In a recent conversation with a chief technology officer, he asserted all digital technology changes in his organisation were being led by IT and not by marketing. It made me wonder: How long a marketing function like this could survive?
Programmatic ad tech vendor, AppNexus, is looking to up the ante around targeting buyers with digital advertising by allowing marketers to bring their own data insights into its real-time bidding platform.
The vendor has launched AppNexus Programmable Bidder (APB), a new technology that allows advertisers and agencies to upload proprietary algorithms directly into its platform in order to better tailor and optimise digital campaigns on the back of their own data.
AppNexus Asia-Pacific vice-president, Dave Osborn, told CMO advertisers have historically had two choices when it comes to employing internal data insights to bid on impressions programmatically and externally. The first is to work with a DSP to port proprietary data into their “black box” technology, without knowing how that data is being used and valued, he said. This could potentially be leaving money on the table, he claimed.
The alternative is to build proprietary ad bidding technology, and wear the expense of sorting through and bidding on more than 30 billion ad impressions per month, he said. Even brands establishing programmatic trading capabilities in-house are limited to the data usage criteria and variables set by third-party vendors, Osborn said.
With APB, marketers don’t have to rely on AppNexus’ bidding criteria, and can instead come up with their own algorithm decision trees, which can then be uploaded using the vendor’s open APIs and accessed via the platform interface. These are used to make decisions around which audience to serve what creative in real-time.
Typical variables used by bidding algorithms for programmatic advertising today include time of day, geography, inventory and segmentation, such as users that have previously purchased products. By opening up the box, marketers using AppNexus’ platform could create any number of variables, such as customers who purchased in-store the day before, who’ve just been on their website, as well as set rules and a series of advertising events around them.
Osborn said the idea for APB stemmed from a number of clients who were finding they couldn’t achieve the level of targeting they wanted programmatically to using the data insights gathered internally. This technology isn’t for everyone, however: To employ APB effectively, organisations are going to need sophisticated data scientists as well as relatively low latency around gaining data insights from Web and digital activities.
“For a lot of companies using the platform they’re happy working with the variables built into the system with subtle tweaks on their own,” Osbon added. “But increasingly, as we get more sophisticated clients and marketers interested in using this technology capability, we’ve got to retain a significant level of flexibility.”
Long-term, AppNexus anticipates a new generation of programmatic media companies could pop up, selling packaged optimisation services and algorithms to marketers for using on the platform.
AppNexus is currently in a closed data trial with select clients for APB, and expects to expand this in coming months before rolling out the open offering by the end of 2015. While he saw broad industry application for this technology, Osborn expected major retailers, telcos, banks and entertainment companies to be early adopters.
In a statement, AppNexus CEO, Brian O'Kelley, said the technical challenges and financial costs of real-time bidding have prevented most buyers from investing in algorithms to date. AppNexus ultimately hopes to see advertisers and agencies evaluating proprietary optimisation formulas against its algorithms and refining and adapting their campaigns accordingly, he added.
“By allowing data scientists to upload complex models directly into our scaled real-time bidding platform, we will finally see the industry take full advantage of the power of algorithms to maximise the impact of every ad impression," he claimed.
Also this week, AppNexus introduced IQ, a suite of major policies, products and features which it claims form the industry's most transparent and reliable marketplace for online inventory. Within the suite are strict domain transparency standards, pre-bid policy enforcement, and AppNexus Spend Protection, a financial guarantee to advertisers.
In a statement, the vendor said alpha testing of IQ produced discernibly higher revenue per 1000 impressions (RPMs). It also claimed publishers participating in IQ's alpha phase saw lifts in RPMs of 29 per cent to 47 per cent, and nearly two-thirds of advertisers have seen increases in response rates.
"Like all markets that join buyers and sellers, the online ad exchange functions best where there is maximum trust and transparency," said O'Kelley.
"Opaque markets are inefficient for all participants. IQ ensures that all inventory available on the AppNexus platform is of the highest quality and visibility. It's very simple: Advertisers should be able to see what they are buying and know what they paid for. It's a principle that benefits buyers and sellers alike, as alpha results have clearly shown."
Follow CMO on Twitter: @CMOAustralia, take part in the CMO Australia conversation on LinkedIn: CMO Australia, join us on Facebook: https://www.facebook.com/CMOAustralia, or check us out on Google+: google.com/+CmoAu