The differences between the digital and pre-digital era of advertising aren’t as big as many like to think.
Underlying any marketing campaign, there is still (as there was in the past) a big idea that unifies communication across every channel, a distribution plan, an audience strategy, a measurement strategy, and a set of objectives.
Digital technology has mostly enhanced our capacity to do things that we have always done in this industry, such as targeting communication to the right audience, and measuring the impact of a campaign. But there is at least one fundamental change in how modern marketing operates, or at least should operate. With the growth of biddable media, the division between strategy and execution has become virtually meaningless.
Biddable media is distinct from traditional media because of the immediacy and depth of the way data is gathered – and can be acted on. Real time feedback enables – and in my opinion necessitates – a high level of adaptability in marketing strategy. I’d like to focus on just two aspects of strategy – creative and audience – in this post, to highlight the differences between the biddable and pre-biddable eras of advertising.
The changed relationship between strategy and execution
In the pre-biddable era, there were sensible divisions between strategy and execution, owing largely to the absence of live feedback on performance.
Let’s take creative content as an example. In the past, an advertiser would likely collaborate with a creative agency to develop a big idea, which the creative agency would develop into a set of communications. It would then pass to the media agency to execute the content by purchasing inventory. It wouldn’t make much sense for a media agency at that time to drive the conversation around the content, beyond the pragmatics of different media inventory (e.g. the dimensions for a newspaper ad etc).
Within a channel like paid search, however, creative strategy needs to be adapted continually during execution. Every second, an advert is acquiring new and useful data based on audience feedback. Different messages, and different variations of the same message, can be tested extensively during the course of a search campaign. So it would be extremely counter-productive to pre-determine this aspect of strategy; making it adaptive to feedback gained during execution is the way to win in the biddable era.
For Econsultancy subscribers
Search ads allow you to test multiple variants of the same ad against each other. Beyond what’s built-in, it’s possible to use scripts to analyse your account data, such as this script we created at Brainlabs (open source, requires registration) which measures search term performance and provides an evidence base for choosing between synonyms or similar phrases. Human creativity will always be valuable (just look at Dynamic Search Ads as an example of over-reliance on machine learning) but it’s even better when guided by data.
For a number of our programmatic clients, insights about different creatives from real-time data always lead to adaptations in creative content, bidding and audience segmentation. Working in financial services, for example, we found that customers were more likely to search for a loan during the first two weekdays of the month, leading to a highly effective change to bidding strategy. For another client, we realised that personalising ads for commuters drove a much higher performance, leading to a re-design of ads related to contactless card payments on the London Underground.
Customer and market research has been a prominent part of the ad industry since at least the second half of the 20th century, and it continues to provide an extremely valuable source of insight to brands. Way before the internet, advertisers have been gathering data on their customers in an attempt to understand them better.
Before biddable media’s take off, though, the insights gained from executing media were less extensive, and not always fed to the team handling the execution in real time.
How things have changed with the growth of programmatic, search and social advertising. Platforms within these channels are generally excellent in providing a detailed breakdown of performance. Organisations are able to upload 1st party data to inform targeting, as well as using the reams of data acquired during a campaign, and even applying second and third party data to enable greater and greater relevance.
In executing any biddable media campaign, there is thus a continuous process of refining audience strategy: testing which audience segments work, or which types of content work best for which types of audience. It’s actually a means of learning new things about your audience, as well as developing existing insights.
Execution on Facebook, for example, can lead to changes in who you target, and what you target them with. For the majority of clients we’ve worked with on Facebook campaigns, there are lessons to be drawn from testing content on established clients, which can then be used to inform prospecting activity.
Facebook is remarkable for the granularity of its audience targeting. This means that a prospecting audience can be gradually refined over time, allowing you to experiment with different combinations of audience segments and compare performance. It also works the other way: you can learn new things about your audience through Facebook, potentially informing strategy in other channels.
The future of media
Search, social and programmatic are forerunners; the rest of media will soon follow suit. It’s already happening, in fact. TV, OOH, radio, audio – all of these are, to some extent, available programmatically, which means their performance can be measured, and their audience can be precisely targeted.
Although ‘data and measurement’ is part of ‘execution’ in Econsultancy’s Modern Marketing Model (M3), I actually consider it to be the link between execution and strategy that closes the feedback loop. Any model that omits the impact of data just doesn’t apply to contemporary marketing.
I think this change is a major factor in the biddable media in-housing trend that we’re seeing. A major part of this is the desire and necessity for advertisers to own their own data – something which is not the default position in most client-agency relationships. Legally owning performance data is one part of it: agencies also need to be experts in using that data to adapt strategy, or else they should relinquish control.
As more media becomes biddable, the importance of execution-as-part-of-strategy is going to increase. Advertisers are well advised to own their own data and start getting good at biddable!