It used to be that if you wanted to know how your customers interacted with your product you only had one questionable option: you asked them.

This approach, formally known as the focus group, had many problems. 

Since a focus group is all volunteers, any group of participants was biased—only a specific type of person volunteers to sit in a room full of strangers and share their feelings about soap, TV shows or automobiles. Social pressures also no doubt kept people from sharing their honest opinions. The way the sessions were conducted, from the content of the questions to the way they were asked, influenced the outcome, and results could be skewed to align with pre-existing opinions. Above all, very few rooms are big enough to house entire markets, so while a focus group might provide a snapshot of what certain consumers thought, it was at best a very small subset.

Technology has merged marketing and market research. Instead of the two happening as separate, disjointed pieces of a feedback loop, the relationship between customer and company can be closer than ever and better for both.

This type of consumer feedback was the “least bad” solution: the only other option being no data collection at all. Phone surveys and feedback cards and their digital alternatives increased the scope of these campaigns, but did little to address the selection and questioning biases at their core. Questions about why a customer chose one product and not another remained too nuanced, impossible to distil into a mass-produced, multiple-choice Q&A card. Any and all information gleaned was at best incomplete.

Today, technology has merged marketing and market research. Instead of the two happening as separate, disjointed pieces of a feedback loop, the relationship between the customer and the company can be closer than ever and better for both.

Online, a practice called “A/B testing” lets marketers run endless experiments to determine how to make their websites more effective at getting users to stick around longer, share a page to social media, or convert from visitors into customers. 

A/B testing works like this: make two versions of a given webpage, version A and version B. Show A to half your users and B to the other half and watch what happens. Different versions have different impacts on customer behavior, pick the version with the best outcome to be the new standard, then create a new experiment and try again. 

It’s this approach that was used to streamline Barack Obama’s 2008 campaign donation page to optimize donations, and this is the same logic that’s used to figure out which headlines you’re shown on the Huffington Post. A/B tests have consumed much of internet and are such an effective technique that they’re now being used offline. 

Real life applications are still new, but the practice is already changing the way product packaging design works.  Back in the days of focus groups, a team of designers and artists would put together some packaging that they thought would appeal to consumers. They’d show it to focus groups, adjust it to the best of their ability, then produce it at scale. With offline A/B testing, marketers are now going to market with a number of alternative package options. Rather than making any assumptions about what customers like best, they let customers speak for themselves—by buying or not. Based on which designs are the most successful, the design of the product evolves by consumer selection. Marketers do the first 80% then let customers guide them the rest of the way home.

Often times picking a product based on its packaging marks the start of the customer experience, not the end of it. It’s what happened once the packaging is torn off that determines if the company got a sale or a customer. Today it’s only consumer electronics that report real-world usage information to their creators, but the incredible rate at which the cost of network-connected computing is dropping means that soon every object will have software and collect usage statistics. 

When this happens, the A/B testing (with appropriate privacy controls) could continue past the point of sale and be an ongoing experiment, with customers everywhere collectively working together to make their products and services better.