Switching to digital advertising is no guarantee of savings

theaustralian.com.au

Over the past five years, billions of dollars in advertising spend have been moved from news media to digital.

If you push advertisers or their media agencies for the rationale for this switch it does not take long to get to the immovable economic argument of CPM — the cost to reach one thousand consumers.

While news media has maintained its CPM at relatively standard rates over the last few years, digital media offers sensationally better value in comparison. It’s not uncommon for a national newspaper, like The Australian for example, to charge about $10,000 for a half-page ad in their newspaper. Assuming a circulation of 100,000 that translates into a CPM of about $100. In contrast, a digital billboard ad placed at the top of a social media site or popular digital page might offer a CPM of $10 or less.

Rarely can any business sustain itself with a tenfold price premium over its competition. But let’s pause for a moment and move our attention from the C to the M in CPM. Measuring media audiences has never been an exact science, but the recent advances in digital media have lowered that already wobbly bar several notches further down.

Not even the most naive client believes that the audience of 1000 Australians that have bought for $10 or $100 really numbers 1000, but how many will actually see their ad long enough to process it?

Let’s start with the digital billboard ad with its $10 CPM. We have to take out the proportion of the 1000 who are actually bots and not human. Ad fraud is a huge and growing issue in digital advertising and the latest data from the ANA in the US estimates that bots account for 11 per cent of the digital display advertising audience, and that the figure increases by 40 per cent for more valuable ads like our $10 digital ad. So we are down to 850 viewers.

The 850 viewers might be human, but many of them have invested in ad blocking software to ensure their browsing experience is digital free. PageFair estimated that in 2015 18 per cent of Australians were using ad blocking software. They also estimate that ad blocker usage is growing annually by 41 per cent so by now we can assume that a quarter of our 850 did not see the ad thanks to their software, leaving us with 638 viewers.

But just because you can view a digital ad does not mean you will view it. Digital ads have an annoying habit of appearing above, below or to the side of the viewing area that the consumer is exposed to.

Metrics are the German analytics company famous for their viewability assessments of digital media. Alas they do not currently offer any Australian data so let’s use their 2016 British data instead. According to that assessment, 47 per cent of served digital ads are actually viewable. So of our 638 viewers just less than half, about 300, can actually view our billboard ad.

But being viewable and being viewed are two entirely different things. How many of our 300 consumers who could look at our digital ad actually saw it?

Here we must turn to Lumen, the hot new eye-tracing company from London, who have enormous data sets examining what audiences actually look at when they consume media. Their new data shows that digital billboard ads, assuming they are viewable, will be seen by 64 per cent of the audience. So our 300 now becomes 200.

But moving your eyes across a digital ad is not the same as processing an ad. How many of those glances lingered long enough to take in some or all of the information in our digital ad? Let’s set the threshold at two seconds or longer. Lumen also has this data and can conclusively show that if 100 people see a digital ad, 4 per cent of them will look at it for more than two seconds. So our 200 glancers become 8 actual viewers. We paid $10 and got eight people to process our digital ad.

Let’s compare that breakdown with news media. Our $100 investment bought us 1000 readers but how many of them actually saw the print ad? Helpfully Lumen also measures print advertising and the proportion is 75 per cent. So we are down to 750 people already.

But hang on, we bought our 1000 readers based on circulation not readership. Usually more than just the purchaser reads the paper they bought. There are various estimates of readers per copy or RPC that vary from 3X to 6X circulation. Let’s apply a conservative estimate of 2X circulation. That means our 750 ad viewers actually rises to 1500.

Again, however, we have to apply the same two second threshold to this group too. How many of the print ad viewers dwelled beyond two seconds? One might expect this to be a lot better than the fast-moving, click-through world of digital. According to Roy Morgan, the average Australian spends 18 minutes a day reading their newspaper and the ads are part of that experience. That higher involvement is demonstrated in Lumen’s 2016 eye-tracing data which shows that 23 per cent of print ads exceeded the two-second marker point, meaning that 345 people saw our half-page ad.

That’s not a fair comparison with our digital ad of course. We paid 10 times the amount for our half-page print ad and could have therefore bought another nine digital digital ads on top of our existing one.

But even then, based on these figures, a $1000 investment in digital digital ads would have delivered a total of 80 people seeing my ad for two seconds or more. The same investment in print would have delivered nearly four times as many people.

Clearly I should apply healthy caveats to this data. Generalising so broadly, even with reliable third party data, is always an imperfect art. Digital offers many additional advantages in terms of better targeting and much more scalable options.

Similarly, we could point to the signalling power of being seen in a major national newspaper as an additional qualitative advantage of news media. There’s a lot more to this debate than crude audience estimations of audience.

That said, let’s take a healthier and more sceptical look at the M the next time the CPM data is trotted out. One thousand viewers is never one thousand eyeballs, but what it actually is must surely be one of the biggest questions in all of media.

The advertising rates quoted in this column are based on publicly available information.