Are publishers using outdated metrics? How should they be innovating and reinventing their business models?
Understandably there has been much debate of late around publishing business models. The rise of the internet, compounded by the global economic woes, are making it increasingly hard to see where the money is in publishing and media going forwards.
You can take your choice of predictions as to what will happen with online advertising spend this year. The IPA's latest Bellwether Survey shows 'record falls' in ad spend where "even internet advertising suffered a record reduction in spend in Q4" of 7%. The likes of NMA or Simon Waldman, of Guardian Media Group, on the other hand are talking about 10% or so growth in online display spend this year.
Metrics, as units, or levers, for callibtrating value and performance, are clearly important to any discussion of business models. Unsurprisingly, there has been a corresponding increase in recent agonisation over metrics.
Should we make the most of the array of new metrics that interactive media afford us? Or do these just confuse the picture for advertisers and media buyers and instead we should make our 'new' metrics the same as the 'traditional' ones? Are our new metrics stillborn - do Ajax, Flash, Air and the like really herald the 'death of the pageview'? Indeed, do RSS, APIs, data feeds, widgets, desktop and mobile apps herald the demise of the site visit and the unique user?
Data, data everywhere...
The metrics challenge has been furthered mired by the repeated, but doomed, attempts thus far to establish a 'single currency' against which online media can be planned and bought. Witness the recent demise of JICIMS - now reincarnated as UKOM, the UK Online Measurement Company. There are plenty who argue that it is futile in any case even to try and establish such a currency.
There are two obvious high level metrics which should be relatively easy to understand, and buy, for an advertiser. If we take correct audience targeting, including any whizzy behavioural targeting et al, for granted for a moment then it is surely about:
- How much advertising am I buying? This is media "input" measured in terms of circulation, unique users etc.
- What outcomes am I getting? This is what happens as a result of the advertising e.g. clicks, sales, measurable shifts in brand metrics etc.
Ultimately, it must surely only be about the latter of these two? What point is there buying 'input' if you have no outcomes in mind? Yet I fear the system still seems overly focused on the 'input' part of the equation. There is still too much talk of how much advertising stuff is available - be that media owners with reams of complex metrics and data that they're struggling to know how to monetise, or advertiers who I think risk over-obsessing about 'reach' and chasing big numbers when those numbers, on closer inspection might not be quite as big as they at first appear.
Too little attention is given to measuring the outcomes. Specifically, digital media and digital publishing offer greater opportunity to track and measure outcomes that are not so readily available in 'traditional' media.
I would argue that central to the success of the cost-per-click model made so wildly successful by Google, or indeed the entire premise of the affiliate model, is that outcomes are being sold, not inputs. The outcome may only be a click but it is a measurable outcome nonetheless. In these recessionary times it is little wonder that there is increased focus on ROI and buying outcomes, be those in the form of lead generation, clicks, coupons, registrations or sales.
A closer look at media owners' traffic figures
Let's take a bit more of a look at the kind of "input" you can buy from publishers online. Specifically, the unique user "who's got the biggest number?" battle that is played out each month when ABCe publishes its figures. This seems to me to be mostly useful in the entertainment value it generates as senior figures scrabble to belittle or big-up their respective numbers.
Let's take the biggest of the big number achievers - Guardian.co.uk. Not to pick on them, indeed I'm quite a fan of The Guardian, but to pore through the big numbers a bit further in the same way it would apply to any other publisher.
In December 2008, according to ABCe figures, Guardian.co.uk got 22.8 million unique users (+43% on the previous year). According to data from the : Newspaper Marketing Agency's new online tracking service the total number of UK consumers who visited a newspaper website in December was 22.6 million. Can it be that the Guardian got more users than there are, er, users?
No, because, of course, a fair chunk of traffic to UK newspaper sites comes from abroad: almost 70% of the MailOnline's 20 million or so uniques come from outside the UK. In the case of the Guardian.co.uk the total number of UK unique users in December is 8.4 million (36% of the total number), so already the big "input" number starts to get smaller. Often the country filter is the only one to be applied in the reported press, but let's see if we can't shrink the number a little more.
