AUSTIN, TX: The Home Depot, the home-improvement retailer, is tapping customer data in increasingly powerful ways to better serve its target audiences.

Melanie Babcock, senior director/agile marketing (audience), social media, adtech & display at The Home Depot, discussed this subject at SXSW 2018.

“We know so much about each of our customers now, and what they need, and their desires. And it isn’t based on demographics or psychographics. It’s based on this data,” she said. For more details, read WARC’s in-depth report: Home Depot’s retooled digital-marketing model.)

“And our ability to think about our customer has expanded exponentially over the last couple of years as we really dig down into: What does a customer mean?”

Elaborating on this theme, Babcock asserted that its granular view of the shopper has moved beyond looking at someone who is moving home to understanding the precise projects they wish to undertake.

“When you’re moving, you’re also redoing your bathroom, or maybe you’re redoing your kitchen, and you’re also buying exterior paint. And, all of a sudden, you’re no longer a customer for [set] products, but you’ve got a whole house renovation going on,” she said.

“So, how do I speak to you about all of that, and how do I prioritise which message I’m going to send to you first? It has really significantly changed what we know about our customer, and how we speak to them.”

The new ways of speaking might include leveraging systems powered by artificial intelligence to identify exactly the right faucet or water heater for a customer, utilising Google Home’s voice-controlled platform, or providing augmented reality tools.

As it considers which tools and channels to deploy, Babcock’s team is using a test-and-learn ethos – a strategy that generates vital learnings even when the return on investment falls short.

“The tests that we do can show zero return. You don’t want a lot of them. But those are great tests. It really helps you stop and think about: Was my data right? Was my messaging right? We build a lot of models in our company. Was the model scored properly?” she said.

“That really helps you stop, and think about every aspect of your job, and what you’re doing, and where the breakdown occurred. And then you can fix it.”

Sourced from WARC