Martin Kihn, Research VP at Gartner, will speak about the state of mar tech and ad tech at Programmatic IO on April 5 in San Francisco.
Ad tech and mar tech may not converge after all.
Two years ago, Gartner analyst Marty Kihn predicted that owned media (mar tech) and paid media (ad tech) technologies would come together (mad tech). But now he’s having doubts.
“It’s a nice, logical, picture. But I changed my mind,” Kihn said.
Instead, he predicts ad tech will grow up in a “parallel universe” to mar tech. The two will meet only in one place: “The nexus will be identity and analytics,” he said.
Kihn talked with AdExchanger about how he’s predicting the two spaces will evolve – and what that means for mergers and acquisitions ahead.
AdExchanger: What do you see in the market supporting your idea that ad tech and mar tech won’t converge?
MARTIN KIHN: The business models are different. Marketing tech works on a subscription model, which means it’s predictable. The media model [in ad tech] is high-risk and difficult to predict. And those budgets come from different places. Mar tech budgets are more capital expenses, with an annual planning cycle. They are often controlled outside the CMO world. CMOs have big paid media budgets, and they’re more discretionary.
Anything more intangible?
The talent is different. I come out of the ad agency world, and I arrived at Gartner, an IT advisory company. It felt like a different world. There is a difference between people happy running an email campaign and those that want to meet celebrities.
Mar tech promised the ability to extend CRM data into paid media. But you’re noting challenges due to the differences between paid media and CRM data.
The nature of the customer data is different. In mar tech, it’s PII [personally identifiable information] around a handful of people. In the ad tech space, it’s anonymized and extended to many more people. There’s different data, different depth and different scale. And the infrastructure in mar tech can be slower. It doesn’t have to be within 50 milliseconds. But it has to be more reliable.
This need to identify consumers may lead to the rise of “customer data platforms” (CDPs). What are CDPs, and how do they fit into the ad tech/mar tech landscape?
CDP is a controversial category that’s basically a DMP focused on first-party data.
Ad tech and mar tech are parallel universes, but they have points of convergence. And the big point is around identity; the marketing record. And the other is around analytics. You need one before you can do the other. The real core is identity, and it’s not just device mapping. It’s getting a deep picture of an individual, and being able to recognize that person. That’s the goal of the CDP: build up a nice comprehensive view of people.
Where are CDPs cropping up?
That’s why Salesforce acquired Krux. It has a lot of elements of a CDP. Adobe has its master audience profile, built on its Audience Manager DMP. The customer identity record can sit within the BlueKai DMP. Tag management systems and some startups market themselves as CDPs.
Based on your prediction that ad tech and mar tech won’t converge, what kind of acquisitions will happen this year?
I see two areas of acquisition. Anything around identity is hot right now: something that can map devices together, or people to records in the world, is a good asset. Oracle acquired Crosswise. Acxiom bought LiveRamp. The other areas would be around analytics, companies developing insight decisions and customers. Those are companies like Tinyclues, Optimove, Adgorithms and Lytics.
What acquisitions won’t happen?
We won’t see mar tech clouds like IBM, Adobe or Oracle acquiring DSPs. Or anything that plugs into the exchange world. The DMPs have almost all been snapped up already.
What does the stress in the ad tech market mean?
It signifies the growing dominance of two players and the bigger question of the role of others; the non-duopoly ecosystem. It’s hard to predict the future of the current ad tech ecosystem outside of Facebook and Google. And whenever something is hard to predict, it’s hard to value. And that’s why you see trouble with stock prices.
People talk a lot about ad tech becoming commoditized. But what’s become commoditized in mar tech?
I think every tech over time becomes commoditized. That’s just the way things go. The older ones are the most commoditized. Email has been around at least 25 years, and you can do it almost for free. Web analytics is a commodity for most marketers. A lot of things that were fertile ground for startups and had good margins become cheaper and cheaper, and then someone like Google or Amazon will start giving it away.
Where do you see commoditization in ad tech?
The value of an off-the-shelf audience, like someone in the market for a car, becomes less valuable over time on a generic level. And DSPs used to differentiate on their access to inventory. Now you can go to Iponweb and use BidSwitch, and any DSP can access inventory outside of the walled gardens. They have to differentiate on service and on analytics.
And how are companies in this space reacting to this commoditization?
They are trying to swim upstream. Salesforce, which has email, is talking about Einstein now. Where there is a lot of complexity, you can differentiate and build up a real competitive advantage.
Adobe bought TubeMogul, a paid media player. Is that the exception to the rule?
I think Adobe is different than other clouds, and the reason is that Adobe is two different companies: the Marketing Cloud and the Creative Cloud. One is a Salt Lake City-based, quantitative culture. The other is San Francisco and a hipster culture. They have two different cultures, and they’ve been able to keep this model working for them. If anything can pull off this parallel universe, it might be Adobe. And they’ll do it probably by keeping them separate.
This interview has been condensed and edited.