An interview with

Encima Group's Domenico Tassone

Aug 13, 2014

On Digital Measure Tools and Display Attribution; A Q&A with Domenico Tassone

Illustration by Graham Erwin

As head of the Digital Capabilities Team at The Encima Group, Domenico Tassone has more hands-on experience with digital measurement tools than perhaps anyone else on earth. And he’s spent a fair amount of time navigating thorny issues, like question if media vendors, agencies or channels are really contributing to sales at the level they claim they are. But at the end of the day his goal is simple: Get the data that answers his clients pertinent questions regarding their buying journeys. Ben Plomion, CMO of Chango, asked him to share some of insights he’s gained in the field.

Let’s start with the Encima Group and your role there.

The Encima Group is an analytics consultancy offering standalone measurement and reporting services, analytics implementation services, tag-management services, as well as more maintenance site analytics and tag management platforms.

Our approach is a bit different because we’re not an agency; we don’t plan or buy media, build websites or creatives. We act as an independent arbiter to help clients understand their customer buying journeys, and accurately measure performance of their agencies, media vendors, channels and websites.

I lead the digital capabilities team, and we ensure clients have the tools and data its analysts need to understand performance

When is The Encima Group called in to a marketing organization? And Why?

Typically we’re called in when clients feel they can’t trust what their media partners and agencies are telling them about performance. Don’t get me wrong; I’m not accusing anyone of not telling the truth. But often clients work with numerous agencies, and they’re poorly coordinated. As a result, the client receives conflicting data, and it falls to the analytics team to figure out what happened. That’s when they’ll call us.

Let’s talk about the measurement tools. Do analysts have what they need to understand attribution?

Most of the tools are pretty basic, especially with regards to digital marketing channels. They tend to be click-focused, which is why we see big gaps in measurement. The truth is, clients spent a lot of money in display, but they’re not getting good measurement as to what happens after exposure to a display ad. The reason, in our experience, is that they don’t have the right tools, and they’re not set up internally to assess all of their channels and media vendors holistically.

Digital has been around for 15 years at least. Why is measurement still falling short?

There aren’t a lot of approaches available on the market to do it well. The media vendors, particularly the ad networks, can provide their part of attribution story when it comes to the passive response of display. The problem is that each ad network can only tell a small part of the story.

The second issue is with ad ops. Ad servers aren’t always set up to traffic campaigns correctly, and to gather consistent data. Added to that, the traffickers are often the newest hires in a media agency, and they lack the technical experience and skills to run complex campaigns.

Finally, there are issues with ad-server functionality. Some ad servers – such DFA and Atlas – use tags to record a second post-impression activity, which gives a better representation of display. But neither one can marry that activity to granular site-side activity.

Is it even possible to link campaigns to site-side activity?

You can, but you only have two independent choices. The first Adobe Genesis integration, which uses a cookie to link the ad server to the site-side analytics system. The other is Coremetrics, which bakes a site analytics tag into the creative delivery tag on the ad server.

In theory these solutions let you see across all media and all campaigns and marry it up to site analytics performance to assess performance across the buying journey. But it’s a level of measurement most clients lack.

Elaborate on the passive affect of display ad.

We know that an ad has been served and we know that it was seen, but how do we assess its influence on a consumer if that consumer didn’t click on it? It’s possible to do view-through measuring using cookies to essentially link what happened between exposure and a visit to a destination. Unfortunately, it’s still an under used metric, but it shouldn’t be since historically it’s been used for TV, radio and print.

Digital is more complex to be sure, but it’s also unique in that we have the potential to measure the incremental value of a display ad, and to see what’s really moving the needle forward.

Who needs to drive the movement, shall we say, to accurate measurement and attribution?

It has to come from the brands. Part of the solution is technical, to be sure. But an important part culture. Too many brands have separate teams for display, paid search, affiliate, email and so on, and no one is coordinating the overall effort. And every team believes they’re responsible for a conversion, because they’re protecting turf, frankly, so there’s no incentive for them to look at all marketing initiatives holistically.

Externally, the brands need to ensure their agencies know they’re serious about this stuff. I worked with a client that needed to nudge its agency into accepting the importance of viewability. Once the agency understood the brand was serious, it implemented a viewability solution, and the number of ad that were out of view dropped from 35 – 40% to just 15 – 20%.

More than that, most agencies have one or two channel specialties, but there’s a lot more going on in the buying journey. So the brands need an independent arbiter to measure what’s really working, and what’s not.

There’s a notion – especially in the C-suite – that everything in digital can be tracked.

It can be, but it’s not easy. When I worked on the brand side we used the Adometry platform, which was an independent tool that helped tell the display story, and measure incrementality.

But all to often, whatever is easy to measure is what’s get tracked. And once people are in that mindset of tracking something they stop asking questions. As a result, measurement gets out of whack with reality. Do people really think that paid search is as causal as it appears to be? If you ask the search agency you’ll hear “or course, it’s even more.”

Albert Einstein said, “Everything that can be counted does not necessarily count; everything that counts cannot necessarily be counted.” This quote is printed on the back of my business card.

What are some of the exciting developments in attribution you’ve seen?

We’re starting to see a solid algorithmic approach to dealing with attribution, which essentially lets the math decide who gets what credit and how that’s divided up. The problem with the algorithmic approach is that there are only two vendors: Adometry and Visual IW.

Most vendors are rules based, and rules are inherently biased to the views of their creators.

How do marketers measure their buying journeys?

The best approaches I’ve seen do statistical analysis of, “Here’s a sequence of events spanning multiple media partners and channels that form a path. What’s the probability of a path leading to a conversion if this particular channel is absent.”

The other aspect is using measurement to back into how display affects in-store sales. Here’s what I mean: If a consumer buys in-store and the conversion event is pulled into a CRM file, you can then link it back to web-browser behavior. You still have to do some incrementality analysis of online-offline, but you can see interesting things.

What are some of the more surprising influences that you’ve seen?

At Sears we saw ad networks outperform some of the pricey television deals that had digital components. Our assumption, in the pre-attribution days, was that the ad networks didn’t work. But we saw strong performance and conversions. It was counter-intuitive, but we measured the influence of display on lower-funnel activities.

How should marketers optimize their campaigns to ensure they get good measurement?

Start with viewability, and if you have good measurement in place you should be able to optimize on some of these more important metrics, such ROAS, engagement rate, conversion rate, and registration rate.

Can marketers optimize across channels?

In theory, yes, because many of the channels ate click based. Display will always more complex because there are so many media vendors that will affect the consumer experience. But in general, you can standardize and optimize if you use Google Analytics and create standard rules for how agencies track click-through URLs for pain search, social, and email.

A more sophisticated tool, such as SiteCatalyst or Coremetics, support CID or SID codes. This is when some kind of a unique key is passed, and found on a mapping file that’s periodically uploaded into the site analytics system. In these cases you get all of the detail, including campaign name, media vendor name or placement name, keyword group, ad group, that kind of stuff. It can be complicated to set up but once it’s done, reporting becomes a lot easier.

Download our white paper on "Attribution for the Masses' here!