The Newspaper Marketing Agency's data has a few further interesting data points which point to stories less often told:
- 56% of newspaper site visits last for under one minute
- On average people visit 1.8 different sites to get their news
- On average they visit these sites only 4.3 times a month
- In December only 23% of visits to these sites started at the site's homepage
Furthermore, one has to consider the, often dramatic, difference between the number of *daily* unique users vs. the number of *monthly* unique users. The MailOnline, for example, gets 1.1m daily uniques but 19.7m monthly uniques. You cannot just multiply the number of daily uniques by the number of days in the month because the same user may visit several times in the same month and be counted multiple times as a 'daily' unique user but only once for the month. So its 'reach' is not 19.7m unless you advertise for a full month.
The data does not paint a picture of loyalty. It shows an audience snacking on content. A readership that is not loyal to the publishing brand. In fact, if I were looking for loyalty then I'd look for properties where the monthly unique users figure was as *close as possible* to the daily unique user figure showing that their users keep coming back - I'd be looking for a *smaller* monthly unique user figure in fact.
Some metrics I'd like to see published which get under the skin of the above:
- What's the "bounce rate" for the site? That is, how many visitors see one page and then leave straight away.
- What's the average page load time (as experience by a customer) and how many visitors leave in less than that time period (i.e. never actually even see the full page rendered)?
- What IP ranges are excluded from the traffiC? Is all employee activity from the office, and from home, being excluded? Along with any of your agencies/suppliers etc.? i.e. anyone who isn't a genuine reader.
- What percentage of traffic comes from Google News or other news aggregators?
- What percentrage of traffic comes from search engines (aka Google) vs. direct? Of those, how many were searches on the publishing brand versus 'long tail' keywords?
- How many unique users visited the site via paid search? After all, anyone can buy traffic.
Some of the above data is published; and some of it is indeed audited/excluded as part of the ABCe process; and some of it is perfectly legitimate traffic. However, I know many site owners, not ad-funded publishers of course, but retailers, travel companies and so on, who exclude 'bouncing' visitors from their web stats when computing their site's conversion rate because they don't represent engaged users: 'bouncers' were never really there to be sold to so shouldn't be counted as a real opportunity. Should publishers do the same? Are you happy as an advertiser to pay for ads 'viewed' by a "bouncing" user who looks at the page for 20 seconds? Can you feel the big numbers rapidly shrinking?
I think if we knew all the figures above it would paint a picture that showed a scary reliance on search and a user base that is much less loyal even than already known. To a large degree I suspect that a newspaper's 'big number' figures are purely a reflection on how well they rank (internationally) in Google, both in natural search and Google News. Nothing wrong with this if one believes that Google promotes the best content perhaps; but, if true, has quite far-reaching implications for business models, policy and politics, and the 'big input numbers' of online advertising.
A new focus on value and outcomes
I think we should be getting away from big number bravado as soon as possible, not least because many markets are fast plateauing in terms of available internet users so publishers are going to need to get used to spinning a positive story of decline which will be all the harder if they've been trumpeting growth too recently.
I think we need to get better at measuring, and delivering, value rather than volume. Outcomes rather than input. Ultimately I don't care how many squillion unique users you've got if it doesn't help me sell car insurance, holidays, TVs etc.
This is a little scary for many publishers as the truth is they don't actually *know* what value they deliver, and they don't really want to find out.
Digital media agencies I've talked to, when I ask them which sites 'work' best for them and their clients, tend to say the smaller, niche sites. But they are time consuming to deal with and the client may not know them so the agencies don't necessarily want to take the time selling them in to their advertising clients. I fear it is still the case that many big name publishing brands are getting away with trading on their brand heritages and advertiser fear of not going with the 'usual suspects'.
This is likley to happen even more so in the recession but, in the end, I cannot believe that efficacy and real value will not become transparent to the advettiser and therefore only those publishers delivering value will survive. Indeed, they will flourish - if all sides know the publisher's true, and measurable, and transparent, ability to drive brand and purchase favourability, or actual sales, then that publisher is in an immensely powerful position. The internet means there will be nowhere to hide in the end if you don't deliver value.
