creating your high fidelity MVP. Of course, you’ll have to figure out how to drive traffic to your blog, but at the same, it’s possible to check other hypotheses: what channels are more effective, what aspects of your message drives more attention, is there an interest in your topic at all? If you are in very early stage and wondering about the importance of the problems you intend to solve, you can search for different forums and browse them. You even don’t need to create your own blog or forum. Just visit a dozen or so forums and check if there is a topic relevant to the problems you want the address. If yes, check to see what people are talking about. Maybe there are many alternative solutions and your chosen problem is not a problem anymore. In case if you can’t find relevant topics, initiate your own in as many forums as possible and wait for the community to respond.
  • An explainer video is a short video that explains what your product does and why people should buy it. Often it‘s a simple, 45 – 90 seconds animation. Sometimes it’s extended with developer and customer interviews. You can hire a freelancer (Freelancer, UpWork, Fiverr) to create quite a decent video with voice over at price of $200 or even lower. Sometimes it’s cheaper to create a video than a website! I don’t think that a professional high-end video representing your product can be made for such a low price, but remember, you are just testing now! Once you are ready to launch your product version with customer confirmed features, then you can create fancier and more expensive video, if you think it’s necessary.
  • A landing page is the main web page where visitors are directed after clicking a link from an ad, e-mail, blog post, or any other another communication channel. The main task of a landing page is to instantly communicate the value of your offering, overcome possible objections, and call visitors to action (usually to submit their contact details, place a pre-order, or even purchase a product). The landing page helps validate your value proposition (product-solution-market fit), sales arguments, pricing of the product, and even to choose most effective communication channels (if you drive traffic to a landing page from different channels, comparing the conversion rate through Google Analytics will let you know which channel brought you the most early adopters). Don’t forget that there are other free tools as well. Weelytics is a site that makes it easy to track the moves that customers take on your site without the need for coding skills.
  • A/B tests are used to measure the effect of any changes to your product or marketing. There are many analytical tools which could be used to test how visitors react to various changes on your website, whether it’s design, introductory video, value proposition, call to action, pricing, guarantees, testimonials, or other elements. Generally speaking, A/B testing allows you to test two versions of a web page or marketing copy and let visitor interactions determine which one performs best. You are not asking your customer which one they like more, you just measure and make decisions based on metrics, not opinions.
  • Digital prototypes such as mock-ups, wireframes, and actionable prototypes can be used to demonstrate the product’s functionality in a way close to the actual experience and perception. These prototypes can be low-fidelity sketches as well as more complicated interactive applications that a beta user could test and feel the experience as close to using a real product (app, platform, website, etc.).
  • Paper prototypes are quite similar to digital prototypes, except these are physical. You can make a simple sketch on paper or make cut-outs to show your product and help to visualize the user experience. The main advantages are that paper prototypes can be used by almost anyone, requires little time and efforts, can be easily and quickly modified, and need very little explaining or not at all.
  • A physical 3D model is a more impactful prototype but is also more expensive than a paper one. Usually, 3D models can be printed or molded from several types of plastic and metals. If your startup is building a technical product that has to be manufactured, you might consider creating a 3D model first. There are companies that specialize in the production of physical 3D models or even fully functional physical prototypes.
  • Ad campaigns are a great way to run market validation surveys. Google and Facebook advertising platforms allow you to choose demographics to target your potential customers. Other advertising channels almost always allow targeting based on the interest of potential customers, be it a section in a news portal, niche blog, magazine, newspaper or other media channel. This lets you run a low-fidelity test to see which features or aspects of your product are most appealing to potential customers (if you use different aspects in your message, but on the same channel). In addition, running a survey and driving traffic from different advertising channels gives you the chance to evaluate customer acquisition through those channels and get more in-depth feedback from potential customers through their answers in the survey. As competition in Google, Facebook and some other major advertising platforms are getting fiercer, it’s important to understand that these campaigns won’t give you great exposure, but will allow testing your hypotheses.
  • “Wizard of Oz” is an MVP when you put up a front that looks like a real working product, but you manually carry out product functions. This is very efficient if you want to check whether you’ve got a desirable product or service before you actually build it. Plus you can also identify the pain points of the product or service, possible problems of delivering them, and unexpected customer actions, among other things.
