Data is rapidly becoming our number one preoccupation. At work, we load our spreadsheets with it. At home, we access it through our mobile devices. If we could, we would turn a tap and see data pouring out. Increasingly we're running our lives with it. Indeed, if authors such as Evgeny Morozov are right, then big data could render our governments redundant. While this could be more science fiction than fact, the message is clear: data is a resource, and marketers in particular need to start using it as effectively as possible.

Making the data cake

People often get hung up on the word 'programmatic' but all it really means is 'automated'. Take a look at Google: that's programmatic search, in which algorithms decide which pages are relevant to your search terms. It works effectively, and efficiently, at scale, and it's the same with programmatic trading. It just requires large amounts of quality data and smart algorithms to feed on that data, to work effectively.

First, a description of the types of data used in programmatic mobile advertising. There are four principal sources of data, coming from the publisher, the advertiser, the ad tech, and third parties.

The data that publishers provide is called contextual data, because that is the 'context' in which an ad will appear. This will typically describe the mobile site or app in terms of its vertical (that is, the industry it's in such as gaming, beauty, automotive and so on), and the size of the ad space. The publisher also provides information such as the user’s location and device, but only if the consumer has consented to this.

Contextual data is the first step in determining whether to bid for an ad: for example if it identifies as an iPhone, in India, on the Scrabble game, and there are no campaigns to match this, the process ends. If there are, the other data is brought to bear on our decision as to the value of that ad space.

First party data comes from the advertiser and describes the audiences they want to reach. If there's a strong match between the first-party data and all the other data sets, the ad space is more relevant and therefore more likely to trigger a bid.

Second-party data is where ad tech comes in. Literally billions of programmatic mobile ad transactions are processed every day, and this data can be mined to add insight and make campaigns more effective. It enables profiles to be built about audiences, for example whether they are male or female, their age, their propensity to engage with certain types of ads and so on, and this is where ad tech can really enhance the precision of mobile ad targeting.

Finally, third-party data, as its name suggests, is bought in from external companies such as Lotame which further describes audiences. This can cover anything and everything, from enhanced marketing segmentation data, to geolocation data, even to weather information (useful when you want to sell fizzy drinks on hot days for example).

Feeding the algorithms

When there's a good match between all the data sets involved, through programmatic trading, a bid is made for the ad. It's the job of the algorithms to do this, based on parameters set by the user.

But the key benefit to number-crunching is the potential for finding the right audiences in the first place, that is, targeting. With programmatic trading, no longer is it a case 'spray and pray', sending out many impressions in the hope that some are relevant. Instead, it is now 'focus and find', using precision to zero in on the right audiences for the right ads.

And programmatic – or, to use the specific implementation gathering pace today called Real-time Bidding (RTB) – also holds huge potential for retargeting, that is, deciding whether someone has already been exposed to an ad and taking appropriate action, such as showing a different ad or even not displaying an ad at all.

This is also where the data works hard, bringing together data about platforms (eg Android, iOS), devices (individual brand models) and exchanges to create IDs that work across all channels while rigidly protecting user privacy.

This is particularly important for mobile because third-party cookies are a lot less abundant than in the desktop environment, which desktop advertising has relied on for retargeting. The data-plus-algorithms approach should point the way to real retargeting regardless of platform or channel.

Retargeting was at the heart of the Sony Xperia Z1 campaign. Part of the brief was to establish how effective the mobile advertising campaign had been, punching through the marketing funnel from top-level awareness through to consideration, preference and conversion. This had never been done before with mobile advertising but, by being able to differentiate between exposed and non-exposed audiences, it was demonstrated that spontaneous brand awareness raised from 37% among non-exposed audiences to 60% among exposed and that, among exposed audiences, those very/somewhat likely to buy increased from 49% to 64%.

Biting a chunk out of the advertising pie

As well as understanding how valuable their data is, marketers need to understand that their digital campaigns should recognise what mobile exclusively offers, because it is uniquely a tactile medium that people carry around with them. It's already known that people use their mobile devices in different ways at different times of the day, for example through the IAB Three Device Lives study. And what tells us this? The data.

Given the astonishing rate of innovation in mobile advertising, expect to see much more nuance around this, particularly first-party advertising data. For example, more granular audience profiling could mean audiences can be identified for offers within campaigns, such as showing certain models of car depending on who's viewing the ad, or introducing more sophisticated valuations for certain audiences given their past interactions.

So while a government run through big data and smart algorithms may be science fiction, it's a fact that big data is having a huge impact on our digital marketing today, and will continue to do so in the future.

 

By Victor Malachard, CEO of Byyd. 

PrivSec Conferences will bring together leading speakers and experts from privacy and security to deliver compelling content via solo presentations, panel discussions, debates, roundtables and workshops.
For more information on upcoming events, visit the website.

comments powered by Disqus