With brands moving more advertising funds from linear to digital media (Digital ad spending reached $209 billion worldwide — 41% of the market — in 2017, while TV brought in $178 billion — 35% of the market — in 2017) the need for a positive return on advertising spend is increasing and the perils of brand safety are rising. 

A greater understanding of the target audiences, often based on ‘inferred’ demographics, has fuelled this rise of programmatic marketing and advertising. Digital is moving to a GRP (Gross Rating Point) – focused planning, i.e. a buying and optimization models based on age, gender, reach and frequency. 

Among the most sophisticated media buying agencies, these models comprise; rich demographics, knowledge about their technology usage, behavioural insights, purchase history and location data.  However, they are often missing forward-looking insights. These are insights which judge future intentions to buy or the extent of positive disposition towards the brand for example.  

Being able to understand how a campaign has influenced intention through comparing the thoughts of ‘exposed’ and ‘non-exposed’ consumers is an extremely useful additional tool to the planning armoury.

The Missing Piece of the Jigsaw

Until now, reliable audience self-reflection has been difficult.  Cookie technology has been around for some time of course, but getting hold of a large enough group who you can be certain of either seeing or not seeing a campaign has been difficult. People might be able to remember an image when prompted, but usually, it is not the name of the campaign, or brand for example. This is particularly exasperated if the exposure was quick, or partial.

Tracking Codes Meets Online Respondents

Over the past nine months, Sapio Research has been working with a well-known international programmatic company to help strengthen its algorithms and measure its outcomes. It’s been using Cint Connect, whose panellists have agreed to cookie tracking. The system monitors how many impressions the ad has received and how many of these are accounted for by its panellists who can be invited for a deep dive.

Each matched panellist also carries over 150 different demographic points from self-declared data, enabling modelling to really step up a gear if need be.

The real power is the ability to understand how the exposed groups’ views differ to the unexposed (who are selected from a different pool). When comparing the thoughts of a control group or ‘match audience sample’ who haven’t been exposed to the campaign, you can judge what has worked well, what needs to be improved and why.

This insight can help to track content, optimise creatives and ads in real time, prove marketing impact, discover more about the audience and strengthen your media plans.

Therefore, this process truly enables the programmatic industry to win back advertisers trust.

  1. You can reliably measure the impact of a campaign on real people who you can ask to explain their behaviours or future intentions.
  2. You can assess the components of a campaign that drives success, so make more informed decisions, not just on placement and timing, but the content of the creatives too.
  3. Involving independent, external data sources and collaborators using this approach improves outcome believability and integrity.

The programmatic industry has been criticised a lot recently so new tools and techniques can’t afford be ignored. This is a great example of how experts in traditional and modern data analysis techniques are coming together to make things more useful.