An influx of advertising technology vendors in recent years has made our job of spot the difference increasingly tricky from the outside looking in. That’s why we often have to take the word of companies like Comcast as gospel, with the means and resources to identify a killer secret sauce when it sees one, something of the TVSquared flavor variety, for example.
As promised, after our tongue-in-cheek suggestion that Comcast should hurry up and buy the Scottish start-up, we got on the phone to TVSquared and dug about for its hidden ingredient.
“We get skeptical too – there’s never one thing. We worked on the Comcast Accelerate project back in 2014 and tackled a number of different attribution data challenges which we adapted to and saw how other companies were struggling to scale,” explained TVSquared CEO Calum Smeaton.
An almost lackadaisical tone attached itself to the way Smeaton addressed the enviable feat of getting a foot in the door at Comcast, although he assured us that breaking into the US market did not come without its fair share of scars.
Founded on the premise of a lack of analytics in linear TV advertising, TVSquared built a measurement model covering linear and OTT video, today claiming the largest enterprise-scale platform for TV measurement. Starting out in e-commerce, TVSquared’s data-crunching credibility was fine-tuned through optimizing high-frequency Wall Street models – eventually evolving into a platform active in 70 countries today with approximately 60% of revenues from the US market.
What sets TVSquared apart in our view is its hankering for longevity. The company focuses on the long-term effect of TV ad campaigns, modelling the impact in the weeks and months following ad airings, while identifying the best days, times, programs, networks and genres for response. Indeed, it claims the long-term effect of a TV campaign is as much as five times more effective than the initial response (from days to weeks) – with performance analytics able to capture ad stock and longer-term brand effect.
Attribution data has been TVSquared’s mantra from day one. Smeaton gave the example of a diet food client in the company’s fledgling years. “Early on, we saw that when data was loaded, these diet food ads were performing poorly on Saturdays, but then improved significantly on Sundays. This was a no-brainer; Saturdays are for junk food, right? So, when we started putting these ads on The Wedding Channel and The Food Channel on Sundays, responses were off the charts,” he said.
This is a super-simplified example in an increasingly convoluted market, but it helps to prove a point that identifying the best TV buy opportunities does not have to be rocket science. Although a set of proprietary machine learning algorithms and working with hundreds of data scientists at major clients certainly helps.
“We saw a trend of cost per word increasing, where Google was plugging all the free gaps online, and soon found that you could scale quickly on TV if you have the means to optimize,” said Smeaton.
Faultline was keen to pick apart TVSquared’s prediction tool, which is used to identify the best TV buy opportunities for optimizing spend – delivering what it calls always-on TV analytics at macro and granular levels. Smeaton wasn’t willing to give much away, noting that lauding a machine learning engine isn’t the answer – again pointing to how adapting based on customer preference is crucial.
For those unfamiliar with our recent coverage of TVSquared, the company has grown rapidly since its founding eight years ago. Its expertise is in driving campaign performance, promising a 25% reduction in cost per response and a 30% increase in sales or registrations – unifying the measurement of linear and OTT ad buys into a single platform.
Smeaton referenced TVSquared’s work with Comcast’s Effectv division (formerly Comcast Spotlight) as testament to TVSquared’s progress, training hundreds of salespeople in the art of attribution data. TVSquared was basically the attribution data brains behind Comcast’s Instant Impact initiative, built to dish out performance feedback to advertisers in a platform now integrated with ad tech vendor Ampersand (née NCC Media).
Again gravitating towards Comcast, our probing on this deployment triggered Smeaton to point to the lengths the world’s largest cable TV operator goes to where ad tech is concerned. Pointing us to a Beet.TV interview with Comcast Effectv’s Andrea Zapata, VP of Data Innovation and Insight, we learned how the operator is extremely selective in how it picks vendors. Ad tech hopefuls have to undergo a grueling process from proof of concept, to pilot, to scale. With Comcast present in north of 60 markets, with thousands of sellers, it is essential that it ultimately chooses technology that can scale appropriately with itself.
In a nutshell, what Comcast and its advertising clients are interested in seeing, by Zapata’s words, is did a TV ad drive traffic to a website? That’s it, along with a fast turnaround. Within 30 days, Comcast’s sales teams and its clients can log in to the same dashboard to view performance, analyzing whether there was immediate action taken within 30 minutes of an ad, and, if not, looking at how it can perform better.
Naturally, Comcast started out applying this approach to linear TV, now moving to VoD and anything with IP. Going above and beyond the core linear product is pivotal to becoming a data-agnostic powerhouse.
Of course, the big boys Nielsen and Comscore are among TVSquared’s biggest challengers. Nielsen acquired attribution firm Visual IQ towards the end of 2016, subsequently launching what it claimed at the time to be the industry’s first person-level TV targeting data management platform. Nielsen has been developing its hyper-targeted DMP to harmonize targeting between linear and digital, just like TVSquared.
“The death of TV has been predicted for years now. The question now is where should we put ads on streaming services and linear TV,” added Smeaton.
Speaking of predictions, Faultline made another of our own before the New Year, suggesting that TVSquared could be on the cusp of landing an ad tech contract supporting the forthcoming Peacock streaming service, set to launch in April. “We’ll wait and see for Peacock,” was all Smeaton had to say on that matter.