Arming the Rebels is a four-part series with Anthony Rotio, co-founder and co-CEO of GrowthLoop, for every retail media network that's fighting for their share. The series is brought to you courtesy of GrowthLoop.
A retail media leader recently told me how they got funding to build their network. They pitched the project internally on what it would do not for ad revenue, but for the marketing team: activating the retailer's own audiences, selling more of its own products, winning back shoppers who had lapsed. Media was a bonus, hardly even included in the main pitch.
I'd guess that pitch usually runs the other way these days, with everyone starry-eyed about all the money a media business can bring in. But a couple of other retail media network leaders have told me their first job when they arrived was fixing the lifecycle marketing setup, because the customer data was scattered everywhere.
Customer data split across an email platform, a transaction system and a loyalty program, none of it joined up, is difficult and expensive to activate. So for the second episode of this series, I asked Anthony Rotio, co-founder and co-CEO of GrowthLoop, why retailers end up here and how long it takes to get out. GrowthLoop's capabilities cut across both the retailer's own marketing, and the audiences its media network sells to advertisers.
This interview has been edited for length and clarity. You can listen to the full interview on the podcast feed or watch on YouTube.
Customer Data Scattered Everywhere
Kiri: A retail media leader arrives to build or revitalize the media business, and finds the customer data isn't ready to use. It sits in different systems: the email and lifecycle marketing platform, the transaction data. It isn't joined up, and it isn't trusted. Why does this happen?
Anthony: Let me back up and take it from the marketing side. I think this number might be dated now, but a CMO historically has been in seat for like 18 months on average in the enterprise. When that happens, sometimes you buy a bunch of different tools over the period of a decade, and data starts to aggregate kind of outside of your ownership. Some of it may be in an Adobe, some may be in a Salesforce, some may be over here in an email platform. And it's tough to look at that and see the light of how I can actually use this data to build this retail media network dream that I have.
The good news is oftentimes partners are closer than they think. It's not an insurmountable task to get the data you need centralized somewhere, whether that's in BigQuery or Snowflake or AWS or Databricks. There are plenty of tools to sync data now. So getting it in is step one: can we get the data landed raw? Step two is you have identities all over the place, so you need to join the identities somehow. But beyond that, the raw event data, the raw transactions data that you have somewhere in your organization if you're a retailer, those things can be used as is.
Typically what we find is when people see the mess, they see a three-year project. This is never gonna happen, how am I even gonna get started? And it actually can be reduced to quarters or months if you box in the problem a little bit.
Kiri: Advertiser audiences get built from the same data the marketing team uses for emails, app notifications and circulars. So the media business rides on top of all that, and inherits its data problems if it isn't set up correctly. How are retailers handling that overlap?
Anthony: We've seen this probably in the last two years or so only, where organizations want to centralize this and have one operating layer for both sides of the house, rather than a stack over here and a separate stack on the retail media side. You want to take the learnings you have on the marketing side from your experiments and write the results back to a place where you can use them, both on the marketing side and for the media side as well.
We're actually getting RFPs for this audience operating system layer for both sides of the house, and we now have about a half dozen customers that use GrowthLoop for both lifecycle marketing and their retail media network. It's really savvy when you think about the sprawl that can happen with this data. How do I simplify the stack in a way that I'm still owning the data and owning the customer learnings as the retailer, and not building and aggregating them in a system that relies on a contract that may not be in existence in two, three, four years down the road?
Kiri: The way you describe it seems so obvious. Why has it only happened in the last couple of years?
Anthony: A lot of the technology wasn't there to enable this previously. Marketing systems that are composable, that plug directly into this customer data layer [rather than keeping their own copy of the data], didn't exist more than six, seven years ago, which is when we pioneered this approach. Now there are a few others out there too.
This is more about working with partners that are going to innovate alongside you than picking the best possible piece of technology and saying it's going to be the right technology in five years, because who knows what five years is going to look like. My daughters are eight and five. As a guy who's been in technology for a long time, I always thought I could see pretty far out into the future of technology. Now I have no idea what world they're gonna grow up in. So when we talk about choosing vendors and choosing partners, it becomes more about who do you trust? Who do their customers trust?
Box in the problem
Anthony's version of the first few months looks like this:
- Get the customer data landed raw in one place, such as a cloud data warehouse.
- Join up the identities, so the retailer can tell that its email subscriber, app user and loyalty member are the same person.
- Use the raw event and transaction data as it is.
Every step on that list helps the marketing team before a single audience is sold to a brand. Which is probably why "marketing first, media as a bonus" was the easier sell for that retailer who found it necessary to pitch media as a data play.
Next up in the 'Arming the Rebels' series: Build, buy or rent. Why a retailer with a big engineering team might still rent some of its ad tech. And thanks to Growthloop for sponsoring this season of Retail Media Breakfast Club!
- Part 1: What's Holy? There's a tension between what customer data a retailer needs to hold onto, and what can be provided to brand advertisers who want to reach that same customer. Read here
