At FT, High-Tech Finally Starts to Simplify and Clarify Data
The high-tech systems that have been adding layers of data and real-time operations to the publishing and advertising markets in recent years also cake on additional costs and a comic level of complexity. The dizzying marketplace schemata developed by Terence Kawaja illustrates is one part map and one part joke. Media buying has become so complex and algorithmic that agencies are looking for math majors. Executives wonder where the money is going as each new tech piece adds to the CPM.
The problem is acute for a publisher like Financial Times, where a complex metered access model has to serve un-registered, registered and paid members differently, and those interactions render mountains of demo, behavioral and ad performance information. To date, the tiered access models has served FT well, as it just recently surpassed 250,000 digital subscribers. “One of the downsides to our model potentially is that all of a sudden you have a lot of data doing a lot of things,” says Jon Slade, Global Digital and Strategic Advertising Sales Director.
Traditionally, FT has leveraged its data well to nano-target for advertisers – to ferret out in its logs the tight niches not accessible elsewhere. “If you are looking for executives in pharmaceutical companies in Ohio, then we can do that for you,” says Slade. “We are in a world where supply is outpacing demand.”
The technology helps them fight off price suppression and maintain premium prices. But the technology also adds complexity and additional buckets of disparate information. Demographics sat in one database, ad inventory in another, and so on. In recent years the addition of new technology has been stultifying. “Up until 12 to 18 months ago I would have said that every layer of technology we put into the business added X% more complexity, and it was difficult from the outset to understand that the investment had been well made.”
Slade says that from a publisher’s perspective he has started to see a “sea change” lately where technology solution are starting to take the reigns of the massive complexity they have built and started to simplify and clarify. He has been working with analytics provider Metamarkets to pull the strands together so that he can see things in the data that just weren’t apparent before.
All of those silos of information, inventory, audience, usage patterns, etc, had to be pieced together in the past to respond to media buyer queries in a day or two. “From an ad director’s point of view, that is not useful. What I need is the supply of inventory that I have against these granular targets. What is the sell-through in the past? What is the optimum price?”
Slade says he and his staff are seeing now what it means to have that kind of data in one place and responsive to queries immediately. “When someone asks on the phone how many of these positions will be available in Paris next Tuesday for chief executives, and I need to know the optimum price, it takes 15 seconds to get an answer out of the system.”
Integrating all of the piles of data now accruing on publisher sites in a way that lets you query across databases in real time changes the workflow, Slade says, and it is especially effective in helping him establish and defend pricing. “I have a client on the phone who wants to renew a sponsorship, and I wanted to increase the price. In ten seconds I can say how many impressions we served across that sponsorship and what the price has been, and the audience she reached, down to their job titles and companies. Then I can justify the price.”
In order for the many data strands to weave together effectively and quickly, a number of technical requirements need to be met, says David Soloff, CEO, Metamarkets, whose analytics engine drives the console Slade uses. The engine works across multiple different databases. Speed is achieved by putting the processes entirely into memory. Also important is “data freshness,” he says. “We don’t batch them up but provide the data sources in real time so you can see what is happening on the platform in real time.” Of course this needs to be done in the cloud to be scalable.
But the most interesting piece of this for a wide range of publishers who are using it now is the dashboard. Via Web interface as well as tablet, Soloff says the customers are learning how to engage the data points in one view so they can explore it more effectively. He says that an analysis of the dashboard use itself suggests that publishers are getting sucked into the data. “Initially, customers did short sessions, but now we are seeing session lengths over 25 minutes.” Once the data they already have is pulled together is put together, they start coming up with new ways of querying the information.
Soloff says that publishers often pay lip service to the importance of analytics, but when the money needs to be invested, it will go to product development before it goes to number crunching. The company is also starting to analyze and flag aberrant behaviors. Anomalous trends in anything from spikes in social network activity at a given location to sudden inventory loads could be exploited if seen in real time.
Now that greater integration across databases is becoming possible, Slade says he is looking forward to cracking some of the more thorny codes in digital marketing, like the actual effectiveness of each ad piece. “We are getting a sense of dynamic pricing,” he says. “Rather than a flat rate card we are putting some real science and real thinking behind the value of an ad impression. In the next 12 months I would love to get us to a greater sense of attribution.”