Blog

Conducting Sound Business Experiments for Strategic Decision-Making

Strategic Decision-Making

The most underutilized practice small businesses can benefit from is the scientific method. It’s a reliable system for testing hypotheses and conducting experiments. In the old days, people didn’t have a plethora of digital analytics tools to choose from. They relied on more rudimentary tactics for improving their businesses. Marketing, for example, certainly existed before the 1900s. There simply wasn’t a consistent word for it yet — nor were there platforms like HubSpot to streamline the process.

Technology has given entrepreneurs the luxury of automation for complex tasks like observing customers, generating leads and producing ad campaigns. We’re bombarded with sophisticated apps and software to assist us in learning more about our customers and improving strategic decision-making. But even without these fancy tools, collecting data and testing hypotheses about your business is totally doable — in fact, smart business owners have been doing it for centuries. All you need are eyes to observe and a brain to interpret.

 

Hypothesis and Setup

The first step to uncovering new information that can help your business is deciding what you want to know. Then, formulate a guess as to what the answer might be. For example, do customers tend to spend more money on days when you offer a discount? This is a simple question, and you might assume that the answer is yes. However, knowing for sure requires very close observation of your customers.

You have to ensure that your test setting has the change element in place (e.g. the discount) and that you have something to directly compare it to (e.g. the control setting without the discount). Essentially, this is A/B testing in real life.

One of the most challenging aspects of doing small business experiments is ensuring no other variables interfere, such as factor z causes a change instead of factor x. While you may not be able to rule out every external factor as scientists in laboratories do, you can at least make sure your research is done in the most typical business setting. Avoid drawing comparisons when external factors can skew results (holidays vs. non-holidays, weekends vs. weekdays, mornings vs. evenings, etc.).

 

Ask One Question at a Time

One mistake businesses often make is failing to focus on a specific factor. Harvard Business Review describes one major bank that was originally on the right track but didn’t quite follow through with their methodology. This led to inconclusive findings.

The bank’s intention was to improve customer service, a useful quest for any service-oriented business to embark upon. They were smart in creating “test” settings, certain banks where the customer service changes were implemented. Other locations were left to function as usual without the changes. However, the downfall of the experiment lay in a simple, but fatal flaw: changing too many things at once. Thus, even when improvements occurred, such as a reduction in customer wait time, the bank was unable to distinguish which change created the positive result.

 

Beware of Assumptions

If not tested, some of the core assumptions you have about your business can cause it to come apart at the seams. Are your foundational beliefs accurate? Have you ever questioned them through analysis? What you find may spur a profound change in your business plan.

For example, a manager might assume that high employee turnover rates are a result of long hours required on the job. But by introducing a simple mechanism for employee feedback, management might discover that this is not a major concern. Another issue, such as poorly functioning machinery, could be the cause of the company’s exasperated workforce. At this point, the real solution can reveal itself and problem-solving can occur.

 

Don’t Drop the Ball

Implementing relevant changes and sharing new knowledge is an obvious but often forgotten step. Once you’ve begun testing and drawing conclusions, you’ll want to create a process for storing and sharing so that the rest of your organization, including new employees, can benefit from the findings. Larger companies have begun keeping knowledge libraries, while others like Famous Footwear simply circulate study results and post them for all employees to see.

 

Build a Testing Process

Companies like Capital One and eBay engage in a constant process of randomized testing. Of course, each company has customized software and the benefit of a massive customer base, which makes split and multivariate testing quick and easy. But small businesses can replicate such efficiency on a smaller scale, keeping a list of key hypotheses, selecting test parameters and recording results that always give a distinct yes or no answer. Tools like SAS can assist you in nailing down a repeatable testing process, or you can manually design your own.

We all know that decision-making without any evidence or prior research involves risk — not to mention the possibility of wasting time and money. That’s why it’s critical to conduct your research carefully and repeat several times before coming to a definite conclusion. Organizations that have the most success and experience continuous growth are the ones that have adopted an experimental mindset, consistently testing, observing and tweaking based on what they learn. Equipped with this mindset (as HBR calls it, a “test-and-learn” mindset), you can sculpt and refine your business according to what works best.

 

Prioritize Experiments

Finally, recognize that not all assumptions are created equal. Some assumptions you make about your customers – if proven wrong – may cost you a few thousand dollars. Others may cost you your entire business. That is precisely why, as I’ve discussed elsewhere, the Business Model Canvas exists. It’s a powerfully simple tool developed with the help of over 470 cocreators to help you identify the riskiest assumptions to your business model, then start creating experiments designed to turn those assumptions into facts – starting with your riskiest assumptions first, then working your way down in importance over time.

To start building out your own Business Model Canvas for your company today, grab a copy of our free Digital Business Model Canvas Template.