In defence of Mark Ritson: Digital metrics still have a long way to go
Squabbling over digital metrics is not going to solve the problem, writes Nico Neumann. Instead, let’s focus our energy on improving standards.
After watching the Mumbrella360 debate Are digital metrics bullshit? and observing various Linkedin rants afterwards, I spotted a behavioural trend which I think really does not help our industry.
So what is this concerning trend?
On the one side there is Mark, a business professor, who has not only been teaching at some of the best schools in the world, but also has first-hand experience in consulting top brands. While he likes to make noise about BS, Mark also tends to back up statements with data and references, and only makes general observations.
Mark can be a bit provocative and likes poking the bear, but that’s his brand and let’s be fair: it is entertaining and probably the reason why Professor Ritson appears to get outstanding teaching evaluations from his students. He may call a room full of people ‘idiots’ to make a point, but he does not insult any single person directly on stage because of a different opinion.
On the other hand, we could hear or read rebuttals from various adland executives, often even directly referring to Mark. Some raised good points, such as possible misunderstandings around the term programmatic.
Yet, in addition to rational arguments, we can also observe a tendency to personally attack or ridicule Mark across social media, events or trade press releases.
Flicking insults – is that what our professional life has become? Perhaps another Donald Trump influence? Is our best future path given by making fun of someone who addresses obvious issues?
I would say some of the responses I have seen only show that Mark has hit a soft spot. Sadly, a lot of people make money with BS and smoke and mirrors in advertising, and these people clearly have an interest in maintaining the status quo.
Of course, this applies to both ‘traditional’ (whatever this means) and digital media, but more business folks related to the latter seem to have pushed the envelope in recent years to make a quick buck at the cost of long-term health. That’s why we ended up with a chief brand officer of one of the biggest advertisers publicly criticising the digital marketing business.
With regard to the Mumbrella360 discussion, we have heard some very interesting points, many of which I have encountered in discussions with friends and colleagues too. I think it’s crucial that we try to sort out some of the misunderstandings and confusions in the background of these discussions.
An eye for an eye?
It is fair to say that the measurement methods for TV, print and radio are flawed too and that digital metrics are not the only ones with errors. However, that’s a separate issue from the many problems that we have created for online media, such as the chaos of definitions, standards and useless vanity engagement KPIs.
Moreover, I feel that this ‘eye for an eye’ approach does not help solve the present challenges we face. Let’s tackle the issues of each medium one by one and not use the flaws of the counterpart as an excuse to do poorly too. If we encourage the philosophy of “an eye for an eye”, we will end up in a blind world.
1. Ad fraud numbers
Indeed, ad fraud numbers seem to vary a lot across studies: is it 2%, or is it 98%?
That’s a valid question and we need to consider a few technical details here: are we talking about clicks or impressions, was the campaign run via an exchange or with a premium publisher and in what year and market was the study conducted?
Fortunately, in 2017 we have much less [identifiable] fraud than in 2014 – largely thanks to the trade press and people like Mark who have made marketers aware of this topic.
Nevertheless, one of the greatest issues for the reported percentage ranges is that anyone can publish any number to get headlines, but we have no idea how a research piece was carried out and whether or not the technology is reliable.
That’s the reason we have accreditation services. In fact, there are only very few companies that are accredited by the MRC for sophisticated invalid traffic detection. You wouldn’t use a non-licensed builder or laymen to construct your home either, so let’s consider only appropriate sources of information for ad fraud discussions.
2. Digital viewability refers to opportunities-to-see, not human eye balls
Unfortunately, too many people confound what viewability refers to. The term just means that the ad was actually completely loaded on a computer or mobile screen (for whatever seconds or percentage). Because only if an ad was presented to a sufficient degree on the open browser tab, consumers have an actual opportunity to see the ad. Based on this definition, TV and print ads have 100% viewability. That does not mean a consumer actually saw the ad on TV, as much as we don’t know whether someone actually notices the 100% viewable ad on a cell phone or computer.
In conclusion, it’s nice to get headlines and we all like defending our turf – otherwise known as the organisations that help us pay our bills. However, there is no reason to get personal with anyone or to play down problems of our ecosystem.
Let’s focus our energy on improving measurements and standards instead. Because if there is one thing the Mumbrella360 debate has illustrated, then it is that we all still have a long way to go.
