Kirk Perry picked an interesting time to leave Google and start as CEO of IRI this week. He’s joining the analytics firm just as packaged-goods and other clients need media analytics more than ever, in part because of his old employer Google’s impending move to pull the plug on cookies, not replace them with any other third-party identifier and launch its own segmentation system.
Perry was president-global client and agency solutions at Google for nearly eight years prior to his new post, working with the biggest clients and agencies to familiarize them with Google’s platforms and transform their companies digitally. So, he knew their concerns. Prior to that, he spent more than two decades at P&G, and his move to Google came as a surprise given that he was widely seen by insiders as a leading candidate to one day become CEO there.
In an interview with Ad Age, Perry talks about what made him make yet another unexpected career move and IRI’s role in bridging the marketing industry’s many data gaps.
What brought you to IRI?
I spent 23 years at Procter, and I’ve been at Google seven and a half years. And so 31 combined years at big companies, great companies. I’ve learned a ton, but I thought he next chapter, I would probably go write a book, continue to do some board work, maybe consult. I’d coach kids sports. I had a whole different path forward. But when I got the call about this, I was intrigued for a number of reasons.
One is, I was intrigued by the size. It wasn’t the size of Google or Procter. The second thing was an intersection of both of my previous career experiences at P&G and at Google. And the third is, for me, the pandemic, being honest, reinforced that I’ve probably still got a number of years left in my tank. I am anxious to get back in the office to get back on a plane again, and kind of do some of the normal things that we used to do—just maybe not to the same degree before.
When I think about my P&G career. It was always about how do you figure out what the consumer did and why they did it, and how do you make the product and marketing and supply chain better. Google was that on steroids, faster.
You know that old axiom of 50% of your marketing dollars are wasted, you just don’t know which 50%. I always said this at Google since I’ve been there, if we can help advertisers figure out what that 50% is that’s the Holy Grail, and that’s something I’m very passionate about.
How would you describe what IRI is today and where its growth is?
When I was deciding whether or not I was going to do this, I talked to a few folks confidentially in the industry. And they talked about how one of the unique things about IRI is a product we have called Liquid Data, an end-to-end consumer planning tool that’s the most sophisticated because it combines hundreds of data sets with IRI’s public cloud. You can dump client data into it, and it really allows you to connect and uncover relevant patterns and use analytics to understand the action steps marketers should take.
At its core, it’s a data and analytics company. But with privacy evolving to where it is, there’s going to be even more criticality placed on first-party data and ability to integrate that [with other data sets] in a very privacy-centric way.
The honest answer is I don’t know yet. I like how you framed it as Google made the decision. Actually Apple made the decision before we did. The reality is that consumer choice is key here. A guy I worked with at P&G … used to talk about the fact that solving contradictions is what gets the greatest breakthrough in business—lighter and stronger, better and cheaper, smaller and more powerful. I think the contradiction [for the marketing industry] is how do you fuel the free and open web in a way that the consumer gets a choice. I can see it from both sides as a CPG advertiser and a digital and tech person. I don’t think the problem for consumers is ads. It’s intrusive, crappy ads. [P&G Chief Brand Officer] Marc Pritchard will say that it’s how do you make advertising that’s actually good? I think as an industry we haven’t done a good job of making sure that it’s relevant, delivered in the right time in the right place and all the things consumers expect.
How important is good old-fashioned retail scanner data to IRI today?
It’s still a big and important part of the business. I was telling someone a story the other day of how when I was a brand assistant on Crest, my IRI contact would drop off two three-inch, three-ring binders loaded with data just on Crest, and I would eagerly away it once a month. I would literally work like 18 hours kind of loading everything into one computer. Obviously, it’s come a long way from that. If you don’t innovate, you’re going to be an AltaVista or a Blackberry or a Netscape Navigator.
On that point, folks are trying to put more omni into omnichannel sales measurement. I just got a presentation recently from Numerator about its TruView product, which is impressive. And IRI recently made a big acquisition in that space. Are you going to be able to capture everywhere your clients’ products are sold?
I don’t think it’s a question of if. It’s a question of how fast and how completely can we do it, because it is critical that there is an omnichannel look at the consumer. Nielsen is looking at it. Obviously Numerator is looking at it. And there’s a project that we were working on with the [Association of National Advertisers and World Federation of Advertisers] on cross-media measurement, so everybody is doing this, and we have to do it well, whether it’s an industry consortium or individual companies. It’s not easy to solve.
It seems like the solution for retail data has to be in the form of really big panels as opposed to direct data feeds, because you’re never going to get them from all the retailers, right?
Yes and no. I will say yes because obviously that’s the technology available today, but no from the standpoint that there could be other technical solutions to bring this together. Maybe if you have [almost] every media player [in a consortium], it’s to their peril not to be in it. And if you have a retailer that’s not part of [data sharing] maybe it’s to their peril if they choose not to be.
Google, with FLoC [Federated Learning of Cohorts] is learning interesting things about not being able to have one-to-one data but being within 95% accuracy of knowing what’s happening with that consumer. And I think as technology moves forward, there may be more ability to do things like that, allow you to not necessarily have these gigantic panels, which are incredibly expensive, incredibly complex to run, but how do you marry a little bit of that panel work with great technology that brings those things together.