This blog post is co-written by Yngve Dahle and Mark Robinson.
If you are launching an all-new business idea, you can easily miss both the right problem and target group. If you do, your business most likely will fail. Therefore you should consult the “Lean Startup” principles. Do not use resources for developing products that solve a problem you think the target group has.
Working lean is the best insurance against solving non-existing problems. Continuously measuring and evaluating objectives, tasks, revenues and cost, you will easily reveal whether the market wants your product or not. So the everyday running of your company will be the best test you can perform.
However, to add another level of security, we suggest testing three different places in the method:
Working with the business idea, we suggest you make sure the target group really does have the problem you think they have. In the business model, you should make sure that your product or service really solves the problem. Here you must review all the boxes in the business model. Will the sales and pricing model work? Have you chosen the correct early adaptors and partners? You must test the business model as a whole.
Finally we recommend that you make sure the market is big enough for you to profit from it, and that it is not overcrowded by competitors. You can do this when working with the Gap.
Looking at Steve Blank’s customer development method again, we see that these three questions correspond with the first three phases of the model. If you have an affirmative answer for each question, you can continue to the next step in the model:
Do you have a problem worth solving?
The first test will assess whether your chosen target groups really have the problems you think they have. We recommend performing this test as simply as possible. Find a few representatives for each target group, and interview them over a cup of coffee. You can also e-mail them a questionnaire. The point is to do something that is practical.
Don’t spend more than half an hour on each interview. Start by selecting your interviewees. You do not need many. Begin with 10-12 and add more if you need clearer answers. The most important thing is that they are good representatives for your target groups.
The interview itself is simple enough. Start by describing the problems you try to solve, and have the customers rank them from “small problem” to “large problem”. Try to get them to provide additional comments to each problem. Let the customer suggest his own, related problems. As you conduct the interviews, you will see each customer groups average ranking of each problem. Based on this data, you can choose to continue developing the idea, improve it, or disregard it and test another idea.
The customers may exaggerate how big the problems are. By asking the questions you may introduce a problem that wasn’t originally there: “wouldn’t a cold beer have been nice right now?” Take care so you do not influence the customer to give you the answer you are looking for. For more information on designing the tests, check out “Running Lean” by Ash Maurya.
Do you solve the problem?
Now we will test the business model by asking: “do you solve the problem?” You must ask these questions to your early adopters.
It is important that we test all the relevant questions in the business model. “Solving the problem” does not only apply to the quality of the product. A good value proposition and a good product is no good if you sell it in the wrong way, or have a catastrophic price model.
You can answer this question in two main ways: You can ask the customers if their problem is solved, or you can observe the customer while simulating using the product.
Ask the customer
This is of course the easiest way to test the model, but you can’t be sure that the customers actually tell the truth. They might consciously give you a wrong impression, either to be polite (well behaved people might actually be a problem here), or because they have another motive. More likely, they will unconsciously misinform you. Perhaps they think their problem is solved, while it really isn’t. When they become aware of this discrepancy, they will stop using the product.
A customer interview is quite straightforward. You contact a number of potential customers and ask them what they think of your product, sales method and price. We suggest you start by demonstrating how you plan to solve the problem. You will probably not have a finished product to show, so the challenge will be to explain the product to the customer. If you have a Minimal Viable Product, you can use that. If you don’t, you will have to make some kind of model, or simply a PowerPoint presentation or video. In the worst case you can just tell the customer what you plan to make.
After the customers have seen the product presentation, you can describe all the features you want to include in it, and ask the customers to rank them from important to unimportant. Ask them to rank your different possible sales and pricing methods the same way. Let each customer explain why they rank the way they do, and let them respond with their own ideas on features, sales methods and pricing. Always try to make the customer elaborate on their answers.
Simulate using the product
If you wonder whether a direct question will work, you can simulate some kind of problem solving. If you want to know if pedestrians on a specific street corner need coffee, you can set up a pop-up coffee bar in an empty store, or park a mobile coffee-truck there for a few days. If you get a decent sale there, a permanent coffee bar might be viable. There are numerous examples of such simulations:
The Landing Page Test
If you market your product on the internet, a landing page test can tell you whether you solve the customer problem and if you have selected the right early adopters. The most famous example of a landing page test is the online file storage service “Dropbox”. The founders thought that people had a need to store and share data in the Cloud. However, they did not want to spend a lot of money on making the product before they knew what functionality the customers really wanted. Thus they “tricked” the customers to believe they had a real product, while they only really had a video simulating the service. Advertising on Google AdWords or Facebook can generate the traffic to the landing page.
If you have a list of potential customers, you can email them directly and try to draw them to your landing page. A landing page test can tell you if your sales methods work or how you should charge for your product. If you get enough traffic to your landing page, you can split the visitors into different groups and test different price models on each group. This is called A/B tests, since group A is presented with one price model and group B with another. If the different price models create very different sales volumes, you can learn a lot about how to price your product.
The Crowdfunding Test
Crowdfunding lets potential customers purchase your product online before it is produced. If you get enough orders, you can produce the product - and those who ordered it get early access to the new product. If you do not get enough orders, the money is paid back.
If you wonder if someone will buy a wristwatch that runs programs from a smartphone, you can sell it on the webpage kickstarter.com without even having a prototype. Everybody will accept that the product only will be built if enough people order it. This service is tailored to test whether you solve the problem, and by 2013 it financed more than 50,000 new products.
Crowdfunding is a perfect way to decide whether the problem you are trying to solve really exists, but it can also teach you a lot about your business model. However, performing a crowdfunding test involves a lot of work. You must define your minimum viable product very clearly. A picture and a description of the product is an absolute minimum. Making a video is even better. A crowdfunding test can help you to test if the product solves the problem or if you have a customer segment willing to pay for it. Since you need to deliver the product if you get enough interest, you can also test your production partners and decide what production costs you will incur. Customers you contact through crowdfunding are often very enthusiastic and will give you useful feedback concerning the functionality and price of the product.
