Having spent time looking at some of the key moments and influencers in the development of the online advertising business, what strikes me from my own personal observation and reflections on the industry is the plateau I feel we have reached right now.
Oh yes, I appreciate the growth of e-commerce and the public in their spending this last Christmas. Yes I too am aware of the growth of our sector where we are to now a $20bn industry with confidence in yet huge growth, shown in part by the $10bn investments last year into the advertising technology sector by the key portals, and even the offer yesterday of $44bn from MSN to purchase Yahoo! (Interestingly they turned down a rumoured previous bid of $80bn a couple of years back, but that was before Yahoo! announced laying off a 1,000 staff these last few weeks).
But in all the buzz and razzamatazz of colossus numbers banded around by a handful of players, it is easy to lose perspective of the client-consumer relationship of which we are charged in developing and the effects that are happening right now brought on in part by web 2.0 thinking and the consumer answering back.
Online advertising has reached puberty, and it’s going to be a painful awkward navigation through it. What is going to make it more increasingly difficult is no-one has been here before directly to offer insight and wisdom as to how to deal with extremities of emotions that are flying around. We must look outside of ourselves and take stock, if ‘web 2.0’ is going to not end up in ‘dot-com crash 2.0’.
I often dissect the meaning behind the desire to create ‘media firsts’ as a starting point in most of my talks. Why? Because it is the one driver that almost all planners and designers converge on when looking at delivering a compelling result for a client. The real definition behind this phrase is “what new format or feature can you recommend that will get me the best click-thru rates?” Herein lays the problem. It is not the desire to be new, inspiring or creative – which are all to be applauded – but it is the desire to do them in and of themselves, art for art’s sake, with no standard to measure them against.
Google and DoubleClick helped define the search and display advertising models with the ‘click’. They gave a starting point for monies to shift into the digital space. It was a simple sound-bite of which to assist clients in their understanding of data and what it could do for their ROI. We were going to save them money! This ‘click’ has become the de-facto convergence point that drive all campaigns with desperation to out-perform the previous click-through result, in hope to both win an award and retain the client account. This is the plateau to which I allude.
The plumb line the industry has made for itself, the benchmark against which all seek to measure is skewed. It is seriously distorted. Clicks do not equal conversions. Clicks do not reveal satisfactory response and clicks most certainly do not reveal branding. Clicks out of context do not even reveal intent. Clicks can be accidental…
Click thru rate, as measured against impressions (Clk/Imp x100) does not even mean anything useful in isolation unless you look them through the eyes of reach and frequency, i.e. how many times did a person need to see the ad before they responded? Define response. Provoke them to click? Provoke me to a point of annoyance by seeing the same annoying advert again I try to get rid of it? Bingo, I have now become a statistic of success to someone, somewhere.
Online advertising is not linear. It is not “I see therefore I click” or even “I click therefore I want” in some kind of nonsensical Descartes theory. If we are going to lead change, we must have a thorough understanding of what is really going on behind the scenes.
Different formats produce different results. An over-the page ad generally will deliver higher ‘clicks’ then an in-the-page format – and so does that mean that is because it is intrusive and therefore more effective or intrusive and more likely to provoke a user to annoyance to get rid of? If we agree on this latter aspect as a possibility, if I reveal a new format that delivers a still high click-thru are we going to accept it at face value, or look at what really motivated the click? Motivation must be creative vs format in context. The same creative and format will produce different results on different pages, so now neither creative or format; two consistent here produce the same results when the factor of placement is altered. No consistency, no benchmark. Our analysis fails unless we look deeper.
It is this superficiality that drives me insane when I read that ‘oh look, I have quadrupled click-thru’s by adding some clever technology’ as I read earlier this week. So what?! You are trying to argue relevance, but in a way that is out of context and measured against a skewed plumb-line. The client had already given away their game plan in terms of wanting to drive conversions. Now it may be they did indeed deliver on that promise, but the point of the article was not we drive conversions, but we drove clicks. Tell me the former and I will ecstatically share their enthusiasm, share the latter and with a little push, the wall of the house will fall like a deck of cards.
Norms, in statistical terms are derived from a 95% agreement. It assumes that for 95% of the time, this is what I expect to happen. 2.5% variants either side of the range are acceptable. That is 2.5% of the time things happen which are wildly off the scale and yet are seen as anomalies that are still factored in when defining a ‘norm’ in the first place – we expect things to be different, 5% of the time. So when something happens outside of the ‘norm’, is it to be praised or investigated?
Analysis of online data cannot be measured in linear paths, any more than human behaviour can be. There are multiple drivers for getting us to be motivated and responsive – and each product will pull at different aspects within us – the brand, the price, the colour – depending upon our needs and requirements at that time, mixed in with where we are along the purchasing lifecycle. I believe all of these aspects can be seen to some degree within a data ‘cube’ in analysing a campaign success. By looking at data from multiple angles, looking at it in 3D if it were, such as the many sides in a cube, we will get a more whole picture of this human behaviour and how it relates to overall campaign performance.
Using Dynamic Logic’s branding measurement systems may be skewed as it may only cover a ‘type’ of people who would normally fill in questionnaires – so how can this be representative of the whole audience we wish to analyse? Your ‘clickers’ – are these people who are investigating further, or looking to make a purchase right there and then – how does the post-click experience relate to where they are at? In search, we are well aware of the need to deep-link, i.e. serve relevant page content to match the text message they clicked on – ye t in display advertising, one landing page is supposed to match all?! It would seem to me that very often campaign micro sites are over bearing with gimmicks delivering games and downloads to people ready to convert is going to do nothing to assist in any kind of conversion, and thereby may be a contributory factor of seeing a down turn in clicks per se – people just do not feel the content in the majority of cases is in any way relevant to where they are at.
We must get back to looking at people as unique individuals, and at best seek to put people into groups online. We must also pick our moments as to when to deliver the message. Behavioural advertising and adaptive creative’s are a powerful aspect of the next wave of digital advertising, but better understanding context within media planning; how the same advert in different environments will produce wildly different results – what is the ‘norm’ for that environment – are people more likely to interact or click when typing an email in Hotmail, for example. Deeper investigation in to previous campaign performance, will reveal a host of attributes that need to be fully understood before you confuse the point of reference with adaptive creative’s – there will be simply be too many variables for you to fully understand what is going on.
Starting with trying to see your data as ‘groups of individuals’ – the Interactors, the Clickers, the Post-Impressionists – trying to match this against page context or purchasing life-cycle – this alone will give you multiple data-sets that will reveal much about how to better optimise a creative within any particular environment. Being able to deliver a branding message or an overview of product features to the people at investigative stage, who may be open to delve further – but how? The majority will be happier to interact on page, only few of which will click away – and that assuming the point of origin was in a position encouraging them to do so (i.e. in an news or information site, as opposed to an email or content creation environment when they are possibly least open to advertising driving clicks away at that particular moment).
Interactivity data is a hugely powerful beast that is yet to be tamed for most online advertising – most people are so over-whelmed, and under such time-pressures that they ignore it and stick to simple ‘benchmarks’ – of which, sadly, there are none. It is a huge concern to the industry as clients become more aware then ever as to the power of digital advertising, delivering half-baked branding studies and impressions and clicks will never deliver the potential, and with it the budgets so desperately representative, of the power that digital advertising could so easily be qualified to demand. We don’t need more data, we just need more front-line analysts to help us push over this plateau…