The latest figures from the IAB Adspend report show the impact of the rise of programmatic trading on display advertising - digital channels are up 17.5% to a record six-month high of over £3 billion for the first half of 2013 with digital display alone growing 23%.
This is largely due to the widespread adoption of programmatic trading.
Most of this growth has been generated by focusing on optimising workflows, lowering CPMs and harvesting the low-hanging fruits of RTB such as retargeting (“Data 101”). These short-term benefits have now been reaped and advertisers are ready to move on to more advanced strategies to deliver on the promise of programmatic trading - reaching the right audience at the right time with the right message.
There are three main steps advertisers and their agencies need to follow to go down this path:
1. Know your audiences
The days of the “women aged 25-49” target group are over. Not because it is not a relevant target audience for an advertiser - after all, the P&G's and Unilever's of the world have made a decent leaving out of it - but because marketers can do so much better now.
Consumers interact with the web leaving data-rich digital footprints, revealing the pages they visit on a website, campaign engagement, subscriptions, cross-device behaviour and so on. Mining these huge amounts of data gives marketers a unique opportunity to understand who their audiences are, how they interact with their brands, and how best to reach them to achieve a particular goal. This could be a direct response goal like a subscription or a sale, or a longer-term one such as product or brand awareness, preference or affinity.
Because these vast amounts of user-centric data are available, audience definition should always start with first party information, that is what the advertiser knows about its audiences. This can be the behaviour of their website visitors, CRM information, login data, purchase data and so on.
For example, one of the brand advertisers Weborama works with had defined 11 CRM segments triggering various direct marketing strategies (mail, email) and wanted to apply the same intelligence to its online display (remarketing) strategy. To do so, Weborama implemented an advanced segmentation of their on-site visitors in order to identify CRM-like behaviour (first visitor looking for prices, repeat visitors intending to book, etc.) and to address targeting (bidding strategy, message, frequency, time). By replicating its CRM re-engagement strategy online, the client was able to enhance its return on adspend.
2. Make these audiences available at scale
In the world of CRM, marketers are limited by the size of their customer base. In media, all internet users can be scored using custom models to determine their propensity to respond favourably to the advertiser’s message.
These scores should therefore determine the level of interest of the marketer for this particular point of contact, the price he or she is willing to pay to acquire this contact, and the best message to deliver to this user. The main question is: how do you create these scores so that they are available at scale?
There are two answers to this question: the first one is to use a large enough calculation base, such as a third party data source that will provide the raw material for the scoring model; the second it to propagate the scores widely enough enable this model to operate on every ad request.
3. Optimise continuously
One of the major differences between media planning and audience planning is the frequency of optimisation. Once a media-plan has been designed, it will most likely be stable – at least until the end of the first phase of the campaign.
An audience-planning strategy can, and should, be optimised on an on-going basis. Because marketers can build a feedback loop based on real-time campaign performance, they are able to constantly refine their audience segmentation and targeting model.
One of the challenges for marketers and agencies is to maintain an “open mind” and to allow for new insights to come and enrich the model and potentially modify the targeting strategy. The best way to do this is to keep a non-targeted, “learning” budget line that will pick up new trends and behaviour and will enrich the understanding of audience characteristics.
Brands can also go increasingly more granular to break successful strategies down further. For example, if a campaign targeted to “18-34 males interested in politics and sport” is successful, it should be broken down into several different line items: “18-24” and “25-34”, or “football” and “other sports”, in order to provide more granularity and more room for optimisation.
Audience targeting has already fundamentally changed the way marketers approach online media, and will very soon change the way all media investments are planned and executed as more and more channels become digital.
To be successful in this new media landscape, marketers must adapt to their environment and learn to leverage the information they already have on their audience, to enrich and extend it with 3rd party data sources, and to challenge and refine their audience targeting strategies on an on-going basis. For the smart marketers who will be able to adopt this approach, the benefits are immense.