Nico Neumann is a senior research analyst at the University of South Australia.
This topic is getting tired. Can we stop giving Ritson so much attention and just get on with collaborating with each other to produce great marketing campaigns? No media channel is perfect – digital channels included. It’s easy to sit back and pick apart from afar. Get in the trenches Mark and see how you go.
User ID not verified.
100% Agree. While it’s a heated subject, no-one (person or company) has addressed Mark’s criticisms to the same extent he does – with quantifiable data. There’s been plenty of personality mixed with petty shit slinging in this debate, and one such personality has even apologised to Mark as a result.
The Mumbrella360 Digital Metrics debate featured a moderator who offered no moderation – or perhaps they were just up there as a sponsor. If I stood up and was subjected to personal attacks instead of a debate I was there to have, I would have pulled the plug. What surprised me was how willing an audience of Australia’s top media and marketing types were happy to accept and validate these attacks too.
It clearly shows the industry has a maturity and professionalism issue. You wouldn’t see, let alone accept this type of behaviour and opinion in legal, accountancy or other professional services businesses. So why as an industry do we tolerate it, and why has mumbrella assisted promoting such behaviour?
If I were Mark, I’d not offer my services to another Mumbrella debate.
User ID not verified.
TV ads can’t be considered 100% viewable if no one knows if my TV was even on.
The “…very few companies that are accredited by the MRC for sophisticated invalid traffic detection.” includes IAS, Moat, DCM, Microsoft, Sizmek, Flashtalking, Innovid and Nielsen, which are the 3rd party adservers and measurement tools for every agency and publisher group in Australia, and therefor nearly all Australian digital advertisers.
I’m with @Overit. This is boring now.
User ID not verified.
your tv comment has nothing to do with the digital viewability as it is measured online. all tv ads are 100% viewable unless your telly is really broken.
how about you check the link before spreading wrong information: only integral ad science, adloox, double verify, whiteops and pixelate are mrc accredited for sophisticated fraud detection for all traffic (mobile and desktop)
User ID not verified.
How is a TV ad 100% viewable if my TV is on and I’m not in the room?
User ID not verified.
I love digital analytic’s because they are factual. My GPS is Android and in real time finds the shortest route, via Google Maps. My contributions to Google Maps of places are not only rewarded, however a view count is available for me to see, so I know how engaging my reviews are. Google Trends can be utilised for a plethora of reasons and things. Here it quite correctly shows a slight increase in the ‘News’ category for ‘Mark Ritson’. Ho hum.
User ID not verified.
I forgot the link: Voila! https://trends.google.com/trends/explore?cat=16&geo=AU&q=Mark%20Ritson
User ID not verified.
have you even read the section in the text on viewability and ots = opportunity to see? viewability has nothing to do with human beings seeing the ad. you need an eye tracker to get that data. you can have a 100% viewable desktop ad and havent seen it either (if you are not in the room)… same for tv of course
User ID not verified.
Jay, so if they are all accredited why do they produce different results.
User ID not verified.
But how does Google measure the whole market and/r competitors. Is it a perfect measure of an imperfect subset?
Digital is good at the micro and crap at the macro.
User ID not verified.
As someone who has suffered through, and presented in, 100’s of post campaign presentations in my 7 year media career, here is how they go 100% of the time:
Client: Did our TV ad run?
Agency: Yes
Client: Did it reach our reach goals?
Agency: Apparently yes.
Client: Moving on. What of Out of Home?
Agency: Here are photos that proves your posters were put where these photos where taken at the time the photos were taken.
Client: Moving on. DIGITAL! What of video completion? How many users landed on our product page? How efficiently did we drive traffic? How many clicks resulted in a sign up? Which publishers drew more engagement? Where in the purchase funnel did we see the most drop off? Which creative variation was our best performer? Which target audience didn’t perform? What was our viewability?
Agency: Here are answers to literally every question plus recommendations moving forward.
Point is, as Ringrose stated, we do have quantifiable data and we’re always learning from it. Stop trying to defend TV – TV is just TV until we can adserve it. I’m more than happy to credit Ritson’s alarm-ism for the standardization of IAS and Moat into Australian media buying. But now that they’re here and keeping Facebook honest to boot, I think it’s time Ritson and all of us moved on to better things.
