5 things you need to know about audience measurement in 2021

1. The death of the third-party cookie is great news for advertisers

If you are an advertiser, one of the things you’ve been asking your partners and vendors the past year is probably: “How will the death of the third-party cookie impact my campaign measurement?”

You may or may not have gotten a straight answer because, well, it’s complicated. But in reality, the death of the third-party cookie benefits anyone with a genuine interest in understanding the effects of advertising.

Online measurement has had to deal with increased fragmentation in online consumption habits, ad blockers, browser initiatives like ITP, regulatory changes including GDPR and ePrivacy, limitations to cookie lifetime and so on for years.

Missing and incomplete data have caused methodological obstacles for more than a decade. The fact is that relying solely on “tagging” users exposed to an ad with a third-party cookie and “tracking” them across websites was never a good way to understand campaign effects. Measuring and correctly attributing campaign effects over time and across media, platforms and devices has always been challenging via the third-party cookie.

The industry at large has been relying on an improved source of information for anything from frequency control to conversion attribution, and estimates of people reached. The results have included a ‘wall of ads’ where website visitors are bombarded with the same ad over and over, campaigns where the estimated reach exceeds the size of the entire population and large platforms claiming credit for conversions based on last-click attributions, to name a few of the issues.

Unfortunately, the ad tech industry’s way of coping with the challenges has included more tracking technology like fingerprinting and other cookie-based identifier systems without too much concern for user privacy.

The death of the third-party cookie and privacy regulations forces the industry to develop better and more sustainable solutions for measurement. This might, in turn, provide advertisers with a more truthful picture of the media, platforms, campaign messages and audience strategies that actually drive performance.

2. There never really was a single source of truth

Remember back when data was ‘the new oil’, and we believed everyone and everything was measurable as long as it was digital? Advertisers and buyers relied on campaign and website analytics tools as a ‘single source of truth’ for evaluating and optimising campaign performance and the overall effects of their marketing activities. To a large extent, the general belief was that cookies could tie together 1-1 the consumer exposure to a campaign with the conversion on a website.

Online measurement and traditional analytics solutions were founded on ‘bean counting’ and largely based on a descriptive analytics approach where the underlying data is a count, or aggregate of log-lines to which basic math is applied to describe things like total impressions served, average CPM, periodical change in cost per click, cost per conversion, etc.

A major fundamental weakness of descriptive analytics is its inability to handle missing or incomplete data, which is why descriptive analytics have never been well suited for measuring online performance even though we kept using it.

The same user accessing a media from multiple devices or browsers will always result in fragmentation as every new identifier added to the mix when switching browsers and using multiple devices instantaneously creates “new” users due to the nature of descriptive statistics.

It’s a simplified approach leading to a lot of reporting error which is why many of the most used analytics tools on the market have always been over-reporting the number of unique users and underreporting on frequency.

The fear that Apple’s tracking prevention initiatives or Google’s determination to kill the third-party cookie will negatively impact advertisers ability to measure campaign and brand performance is based on the false assumption that cookies were the key to a universal truth. In reality, a single source of truth never existed.

3. The solution is advanced marketing modelling

While the impact on measurement that relies heavily on third-party cookies and uses descriptive statistics is severe, advanced inferential statistics offers a methodology that is a much better fit to handle incomplete and fragmented data.

Inferential statistics use samples to describe the population, which means there’s no need for perfect data or the ability to track individual behaviour tied to an identifier like the third-party cookie.

Graph technology allows for privacy safe deduplication across devices, media and platforms, including walled gardens where deep integrations are possible, and recent advances in machine learning offer solutions for handling environments with even very sparse data.

In other words, advanced marketing modelling provides a framework for understanding sources of uncertainty, sparse and incomplete datasets while still inferring precise knowledge of the population, which is what is needed to solve the challenges in measurement that the industry faces today.

4. Cross-media audience measurement is more important than ever

As media content is being consumed in an increasingly fragmented fashion across multiple devices like smart TVs, desktops, laptops, tablets and smartphones as well as multiple media platforms, including walled gardens like Facebook, YouTube and others, the need for cross-media audience measurement capabilities has never been greater.

Validating the true reach and frequency of campaigns is of increasing concern to advertisers who need continuous and standardised measurement to determine how and where advertising budgets are spent most effectively.

The WFA North Star initiative is a global example of advertisers’ demand for better cross-media measurement. The ability to measure consumers’ ad exposure across all media and platforms where the consumer could potentially be exposed to the ad, including premium video audiences, is vital when buyers want to understand how campaign budgets are best allocated to achieve maximum effect at the lowest possible price.

5. Walled gardens are opening up

It’s been three years since the Cambridge Analytica data misuse scandal broke, and the walled gardens closed their doors on third-party measurement companies. Since then, advertisers have had a hard time getting third-party validation of campaigns and understanding campaign effects across open and closed platforms.

Although the walls around the ad tech giants’ gardens are still very much there, platforms like Facebook and YouTube have started to open up to select vendors. For example, AudienceProject is currently one of a select few measurement companies integrated with Google Ads Data Hub (YouTube) and one of a few measurement companies delivering audience verification across Facebook, YouTube and the open web.

Deep integrations and ad server integrations to walled gardens are necessary for measurement companies that want to deliver true cross-media audience measurement in a future where more and more advertiser budgets are allocated to the big platforms.

Learn more

Want to learn more about the current challenges with cross-media audience measurement and how AudienceProject is solving these, please download our white paper “Advertisers’ North Star within reach” going into more detail on this.

Or, if you want to start your cross-media audience measurement journey today, please reach out on hello@audienceproject.com.