By: Aidan Mark – Global Director, Performance Strategy at CvE
6 min read
With enough data and a lack of understanding from your audience, it is possible to create almost any marketing effectiveness narrative that you care to spin.
One of my favourite marketing quotes is not a particularly famous one. The quote comes from the brilliantly insightful Tom Fishburne, creator of The Marketoonist. The quote appears within the subtext of a cartoon he created that pokes fun at marketing fads being presented in the boardroom, with promises around marketing’s ability to drive incremental growth:
‘Marketing is full of myths, anecdotes and long-held beliefs about what drives growth’
The quote resonates with me and reaffirms by lived experience having spent over 20 years in performance and marketing effectiveness roles. The accepted truths of what drives growth differ enormously from one business to another.
I believe this situation is born out of multiple forms of fragmentation. Fragmentation exists within the media people consume. Most estimates suggest that the average person is exposed to around 10,000 commercial messages each day, across a variety of digital and offline touchpoints. It goes without saying that digital media is primarily responsible for growth in that number in recent decades.
Fragmentation also exists within businesses. Where once there was a single marketing team, these days businesses tend to have different and often competing specialist departments. Separate brand and performance teams are common, and within the new digital team it is not unusual for the largest brands to have teams representing each performance discipline; dedicated PPC, SEO, social, CRM and programmatic teams are common practice amongst larger advertisers.
These setups inevitably lead to internal politics. The most egregious examples occur when commercial pressure is placed upon the collective team; Post campaign analysis needs to be written and agreed, budgets need to be allocated, headcount needs to be reduced. These pressures force teams to compete for limited resources, with decisions usually made based on leadership perceptions around which teams are using the limited budget most effectively.
This culture does not encourage honest reporting. When political capital is there to be won, potentially with jobs on the line, it is simply not in the interests of any team to report on the true effectiveness of their own marketing. A much safer political play is to inflate your numbers beyond reality, in order to make their efforts look good in the eyes of budget holders.
Unfortunately, management is often not knowledgeable enough to interpret all of nuances within the vast new data sets being presented. Budget stakeholders tend not to understand the intricacies and assumptions made when the team arrived at the ROI figures that are being presented. The result is that the teams reporting the highest ROI tend to secure the budget.
The old quote about there being ‘lies, damned lies and statistics’ is as true in marketing as it is in other walks of life. With enough data and a lack of understanding from your audience, it is possible to create almost any marketing effectiveness narrative that you care to spin. And this is exactly what is going on at scale, as powered by the most data-rich organisations to ever exist – Google and Facebook.
Google and Facebook data is widely used by and reported on by marketers. In 2020 Google drove $146bn in advertising revenue, whilst Facebook’s 2020 revenue was more modest at $25bn. How do advertisers assess the impact of this eye wateringly large spend? Of course, measurement varies from business to business, but the ‘free’ and ubiquitous Google Analytics and Facebook Attribution are undoubtedly the main data points being used to mark the digital marketing teams’ homework.
If something is free, you are the product
Google and Facebook are smart in how they present their free attribution data. In a Google Analytics report, every sale is attributable to some form of digital touchpoint. In Facebook attribution, the default settings dictate that if Facebook have shown someone an ad within a 24-hour window prior an online sale, the sale is attributable to Facebook’s ads. If Facebook decided to target everyone in the UK today with an ad for your brand, they would claim all of your digital sales tomorrow as being attributable to Facebook media. Clearly, such a claim would be ridiculous.
Google & Facebook are experts at dressing up attribution in a way that budget stakeholders believe, often positioned as ‘best practice’ for their platform, but ultimately over claiming the influence and effectiveness of their media.
In a Google Analytics view of the world, the most effective marketing channel is paid search. In Facebook attribution, Facebook ads are king. Yet both versions of the truth are often accepted, and marketers continue to work with hugely inflated attribution numbers relating to digital media.
Mark Ritson alludes to this problem in a recent article where he talks about the need for marketers to be comfortable with imprecision. He tasks his MBA students with solving a problem with numbers that deliberately do not add up. This reflects the realities of marketers working with Google and Facebook data – often the sales reported from the two walled gardens add up to something greater than the real business total. This is because double counting and over-claiming of success is afoot. Ritson rightly states that marketers need to be comfortable with imprecision. This pulls against the forces from Google and Facebook who encourage us to use precise figures; precise figures, that are ultimately inaccurate.
Breaking down the silos
There is unfortunately no quick fix to these challenges, but there are some basic steps that marketers can take to help remedy the measurement challenges created by fragmentation.
Perhaps the simplest step a brand can take is to introduce some independent adjudication. An independent view of attribution can be delivered internally or externally, and there is merit in both approaches. The crucial aspect is that the adjudicator should only ever have KPI’s against total business outcomes, rather than the perceived contribution from a single marketing channel. It sounds obvious, but it is vital that brands align stakeholder incentives with their wider business objectives, a practice that is astonishingly rare.
Naturally that adjudicator needs to be well versed in the nuances in how each marketing channel reports ROI, as well as how the marketing activities tie together into a wider marketing strategy. By way of an example, advertising with a brand building objective will always be under valued when assessed through the lens of sales metrics.
Given that experts such as this are rare and therefore hard to find, a more cost-effective solution may involve hiring an expert consultant. The added benefit of this option is that you can structure incentives in an appropriate and unbiased way; it is easier to be neutral and objective with ROI when you are not on the monthly payroll. True independence is hard to achieve with in-house personnel.
To tackle the issue of media fragmentation, it is vital that brands have a measure of success that sits above the tech giants. Having this higher-level measure combats the problem of double counting, as only the true number of sales are available to be allocated to each marketing activity. In addition, if you have large line items in your media plan that are currently getting zero sales credit, that should be seen as a red flag, and highlights that there are likely issues of accurate attribution.
And finally, you may also want to run your own experiments on Google and Facebook media, free of the biases created by their free measurement tools. When you understand the ratio of Google/Facebook reported sales to the actual sales you measure from an experiment, you can start to use attribution data from the tech giants with more confidence.
In taking any of these steps, you would be taking a huge leap of progress in moving your organisation towards an objective and unbiased measure of success. At Control v Exposed, we believe such an approach will mark a new era of data driven marketing.