The Concierge Test
The concierge is the hotel employee that fixes everything to make your stay unforgettable. He can obtain tickets to the theatre or ice hockey matches and help you with all sorts of useful things. All these tasks are done as a personal, manual service to each individual guest. Performing a Concierge test involves much of the same. You simulate an automated service by doing the job manually. The best known example is probably “Food on the Table”. The founder of this American company had an idea that people want a guide telling them how to use discounted groceries from their local supermarkets in recipes tailor made for their taste. He did not want to spend a lot of money before he knew what kind of product the customers really wanted. So he “tricked” the customers into believing there was an automated solution while he actually did all the work manually in Excel. A concierge test will give you feedback on whether you find the right customer segment, charge the right price, and have the right partners. You will also get a good estimate of how much money you will need to finalize the product. The term “Wizard of Oz experiment” is often used to describe a concierge test.
Is the market big enough?
Previously you asked whether you had a problem worth solving and whether you actually solved the problem. You tried to answer both those questions by doing tests with your customers. Now you can test whether it is possible to make money from the chosen target market:
To answer this question, you must attack the problem systematically. You need to go through these five steps:
How many people have the problem?
Finding the total number of potential customers for your business is far from simple. Again, you must start with what problem you solve and for whom you solve it. If you have a clear answer for these two questions, you can use publicly available statistics to figure out how many potential customers you can get. Clayton Christensen could say that he was in the market of fighting traffic-boredom for morning commuters. The total number of potential buyers would be the sum of all bored morning commuters in cars between six and eight a.m. in the geographical areas he would operate. Perhaps you could say that the bored commuters would be the people already buying milkshake, coffee, pastry, chocolate and fruit?
If Christensen decided that he would rather solve the problem of satisfying the hunger of families with children, the market would naturally look completely different.
This example shows that the number of potential customers is defined by the size of your demographic target group in the geographical area you want to cover.
If you are solving a brand new problem, this becomes even more difficult. When Apple launched the I-pad in 2010 it must have been close to impossible to predict the size of the tablet market. No one had ever heard of such a gadget before! It was a similar situation when Facebook was launched in 2004. If you had told the founders that a third of the population of the world would be their customers by 2015, they would hardly believe you.
The only way to forecast a total market in such a situation is to ask the customers whether they will buy or not, and then calculate it. This is not by any means an exact science, and it may be wise to use some hefty security margins.
How many fit the business model?
Of course, you will not be able to serve all of the people in your target group. The first limiting factor will be the business model you have chosen. The product features you include in your product will not satisfy all the customers. The price level and model you selected will disqualify further customers. Not all will be able to afford your product. Finally your chosen sales model and distribution partners will not be able to reach everyone in your target group. All these things will limit the number of potential customers.
How many will not go to competitors?
The next thing to take a bite out of your pool of potential customers will be your competitors. First define your competitors and discussed where you are stronger and weaker than them. Can you use this information to assess what share of the potential customers will go to them, and what share will go to you? This will never be an exact science, but will you end up with 10% of the market? 90 %? Try to play around a little bit with different alternatives.
The precision of these estimates depends on how clearly defined your industry is, and how well established you are in this industry. Global market shares for computers and cellular phones are defined down to decimals. Unfortunately, most start-ups do not have this data. They are most often in markets where the total customer size is uncertain and the market share of the different suppliers is even more unclear. How many customers can you handle with the resources you have? Perhaps you have enough resources to handle them all, but there is a chance that the number of employees, machines or buildings will make it impossible to deliver to all of them.
For example, imagine you run a men's hair salon in which you are the only hairdresser. Say that your hair salon is in a small town with 8,000 people. Half of the people are men, and out of these 50% are adults that go regularly to the hairdresser. That means the hair salon has 2000 potential customers. Your business model is to only cater to drop-in customers in the shopping mall where it is situated. You know that a quarter of the population of the town never visits the shopping mall. This means that the chosen business model cuts the potential customer group down to 1500 people. The shopping center also has another hair salon. This is a mixed sex salon with one hairdresser. You estimate that you can get two thirds of the men coming to the mall as customers, which translates to 1000 customers each year. If each of them cut their hair 4 times a year, you can sell 4000 haircuts each year.
Then you estimate how many customers you can serve each day. If you spend half an hour on each customer you can theoretically serve 15 customers. If you include breaks and a little downtime, a realistic target is 12 customers per day. With 20 working days each month for 11 months a year, you can manage 2600 haircuts a year. This is fewer haircuts than you think you can sell. Now you have an important decision to make: Should you employ a half-time hairdresser for the 1400 extra haircuts? You have space for him in the salon. You see how this test may make you want to tweak your business model or add to your resources? If you do not have enough capacity, is there a specific cluster amongst your users you should focus on? Maybe one customer segment will be willing to pay a higher price than the others? Perhaps you can find customers that will be better suited as references than others, or that can give extra valuable input to the product development?
What is the value created for each customer?
The last thing you should do is to try to determine what revenue you can get from the potential customer pool you calculated. This will allow you to find the economic value of the problem you are solving.
Continuing the hairdresser example: With a price per haircut of $30 (excluding VAT), you will make approximately $360 per day. This generates approximately $1,800 per week or $80,000 a year. If you employ the half-time hairdresser, the yearly revenue can grow to $120,000.
We hope that this blog post may inspire you, and perhaps start a discussion. If some of the entrepreneurship terminology used is new to you, you will find it better explained in the Business Idea chapter of our Lean Business Planning book. Please download a free preview PDF of the book.