  • Concierge MVP means that you start with a manual service instead of providing a product. The main thing here is to make the service consist of exactly the same steps people would go through with your product. Usually, the product or service is delivered as a highly customized service to selected customers. It requires quite much time and effort to handle everything manually, but, actually running through the process manually reveals various aspects of the customer experience and provides highly valuable insights in many cases. So instead of using all your resources to build a real product, it might be wise to allocate some time and energy to run this kind of MVP testing to figure out if you are building the product that customers will love and be willing to pay for.
  • Piecemeal MVP is something between the “Wizard of Oz” and the concierge MVP. You make people go through the process of using your product, but instead of delivering them manually, you create them using existing tools. Why should you invest time and money into building your own infrastructure, when the product can be built or shown using other existing platforms and services? Putting your product image and description on eBay would save you from creating your own website and driving traffic to it. Check to see how much attention (views, clicks, orders) your product gets on such e-commerce platforms and you’ll have a good indicator whether it’s interesting for anybody!
  • Single featured MVP tests a single, essential feature of the product. The theory is that, if you can’t find a single, super important feature that can stand on its own (at least for early adopters), adding more features won’t make your product much more desirable or essential. Using this type of MVP also prevents users from being distracted by other, maybe non-essential features, and makes them pay attention to the essence of the problem or solution. This restriction allows you to test major aspects one by one, gaining a clear understanding of which one of them is most important (which would be much harder to determine if you were running a test with a multi-feature MVP).
  • Crowdfunding MVP could be described as selling it before you build it. The idea is very simple: launch a crowdfunding campaign on platforms such as Kickstarter, IndieGoGo, RocketHub or any other, and raise money to create a product. For the best effect, you’ll need to create a prototype or demo video, and then see the reaction you get. Unlike a normal crowdfunding campaign, you shouldn’t do everything in order to drive the interest of backers, because this is just a test to see whether a market for your idea exists naturally. This type of MVP will, first, help you validate if customers want to buy your product or not. Secondly, if you are on the right path, you will also raise money. Thirdly, if your crowdfunding campaign is successful, you will create a passionate group of early adopters and maybe even raving fans who will spread the word about your product.
  • “Fake door” is an easy way to measure whether your existing customers will be interested in a new product or feature. The main idea of “fake door” is to show the access to a particular feature, benefit, or product when you don’t have it yet and to measure how many of your customers would try to access it. For instance, you could put a call to action button on your website, but when a visitor clicks it, they get the message “coming soon.” Actually, you even don’t need to have a website to employ “fake door” MVP. It also works well if you’ve got a large social media audience: share the news about the product or feature on your networks using customized links and track the click-through rate. You’ll easily find out how many of your followers would be interested in trying, buying, or signing up. If the conversion rate is good enough, you have a clear verification to build the real product or feature. But, a word of caution about using this type of MVP: don’t get your existing customers too confused or frustrated. You should always explain and be honest with them.
  • Audience building is a simple idea. Before creating a product, build a virtual community (tied to the problem you intend to solve or even to your end product). If you can create such community using low-cost or even free methods (Facebook or Linkedin groups, newsletter, blog, etc.) it would be an indicator that there is a group of people interested in the topic and, you would already have a list of potential customers, once you create a higher fidelity MVP or even the product itself. Creating such group is not just a one-time test because you’ll be able to run surveys, in-depth interviews, and experiments within the community to make your product more valuable to customers.
  • Micro-survey is just one or two questions long, thus it has better response rate, is more reliable, and can be precisely targeted to the specific customer situation. For example, if you have a website, you can use a pop-up with a single question when the visitors performs an action (signs up for a free trial, downloads free material, or visits the pricing page). To answer a single question in the pop-up window doesn’t require much effort from the visitor, especially if the answer is yes or no. Once the answer is submitted, you can ask one more question and it may be an open-ended. Always consider making your micro-survey questions specific but open-ended in case if you want to get more insights and more feedback. In addition to pop-ups, you can also deliver your micro-surveys via email.

