Many marketers are still unclear on what they should be expecting from the data they are buying, says Audience Project UK commercial director, Martyn Bentley.
Ahead of his appearance at The Drum Programmatic Punch on 3 December, Bentley speaks to The Drum about how the application of data can make media buying more complex, why it’s essential for marketers to understand what is in the data they are buying and what the most common mistakes are when working with data.
Transparency is the word of 2018, but do you think that the industry is focusing far too much on media buying and less on the data being applied?
“Follow the money” is a well-used saying because it’s often true. There is money to be made on media buying and it’s how media buyers have been compensated for years, although there is some gradual shift in the model now. Agencies benefit/benefitted from buying more media (if they are/were on a commission), brands are interested in the effectiveness of their media, and procurement teams are laser-focused on the price of the media.
Now, there is money to be made on data of course, but the application of data can also make media buying more complex and disrupt established remuneration models - e.g. apply a data layer, buy less media, and earn a lower commission.
The way various players earn money on the work they are doing is often misaligned - data vendors, funnily enough, want to sell as much data as possible. Are they always focused on the outcomes or just the revenue? Probably a mix, to be fair.
Publishers benefit from selling media and generating exclusive data, and also it could be argued that some publishers benefit from not revealing too much about their audiences with their data. And for sure, some publishers have also been victims of their data being re-sold without fair recompense. Also, publishers can lose out on selling media at scale when data is applied by buyers that unnecessarily restricts media volume - the well-known issue of PMP buys of quality inventory packages being reduced to a few hundred impressions my multiple data layers.
There're conflicting models at the moment, so the answer is complicated and, what’s more, there are now mega-publishers with huge scale and pretty formidable data, making it quite irresistible to resource-strapped marketers and agencies to put a lot of their spend there, rather than invest fairly heavily in data-science and people to find the perfect time, place and media across the whole ecosystem. Why would agencies do that if they might not get rewarded for it by procurement departments, who just want cheap media?
On balance, I’d say the true value of data to drive effectiveness for marketers and fair price for publishers hasn't fully developed yet, so media remains king for now, albeit with the data usurper waiting in the wings. However, if you have media volume, quality environment and first party data, you could be media-king for a while still.
Are marketers still unclear on what they should be expecting from the data they are buying?
I think many are. For example, you can be clever about how to use data online but if it isn’t adding incremental reach to the offline marketing activity or isn’t driving tangible conversion activity or brand/consideration lift, what’s the point? It’s probably just making money for a middleman or vendor.
I think it’s essential for marketers to understand what is in the data they are buying. E.g. do all buyers and marketers know that a car intender segment is never comprised only of definite in-market car intenders - it is an inferred intent or a segment that has been modelled on some inputs. To work out what value and strategy to assign to a set of data, one must understand what inputs have gone into the segment - what are the deterministic signals (if any), how much has that been diluted with modelled IDs, and how should it best be deployed?
Should brands be taking their technologies that touch their most precious strategic asset - their own data - in-house?
That totally depends on what else their marketing mix consists of and how much expertise they have in-house.
If an expert vendor or agency is enabled to get close to the heart of the client business and will be rewarded on the correct KPIs that drive the client’s business, then that’s potentially going to be more cost-effective than a brand having to build and learn everything itself, pretty much from scratch.
Advertisers, agencies and media companies invest heavily in tech and data. But what are the most common mistakes when working with data?
We think that the most common mistake is not asking where the data comes from and how it is made (as above). Or understanding the value of transparent thinking.
Secondly, 'silver bullet’ thinking can be dangerous and that may come down to a lack of knowledge and confidence. And of course, the ad tech/data community is as we know, an excellent marketer of its own products, and buyers and sellers have definitely been swept up in the rising (and persuasive) tide of 'this solves everything’ selling! Hopefully, now that we are entering an era of transparency, better questions are starting to be asked of vendors, so that the cream will eventually rise to the top.
Constant education and aligning the business objectives of your marketing spend (nearly always to sell more products at a higher margin), with the reward system for your employees and vendors/partners are critical. And don't forget that data can also be about learning, not just targeting. Digital marketing and consumer activity generate masses of data, all of which can contain highly valuable insight to be capitalised on if it can be gathered and interrogated meaningfully.