User ID not verified.
So TV metrics are bullshit too then, right?
User ID not verified.
“Nevertheless, one of the greatest issues for the reported percentage ranges is that anyone can publish any number to get headlines, but we have no idea how a research piece was carried out and whether or not the technology is reliable”. So true with a lot of agencies these days, lol!
User ID not verified.
The current articles and previous ones point out the variability of ‘view count’ and whether it means people are actually reading your blog posts vs. some search bots or people that load the page then close it because they were expecting something else…
Don’t just ‘accept’ it as fact. Test it.
User ID not verified.
Great article, well done.
We have noticed a big trend in Agencies trying to defend loose, poorly defined and untrustworthy figures to try and justify the worth of their work.
Put simply, if you want to keep your agency in the job – don’t go the lazy way and open a Google Analytics or even worse… Social analytics page. Get out of the office and start talking to those customers and prove your message got through to them.
My message could very easily have reached them, but been instantly forgotten at the consideration stage because it didn’t stack up relative to my competitors ad they just saw… hence invalidating all those useless ‘views’.
We often engage the services of market researchers, pick up the phone, email and talk to people over the phone or live-chat to customers and simply ask them one of the following:
1. “How do you hear about us, what communication motivated you to purchase our products?” (remember this is recalled memory, not actual impact – so you’ll need to use other tests to validate in research)
2. Setup your sales channels with tracking links, unique offers to the channel that the customer must quote, require a special link or phone number to claim. etc.. (Setting up a special 1300 number for a campaign costs almost nothing).
3. If you are trying to address brand perceptions or other intangible emotions etc.. rather than drive a direct purchase at a point in time. Then figure out that sample size and get contacting people that could be reached in the channel area. (But isn’t that hard? Yep. So start practising and develop some repeatable processes)
*Prove it, or lose it*
User ID not verified.
LOL. Nothing to see here. Move along. Move along.
User ID not verified.
This ?
User ID not verified.
The article does not seem about defending TV mate. Actually it raises TV issues too.
“we do have quantifiable data and we’re always learning from it”
How naive are you? have you ever worked with the data yourself or read the trade press recently?
“How many users came to the landing page”
You mean cookies, not users…. or better, bots?
Video completion?
Sure, if people could only define the real numbers, see Facebook errors
Clicks and sign ups?
Sorry, is this 2007? Please please measure the impact of your media channels on sales and low funnel metrics. Any marketer still requesting click through rates should be fined a penalty or send to prison
And finally, let me tell you: agencies love promoting ‘digital’ because they have the highest margins on it and can control the numbers as clients let all the data sit with then…
Take control of your own data and take things inhouse. It will be an eye opener
User ID not verified.
We’re an agency that manages client’s own CRM data. My whole job is managing data. We weaned client off CTRs but there’s value in managing funnel metrics. We identify bots (it’s about .05-1% of our 2017 data so far). We use IAS to report on Facebook actuals, not Facebook. I do like the trade press, but they tend to be a step or two behind the actual practitioners.
User ID not verified.
Pleased to see this piece Nico. Very timely.
It’s important to recognise that Mark Ritson takes the professorial approach to all of his work and in each of the presentations I’ve seen he uses empirical research to support his views. It’s also however important to recognise that advertising effectiveness research in “old media” was rarely truly robust, with industry approaches dominated by various iterations of pre and post awareness. And don’t get me started on “readership” audits that decided the fate of gazillions of dollars of spending in magazines based on the question “which of these have you seen or read in the last six months”….
The biggest issue for the advertising industry remains being able to show clients (especially the Finance Director) that money invested in marketing drove a positive return on investment. If it helped build brand reputation, extensions, find new purchases, all brilliant.
But it’s really not a competition between “digital” and “old”. It’s about spending your clients money well in a way that helps them grow their business. And being able to measure it.
Let’s all move on.
User ID not verified.
Nice article! Let’s hope for better metrics for every marketing channel
User ID not verified.
still confused about the ad fraud numbers. even the numbers of accredited vendors seem to be all over the place
User ID not verified.
I don’t understand why we place such high stock in the accreditations of a small company in the US.
Maybe a good place to start is an localised body that understands local market dynamics providing this service.
The MRC are really only saying yes, at the time we looked, it did what it says on the tin.
User ID not verified.