4. Find and engage potential customers

Once you’ve filled your experiment cards, it’s time to start the real job: engage your potential customers and run the experiments. You will need to find a significant number of potential customers and not fall into the trap of just showing your assumptions, solutions, and prototypes to your friends, relatives, and classmates. Generally, it’s far better to interview only 10 potential customers rather than 100 friends who might never need your product.

There are different ways to find potential clients depending on what kind of engagement you are looking for and what kind of experiment you are running. Consider such channels as:

  • Your personal network is very good for low fidelity prototype testing, but as we have already discussed, be careful about that.
  • You can use Google Adwords to find out if people are searching for how to solve a particular problem and then you can show them your solution and see how it is accepted.
  • Social network groups are very convenient for segmentation because each of them unites people according to particular criteria and gives you scalability.
  • Partnerships allow you to save costs and get in front of target customers. Take a minute to think about who is already serving your potential customers but is not competing with you. What benefit could you give to such ventures or individuals and how could they deliver your message to your target customer
  • Chosen advertising channels include e-commerce portals, physical shops, news portals, blogs, and so on. It could be a direct gateway to potential customers, especially if you have a prototype and want to test the market fit of your value proposition.
  • “Go outside the building.” This may be one of the easiest ways to test some of your hypotheses: just go to the street and ask people. But be careful here. In order to deliver reliable results, you must be sure you’ve talked to your target group, not just random people.
  • Paid users might be a solution when none of the others is available in your situation. There are different platforms that help developers test their mobile apps (UserTesting, Killerstartups, Ubertesters, PreApps), websites (Enroll, StartUpLift, TestingTime), software, or even a physical product (Vyprclients, UserTesting). Most of those platforms are designed to help developers to find and fix technical issues. But some of them also allow testing the app, website, software, or in some cases even a physical product, by bringing beta testers from all around the world. Most of the platforms help developers track everything beta testers submit, like screenshots, photos, bug reports and also to have individual Skype interviews. But, in most cases, it’s much better to run experiments on real users, not paid ones.

Once you’ve found the source of potential customers, it’s time to engage them. Generally, there are four main types of engagement:

  1. Talk to potential customers—Describe the problem and solution and get their feedback. Check if you are on the right track. It is a rule of thumb that if you talk to at least 50 potential customers and 30—40% of them say that your idea is great and they’d love to try your product or service, it’s a good indicator.
  2. Show some kind of visual prototype—It might be just a napkin sketch, design mockup, presentation, or video. Get customer feedback once they see (not only hear) what you plan to deliver.
  3. Offer to buy or preorder—This is an essential point. Anyone can say that you offer nice products and they’d love to use them, but the moment of truth comes when you ask them to pay money for the solution.
  4. Ask to be recommended—Check to see if those customers who showed initiative to pay for your product would recommend you to their friends and colleagues. A net promoter score (NPS) is usually used in this case. The main question to be asked is, “How likely from 1 to 10 is it that you would recommend this product to your friends?” You could even go further and ask them to name their friends to whom you could talk to and show your product.

5. Collect new insights and update your business model

The more hypotheses you test, the more additional insights you get from your experiments and endeavors. Don’t lose them! Be ready to capture them! Prepare one file or folder where you will store all of these valuable insights and ideas. If you wish, you can use “Insights table” (template No. 2). It’s a convenient table that will help you make notes about all of the valuable ideas and findings you discovered while developing your startup. Make it as a separate file and upload it to Dropbox, Google drive, or any other cloud-based storage device to share it with your team. That will help you capture even more ideas and use them in more efficient way.

If experiment results end up changing your initial assumptions, you should replace your previous assumptions in business model with any newly discovered facts. Usually, after finalizing the problem-solution-market fit, the business model might have some major changes. You might have found and validated customers’ new problems and product features that would lead to you updating your value proposition. In case you found the minimum feature set that resonates with customers, it might influence not only the value proposition, but other parts of your business model (for example, cost structure, strategic partnerships, communication channels). Be sure to review and update them.


After successfully finishing these tasks you will have a validated business model that is based not on assumptions anymore, but solid facts. You may also come to totally new insights that will dramatically change your initial business model. If you keep the record of your experiments results, this will help you to build a solid pitch for investors and meet them with confidence.

Available to download Experiment card