What should be tested before investing in a new business?
Test customer problems, willingness to pay, revenue assumptions and operating capabilities before scaling. Compare interviews, surveys, pre-sales and small pilots.
What should be tested first in a new business?
Before investing in a new business idea, the order of business is to verify the hypothesis, which could shake up the entire business if it is wrong. Write down the customer's problems, willingness to pay, profit structure, and our company's execution capabilities separately, and test those with the lowest verification costs first. This order is also supported by randomized experiments showing that startup teams that are trained to formulate hypotheses and test them rigorously perform better.
Why should new business ideas be broken down into hypotheses?
There are several assumptions tied together within the term “good idea.” The assumption that the customer is actually experiencing the problem, that they will pay to solve the problem, and that we can keep the money after receiving the price must all be correct for a business to be successful.
A paper (Zott, Amit, Massa, 2011) that broadly reviews business model research considers that a business model is a system that explains not only how a company reaps value, but also how it creates it. If you write down the side that creates value and the side that reaps value, it will be revealed which assumption is weak.
Let's say a manufacturing company has designed a subscription service that analyzes equipment inspection data. The assumption that equipment failure is a major loss to the customer, that the customer will submit data to external analysis, and that the monthly subscription fee can cover the cost of analysis staff must be confirmed in different ways.
What should I divide the new business hypothesis into?
When reviewing new businesses and business models, Korea Research Institute of Business & Branding divides customer value, profit model, business opportunity, and market verification. The four things can be translated to be useful in decision making as follows.
Four axes that divide new business hypotheses
| hypothetical axis | Questions to Write Down | What happens when you're wrong |
|---|---|---|
| customer problem | Who experiences this problem and under what circumstances? | Even if you make a product, no one is looking for it |
| willingness to pay | Are you currently spending money or time to solve that problem? | There is a lot of interest but no sales. |
| profit structure | Is the remaining profit from one customer greater than the cost of acquiring it? | The more you sell, the bigger the loss. |
| Execution Capability | Does our organization have the people and system to run this business? | There is demand, but supply cannot keep up. |
Which hypothesis should you test first to save money?
There are two criteria for determining order. How fatal it is to be wrong and how uncertain it is now. In both cases, the higher hypothesis comes first. If the rankings are the same, the cheapest and quickest way to check is used first.
Cost of each verification method and what you can know
| Verification method | cost | What can be known | limit |
|---|---|---|---|
| customer interview | low | What the problem is and how it is currently being solved | What you say and what you actually do may differ |
| survey | low to medium | Breadth of customers experiencing problems | Results may vary depending on question design |
| Pre-sale/reservation | middle | actual willingness to pay | Initial response does not guarantee repeat purchases |
| Small-scale pilot operation | medium to high | Operating capabilities and costs | As the scale increases, conditions change. |
The methods in the table do not replace one another but are used sequentially. The common flow is to confirm the problem through an interview, measure the extent of the problem through a questionnaire, determine willingness to pay through pre-sale, and then confirm cost and operational capabilities through a pilot operation.
Do hypothesis testing methods really change performance?
There is a randomized experiment (Camuffo et al., 2020) that followed 116 early-stage Italian startups for about a year. Both groups received 10 sessions of training to obtain market feedback, while only one group was taught how to create a framework for predicting the performance of their ideas and rigorously test their hypotheses. The results showed that startup teams that thought like scientists performed better, were more likely to pivot to other ideas, and were no more likely to give up in the early stages.
However, the target of the study is early start-up companies. When transferring and applying it to an existing company, differences in organizational size and decision-making structure must be considered.
If the verification results are ambiguous, how do you decide whether to proceed?
Before you begin the test, write down what results you will want to continue, change direction, or stop. If you receive results without any standards, it is easy to interpret them in the way you originally believed. If you see a lot of customers showing interest but few making advance reservations, it could be a sign that the problem is real, but the solution or price isn't right. If price seems to be an issue, try using the calculations for Things to consider before raising prices. Korea Research Institute of Business & Branding divides the review results into alternatives such as continue, modify, and discontinue, and writes down conditions, costs, expected effects, risks, and execution order for each alternative. For detailed instructions, please check Management strategy advisory.
How is the verification of a new business for an existing company different from that of a start-up?
Established companies already have the customer base, brand, and funding to speed up verification. Conversely, resources must be shared with existing businesses, and the possibility that new businesses may encroach on existing customers must be considered. As discussed (Teece, 2010) that a business model refers to a method of delivering value and converting payments received into profits, it is recommended to list the points where the existing business model and the new business model collide. If the in-house new business team is evaluated based on the same sales standards as the existing business division, small experiments in the verification stage are likely to look like failures, so it is also a good idea to have a separate indicator appropriate for the stage. The initial tasks of starting a business can be viewed in the field of start-up and small and medium-sized business growth, and the entire field can be viewed at Management research and advisory field.
New business verification Frequently asked questions
How long does it take to verify a new business?
It depends on the number of hypotheses and the method of testing them. Confirming critical hypotheses first through a quick method like an interview can help you find direction before making a major investment.
Are customer interviews enough?
Interviews are great for confirming the reality of the problem, but they don't demonstrate willingness to pay. It may be a good idea to combine this with methods that see real action, such as pre-sales or pilots.
If the verification results are negative, do I have to close my business?
It may mean that one hypothesis is wrong. Look at which assumptions can be changed; customer, price, or delivery method; and compare alternative modifications.
Can new business verification be done with Korea Research Institute of Business & Branding?
Korea Research Institute of Business & Branding handles new business and business model development in strategy and business development advisory, and initial customer and market verification and business model and profit structure inspection in support of growth of start-ups and small and medium-sized businesses.
What should I prepare when entrusting verification to an external agency?
If you prepare a document listing your ideas as a hypothesis list and customer data you already have, verification design will be faster.
References
- Management Science, 「A Scientific Approach to Entrepreneurial Decision Making: Evidence from a Randomized Control Trial」(Camuffo et al.), 2020. https://doi.org/10.1287/mnsc.2018.3249
- Journal of Management, 「The Business Model: Recent Developments and Future Research」(Zott, Amit, Massa), 2011. https://doi.org/10.1177/0149206311406265
- Long Range Planning, 「Business Models, Business Strategy and Innovation」(Teece), 2010. https://doi.org/10.1016/j.lrp.2009.07.003