To go back to the Guardian... I was recently trying to find a good family-friendly hotel in the South of France. Incidentally this taught me, among other things, that there is a fair degree of truth in the 'search is broken' argument as Google is still really poor for such searches. Nonetheless a Guardian article came up and I duly visited (1 of my 4.3 visits a month). I looked at one page which gave me enough information to make a decision about where/what to book and duly went off elsewhere to do so. The Guardian didn't just influence my buying decision; it pretty much made it for me. So they delivered a transaction worth four figures and got 1 page view in return. Doesn't seem fair on the Guardian losing all that value?
And recently, talking to a friend of mine who works in PR at a large public gallery, she said that they could attribute thousands of extra visits to the gallery following coverage in the Guardian, worth tens of thousands of pounds. And yet this was, in the end, measured as 'equivalent advertising value' i.e. a media input, not the actual outcome, with a value much lower than actually delivered. Again, doesn't seem quite fair on the Guardian?
Publishers need to think more like retailers
I think publishers need to start thinking a lot more like retailers. They need to look at their 'readers' as "customers" with RFM (recency, frequency, monetary) data and segmentation - when did this customer last visit? How often do they visit? How much value have they created (directly or indirectly)? How much more, therefore, can I charge in advertising for this customer / segment?
Publishers need to look much more at affiliate marketing and lead generation to track outcomes, not input. They need to get better at capturing user data, surveying their customers, to find out more about how they are influencing buying behaviour. Why not do more post-reading questionnaires to ask them what they plan to do next as a result of their reading or advertising exposure? Can an uplift in internet buzz or searches around the brand be attributed to online advertising?
Publishers need to get more innovative - the change in the FT.com's subscription model being the most laudable recent innovation in publishing business models recently in my view.
Crossing the editorial / commercial divide
I also wonder whether there isn't an opportunity, in this age of blogging, Twitter, Facebook et al, to re-evaluate the unwritten "contract" between the readers, the editorial team, and the business model. 'Editorial' has historically had to at least maintain a "church and state" / "whiter than white" separation from advertising/commercial. And yet I'm not convinced the average reader *really* trusts some journalists any more than some politicians. We know certain newspapers, or journalists, have a particular 'angle' or vested interest that spins their words.
Meanwhile, we revel in reading the wild rants of our favourite bloggers, basking in their manifest, if sometimes misguided, honesty and almost pathetic transparency. Sometimes we don't even know who they are. Do we care? I think we care about the content itself and we're evolving very sensitive "fakery antennae" i.e. we can spot when we're being manipulated e.g. all those lame corporate blogs that have been 'outed'.
Equally, I think readers realise someone has to pay for the content, but, in many cases, they'd just rather it wasn't them. All of which means that I think it would be perfectly acceptable for newspapers to create content that is editorially driven and originated but then seeks to monetise it via clicks, affiliate deals etc.
If only the Guardian had put a bunch of links at the end of its travel article telling me where I should book/buy I would gladly have clicked and given my bit of value back to them. Digital Photography Review site dpreview.com did this very well, though it has since been acquired by Amazon so the affiliate links point, not surprisingly, at Amazon.
To me there are all sorts of interesting PPC or affiliate-type opportunities for publishers willing to innovate. For example, why not allow affiliate-tracked links to be included in user generated content on the publisher's site? Perhaps the publisher could split the revenue generated with the user? This would encourage participation and reward it. You get free content, more traffic, and higher commerce revenues. How would you police it and maintain quality?
Just like the many other communities, rating and review sites out there - it's not perfect but it is doable and manageable. As has already been discussed in publishing circles I think the resourcing model will move away from loads of editorial staffers, production people and sales people, and more towards 'community managers' and 'sub-editors' who will manage the quality and input of users because they will be generating revenue for the publisher.
A gradual transition
Clearly none of the above changes can happen overnight. Not least because, however forward-thinking the publisher, the advertisers and media agencies themselves are often more backward and conservative than one would hope so publishers need to stick where the money is, however misplaced.
But I hope we'll at least see some more hybrid models starting to emerge which mix outcomes with input, which champion value over volume. And I think publishers should spend more time thinking of themselves as retailers, or membership organisations and communities, where sales, customer value and loyalty, rather than big 'reach' numbers are what count most.