Industry figures share their views on the latest issues. If you have an idea for a guest column, email email@example.com
Ad effectiveness: It's time to move on from obsessing over CPMs and focus on the human equation
You get what you pay for, which is why the long-embraced, industry standard CPM is looking more and more obsolete in today’s programmatic universe.
CPM is a “race to the bottom” when it comes to price, and while no one wants to pay over the odds, the old ways of working no longer translate effectively into today’s media currency.
In this day and age the lower the CPM, often the more the inventory is exposed to viewability and fraud issues. While pitches continue to be won and lost over who can buy the cheapest, advertisers should demand metrics that favour quality and performance over volume or price.
Balance is key
Whilst everyone talks about the importance of viewability (and I tend to agree), it is worth considering why it should not always be the highest priority. Viewability represents a chance for an ad to be seen by a person, but the frustrating thing is that fraudulent impressions in many cases can actually appear to perform better depending on the KPIs.
For example, a bot doesn’t care how long a video ad is, it will sit there all day if it means a completed video view. If we optimise out the fraud (in this case the patient bot), on face value the campaign performance (video completion rate or cost per completed view) actually looks worse.
But what is better for a brand? A cheap bot that watched a video for hours but has no wallet? Or a human who wants to make a purchase? It seems obvious, but higher CPMs are often more effective if they eliminate wastage and result in ads that are seen by a greater number of people.
The trick is to balance price with the other metrics in a more meaningful way, driving campaigns to become more effective and not just less expensive. Viewability is one of a number of metrics that fit together to determine real world campaign success. In finding that balance, marketers need to bring back the human.
The value of a human CPM
After all, that’s what it’s all about, isn’t it? Those ads need to reach real eyeballs, tug on real heartstrings and ultimately, open real wallets.
Advertisers and agencies need a new framework that allows them to see the true cost of their media in relation to the good impressions (viewed) and the bad (not seen by anyone). The human CPM (hCPM) marries spend to viewability and fraud data in a more meaningful way to see how much inventory really costs once the bad impressions are removed from the equation.
This really gets to the nub of the issue, understanding if the ads have really landed. At VivaKi we use hCPM to help see past the face value of impressions and optimise towards the inventory that actually has an opportunity to influence the minds of consumers.
In its simplest form, you can calculate hCPM using the following formula:
(CPM/Viewability)/Fraud = hCPM
For example, if you buy 1,000 impressions for £10 but half of them are fraudulent, you have really paid £10 for 500 impressions. If you then imagine that only half of those are viewed, you have actually paid £10 for 250 impressions. In reality, your hCPM in this case is £40.
For brand advertisers, I would argue that hCPM should be one of the top metrics any campaign is optimised and evaluated on - the cost of the ads that were actually seen. It slices through the BS and gets to what really matters, the good, the bad and the downright ugly. It gives us the shiny new brand metric that the industry has been craving, but it’s only part of the answer.
To be clear, hCPM will not solve everything. You could question how much viewability matters for a bank that is solely focused on driving credit card applications, or a telco selling phone contracts. When sales are all that matters, should an advertiser care if they only achieve 30% viewability if those cheaper impressions result in a better cost per sale?
Whilst hCPM might not be the primary metric for every advertiser, it is a simple equation that can be applied manually to any campaign, placement or URL, and in a real-time, programmatic environment it is one that works best when automated and integrated into a wider optimisation toolkit.
At VivaKi, hCPM lives in our proprietary evaluation process – Quality Index (QI) – that vets all inventory sources according to calculated metrics that assess viewability, page content quality, brand safety and historical performance to understand true quality of any impression. Only when all of these are considered are marketers able to make meaningful and accurate optimisations based on all the metrics that matter.
Unfortunately the widely accepted ways of measuring agency effectiveness only compound the problem. Auditors have no idea how to assess agencies except on price, and force them to buy poorer quality inventory in order to hit the numbers they promised to win the pitch. Buying low cost or bulk inventory (especially in a non-transparent way) may on face value seem like cost saving, but if those ads aren’t actually seen by anyone, then their true campaign efficiency is certain to be way off the mark.
hCPM will help advertisers see past the fog of viewability and fraud, but it’s only when they find the balance between their campaign objectives and all of the available KPIs will they really start to understand true impression quality, real world efficiency and ultimately accurate media performance.
Geoff Smith is VP solutions consulting, EMEA, at VivaKi
Have your say
Do you have a strong opinion on a topical industry issue? To submit a comment piece, please send a short summary of your idea to firstname.lastname@example.org. Views of writers are not necessarily those of The Drum.