Solving the attribution conundrum with incrementality

How can I prove the true value of my digital advertising?

This is a question Chris Ball, Business Development Director at Infectious Media, hears often when speaking to marketers.

His answer? Break away from outdated approaches, to measure what actually matters to your business...

IMG_3340.jpg

For too long, performance advertisers have been using a post-click attribution model that overvalues lower funnel tactics such as paid search, affiliate marketing, and retargeting.

But it gets worse.

Post-click attribution hugely undervalues the influence of activity higher up the funnel. No one in this industry should believe that the power of display advertising lies within its ability to generate intentional, human clicks.

But our recent survey of over 200 digital marketers shows is it a persistent approach. 57% of them still use total number of clicks as a key metric to measure programmatic performance.>

Slowly, advertisers have been breaking away from this model, attracted by the post-impression approach. Whilst this has helped to promote the importance of display advertising, it also made it easy to run poor quality campaigns that, on the surface, look extremely effective.

Those who are ill-informed end up buying cheap inventory en masse, and ‘cookie bombing’ users, serving ads to as many people as possible. This results in the vast majority of an advertiser’s audience being served at least one ad, often non-viewable and therefore having zero effect on their behaviour. When someone in the audience converts, it looks like the tactic is working. In addition, this requires extremely low cost (CPM) inventory, with a corresponding drop in the quality of the inventory, resulting in a far higher instance of fraud.

We know that unseen adverts do not impact users in their journey to conversion - which of course sounds like common sense. But knowing this, why would an advertiser be happy to waste budget on such a practice?

 

New Call-to-action

 

So if post-click and post-impression attribution models are fraught with problems, is viewability the solution?

Many advertisers would say yes: by using a post-visible attribution model, advertisers can encourage their media buyers to only buy high quality, viewable inventory that is most likely to influence their audience.

In fact, a number of advertisers and media buyers are taking the position of only buying inventory that is at least 80% viewable. 

But even this isn’t quite as it seems.

Place yourself in the shoes of a talented optimisation manager at a digital agency.

The problem with focussing on viewable inventory is that it’s disproportionately expensive and means Cost Per Thousand (CPM) and Cost Per Acquisition (CPA) targets become impossible  for you to hit. Additionally, for those working for less reputable, less forward thinking digital agencies, focusing your whole programmatic strategy on buying high viewability will encourage you to bid on fraudulent inventory and bot traffic which is designed from the ground up to appear highly viewable.

 

"Why spend money on those users that would have converted regardless of being served an advertisement?"

 

So how should advertisers be using viewability metrics?

Instead of chasing high viewability as a way to define performance, it should always be linked to the end goal: conversions. Then, what you are concerned with is your viewable CPA and your viewable ROI.

Getting to this stage is an extremely commendable position for an advertiser. But it is still not the perfect solution to the attribution conundrum.

This is because each of these attribution models still provide credit for those conversions for which an advertising campaign was not necessary.

So let’s address the elephant in the room of every display advertising campaign.

There will always be a baseline of users who become a customer whether or not they saw your advert. That’s a big statement for a programmatic agency to make but we pride ourselves on straight-talking here at Infectious Media.

Now, for those customers that didn’t seen an advert, it wouldn’t be accurate for the campaign to claim credit for those conversions. Even those that had seen an advert, some of them may have been looking to buy from you anyway, which suggests that the campaign shouldn’t claim credit for that either. And why spend money on those users that would have converted regardless of being served an advertisement?

Just this week (27th March) P&G’s Mark Pritchard has demanded to see genuine proof that his digital investments are resulting in actual sales. But even Mark’s behind the curve. Proving a sale is one thing but proving it happened because an advert changed someone’s mind is another.

 

The real solution is incremental measurement.

An advertiser’s goal should be to create new customers that have been directly affected by an advertising campaign. One that has led to incremental sales, over and above the general baseline.

Once an advertiser has a good idea of the amount of incremental conversions they are earning, they can prove the value of their digital advertising - the holy grail for the digital marketer. And once this has been figured out, advertisers can optimise towards it like any other campaign metric.

That’s not to say working towards incrementality isn’t a challenge. It requires data at a level of detail that can overwhelm even the largest network agency. It also requires specialist analytical technology and skills. You cannot take your eye off the ball for a moment.

But for the digital marketer in this ever-more accountable industry, incrementality has to be worth the challenge. And therein lies the answer that proves the true value of your digital advertising.

To find out more about incremental measurement, read our new whitepaper “Measuring what matters”.

 

New Call-to-action

 

Stay Updated