Beyond insight

At KINESSO, we’re experts in curating and mining data to produce audience insight. Recently, I wrote about the power of the deep datasets we manage – how, backed by Acxiom, we access unrivalled intelligence about consumers. I wrote about the combinatory approach we’ve developed, blending our rich ‘person-based’ assets with a ‘geo-based’ layer to deliver clients the best from both datasets. You can read more here.

We understand that insight alone isn’t enough. To get maximum utility out of data, we need to do more. We need to compliantly mobilise the audiences at our disposal beyond the constrained analytics environments in which they reside – to release them to be effective in other spaces.

Why connect?

We may draw a lesson on the importance of connectivity from that most paradigmatic of electrical-age technologies, the telephone.

Although generally attributed to the Scot Alexander Graham Bell, who invented the telephone is a matter subject to dispute. But one thing is certain: when talking of phones and their value in the world today, we cannot consider them in isolation of the concept of the telephone exchange and switchboard – technologies that were first conceived by a Hungarian engineer named Tivadar Puskás.

Before switchboards, pairs of telephones were connected directly with each other – restricting their usage to the equivalent of a very basic telecom device. Only when the telephone exchange and switchboard came along could lots of devices be joined up, and calls made across a network.

Over time, of course, the telephone network grew and grew from US to worldwide coverage. And the infrastructure that it produced in turn paved the way for the most paradigmatic of twenty-first century technologies, the internet-enabled smartphone.

The simple lesson from all of this is that the telephone – and the informational content for which that specific device is a node – only became truly useful once there was a system of connectivity in place. Information alone isn’t enough; you need a means of distributing it. Without this, what insight you possess remains locked in a closed system, unable to realise its full value.

Our Combinatory Approach in Action

The same is true of audience data.

With this idea in mind, let us return to our example from last time. We considered an insurance provider looking to engage in-market prospects that are defined by criteria outside traditional demographic indicators. We demonstrated the power and sensitivity of an audience design approach based on the smart combination of ‘person-based’ and ‘geo-based’ data.

The next step is to mobilise that combinatory audience across media platforms.

Activating Combinatory Audiences Across Media Platforms

To power this endeavour at KINESSO, we are lucky to be able to draw on the peerless expertise and assets of our sister company, Acxiom. Acxiom provide identity services to support clients in achieving their first-party data ambitions – unifying, enriching, enabling. Acxiom also power our media activation capabilities. At the heart of the Acxiom business is an identity spine that enables us to connect with over 2.6 billion addressable people worldwide across over 500 media platforms.

Broadly speaking, we have two ways of connecting audiences into the media ecosystem for activation.

The first is to send individual IDs for direct matching with person-based platform IDs – or for translation to other digital identifiers (if the destination platform can only handle certain types of records, such as device IDs or specific ID solutions). This route is a mainstay of our activities.

The second way is to distribute audiences as granular geos, such as postcode sectors or latitude-longitude combinations. Some advantages of this second route include utility for localised targeting strategies (such as store locations); minimisation of the privacy governance burden (aggregated geos do not count as personal data under the GDPR); and an ability to more tightly align your activity with offline media options (such as direct mail, door drops, and OOH formats).

In the case of our insurance provider client’s campaign, distributing audience data to media platform in geo form fits neatly with the strategy we devised earlier – that of targeting proximate people who form micro-communities based on shared propensities associated with affluence and policy renewal.

The audience data we compliantly connect into the media ecosystem allows us to power our advertising buys with the same custom targets in every media platform from global tech players to local partners.

Optimising Delivery with Intelligence and Measurement

In the process of mobilisation, we also use our data to add layers of campaign intelligence. To deliver a truly effective and efficient campaign for our insurance provider, we also need to ask: Which media should we be prioritising where? And for each media platform, in each geo micro-region, how should we weight our delivery? We apply an AI Decisioning Engine to provide these answers, and the outputs we use to inform the buy, and to shape the DCO strategies we overlay. Measurement thereafter takes the form of media, brand, and sales uplift analysis. We use lookalike algorithms to identify control segments, which can be created at both person- and geo-levels of aggregation.

Pic for Person+Geo_Article2_01-05-25

All this amounts to an end-to-end approach for designing and mobilising combinatory audiences, which we use to powers effective advertising strategies for our clients.

And crucially, it all hinges on connectivity.

Recall and compare the early days of telephony we touched on earlier.

The first electric telephone call was a single sentence, delivered by Alexander Graham Bell on a closed line between two adjacent rooms. ‘Mr Watson, come here – I want to see you.’ From these humble beginnings, the advertising industry has built a connected capability that reaches into the lives of consumers. The advertising messages we share with audiences today are more sophisticated and engaging in their content, and more wide-ranging and targeted in their distribution, than ever before.