Sean Byrnes
CEO and Co-Founder, Outlier
One of the great benefits of modern data-driven retail business strategies is the ability to test new ideas for consumers quickly. We are no longer reliant on the instincts of a few decision makers and are instead empowered by technology like analytics, A/B testing and behavioral research that gives us deep insights, very quickly.
There are so many facets to modern, omnichannel retail that we have only just started to scratch the surface of what’s possible in terms of understanding buyers. There may be tens of thousands of different adjustments a retailer can make every day to optimize sales, marketing, inventory and more. But most businesses only run dozens of tests, at most. Why?
The most obvious answer is that there are only so many hours in the day, and the people running tests and making adjustments are limited by time. But, modern tools like dynamic pricing, dynamic personalization and automated testing are helping to reduce the time and effort, allowing companies to run hundreds of tests instead of dozens.
In fact, the true bottleneck of testing is the rate of new ideas. Every test starts with a new idea or an insight that becomes a test that then becomes a decision. The more ideas a company generates, the most tests it can run, and the faster it can improve its business models.
What if there was a way to remove the bottleneck on ideas? With the rise of artificial intelligence and machine learning, there is. It’s easy to think of large retail businesses as machines that run according to a set of processes and hierarchies. But most businesses are really a massive collection of small entities, all running in parallel. Among those thousands of small entities are thousands of experiments being run, either on purpose or by accident. We only need to tap into those ideas in an efficient manner.
For example, one of our customers sells hundreds of items across tens of thousands of physical store locations. Using AI-driven automated business analysis leaders were alerted to an unexpected change in sales data. They noticed that a store that had temporarily closed for three weeks was now selling twice as many of a particular SKU than it had prior to the closure. This presented a significant opportunity as a high margin product. Upon further investigation, leaders found that the closure and renovations had created a more customer-friendly store layout, leading to higher sales of the product.
This single observation from an unexpected data change spurred changes for the entire organization, driving new revenue.
Such a story would never have been told before AI and automation. The effort to look at every item sold in every store in every place is too much even for the largest teams of humans. But AI can analyze enormous volumes of data and serve up actionable observations every day. In effect, these insights turn every store into an accidental experiment. Some of those experiments will succeed, and others will fail, but in aggregate, more tests will offer deeper and better insights. More ideas will be generated and tested, unlocking a much faster process of improvement and optimization.
Enabling a business to succeed with AI-driven business strategies requires preparation and planning, as there is no magic behind the scenes. To get a business ready to take advantage of AI, follow these steps:
- Collect raw data: AI is mathematics that works on data. Without raw data, there is nothing AI can do. It’s critical to collect the raw data of your business, such as point-of-sale data, individual page clicks and other records of consumer behavior. There is no way to know what data might be important to AI analysis, so gathering the rawest form of data ensures you have whatever you need when you need it
- Create a culture of insights: Organizations that are open to new ideas find it easier to adopt AI and uncover insights you did not expect. If your organization is inflexible or reliant on a concrete plan, you will find it difficult to take advantage of the new, unexpected insights driven by AI.
- Try before you buy: While we use the term AI consistently, it really is a category of dozens of different techniques, approaches and tools. Whether any single one of those tools or approaches works for your business might be hard to determine, so trying a number of tools is critical. The insights you need to run your business may be unique to your business, so if you can try before you buy, its critical to getting value fast.
AI promises to change the way we run our businesses, empowering us to make better decisions more quickly, leading to customer satisfaction and loyalty. The true promise of AI is to improve how we as humans run our businesses today, and to allow us to be more productive in our work than we otherwise would be on our own.
That’s an exciting future.
