“Data Is A Chronic Disease That Marketers Will Simply Have To Live With”

Big data is like a chronic disease — we are just going to have to live with it forever, says Jay Calavas, chief innovation officer at US marketing tech firm Tealium.

Calavas – speaking in Sydney last week – says the company is well placed to take all it has learnt from solving the problems of tag management (where it has its antecedents) and apply it to bringing the firehose of continuous data flows that confront marketers under control.

Whether customers are using PCs, mobiles, set-top boxes, kiosks or i-beacons, the collection footprint and selection is endless, Calavas says.

“What we do really well is take all those different data sources, correlate and integrate them, enrich them into insights, and then make them immediately actionable across 900 different marketing technologies. That’s why we talk about Tealium being a data supply chain.”

Calavas had his first entrée into digital marketing with WebSideStory, which pioneered the field of real-time web analytics and was later acquired by Omniture. It is still used today in the Adobe Marketing Cloud. “That was a big moment of disruption in the early web marketing world,” he says.

After the business was sold, and after a stint at another user analytics business, Calavas landed at Salesforce.com — an experience he says gave him a good oversight into the kinds of issues confronting marketing cloud platform providers.

“What I saw was a state of disarray. There were point solutions but there was no true unification across the tools. They hadn’t built them. Instead they were buying technology companies and getting them to work together.”

For end users, though, that meant they were dealing with problems of fragmentation and miscommunication which made it hard to unify marketing activity.

He says the work that Tealium did in its early days — tying together different Javascript tag technologies from third party sites and then unifying that data and distributing in a common language and a common format — began to break down these data silos. “Now there is a common data dialect or data dictionary being used from a company’s perspective to communicate data off to third party providers. That solves one a big problem, which is marketing agility.”

The real payoff, however, is what happens next, he says. “We are always looking thematically at these big huge barriers and we want to punch our way through them with new innovation. The biggest problem of them all is the big data problem. It was a problem we all used to talk about a lot as an industry — now I feel we all whisper about it.”

The specific problem he is referring to is the ever-increasing flow of data from increasingly disparate sources.

According to Calavas, many marketers find that, by the time they have put the data into their systems, processed it and created segments, then found the particular needle they were seeking in the haystack, days — if not weeks — will have passed. “At the very least, we like to call it the 24-hour flu. It takes 24-hours before you can receive and act on the data, which is too long.”

Tealium’s solution is Audience Stream. It differs to previous approaches in that it does not hive the data off for processing elsewhere. Instead says, Calavas, “It leverages that common data dictionary and it allows you to begin selecting pieces of that data dictionary and enriching it in real time with customer segments, insights, and attributes.”

First versus third

On a first pass, Tealium may look like it is solving many of the problems which data management platforms (DMPs) were built to address. However, Calavas quickly dismisses the suggestion.

DMPs were created from retargeting engines, he says. “In reality, what they were doing was creating third-party cookie pools but putting their tags all over the web and then allowing companies to service them. They help you find people that look like this or that, or who have been to your web site recently and they help you put an ad in front of them. Basically the DMPs became a retargeting platform for display advertising.”

For that one use case they did a really good job, he concedes (though he adds that it was a good job based on a blurry view of the third-party data.)

And this is where Tealium is fundamentally different, he says. “We work with first-party data. We are a deterministic — not a probabilistic — company. That means we can say, ‘for sure we know this to be true and we are going to act’. It doesn’t mean that third-party data is not useful or usable in the right format but first-party data is king.”

Many Tealium customers still leverage DMPs, he acknowledges, however they are able to use Tealium to provide their platforms with first-party data to do their retargeting.

Looking ahead, Calavas says he expects to see a huge collapse between Demand-Side Platforms and DMPs. “DSPs are taking on more than that native retargeting functionality and if you leverage a first-party tool like Tealium with a DSP you can emulate a lot of that functionality.”

The piece that was missing, he says, was look-alike audience capabilities — and companies like Facebook are building in those features into their platforms, according to Calavas. “I think in the future, third-party data moves out of focus. All of the major Internet warehouses like Facebook, Google and Yahoo — the guys with the eyeballs — they are building one-to-one integrations from an identification perspective. What you will see is companies building direct integrations into each of the walled gardens. That is what Tealium has done.”

Calavas spoke to Andrew Birmingham from B&T’s sister business site www.which-50.com for this article.