Programmatic advertising (aka strapping an engine to a horse)

Strange brandsMy thanks to the team at Digital Doughnut for drawing my attention to this relatively new report by eConsultancy and Quantcast. It is called ‘Programmatic Branding: driving upper-funnel engagement’.

I read the report and it gave me a vivid reminder of a feeling from my agency days where you find yourself in a conference room (probably in New York) listening to a presentation about ‘the new big thing’ (albeit presented in such a way as to suggest that this is not sufficiently new that you shouldn’t already know quite a lot about it) and find yourself struggling to resolve two competing reactions. The first is panic: OMG, the fact that I can barely understand what this guy (probably a guy) is talking about shows that I am just so out-of-touch I may as well resign right now. The second is: OMG I suspect this guy himself barely understands what he is talking about and is just making up a whole new set of words to either disguise his ignorance or take something which is actually very simple and make it sound incredibly complicated.   Experience has shown that the latter usually prevails.

Anyway – the only way I have learned to resolve these reactions is take what is presented apart, to not be afraid to ask the obvious questions and, critically, interrogate the assumptions upon which it is based.

I have done this with this report and my main conclusion is that attaching the word ‘programmatic’ to the word ‘branding’ or ‘advertising’ may be a bit like attaching an internal combustion engine, gears and wheels to a horse. An enormously complicated exercise that may, indeed, allow a horse to go faster but still never fast enough to compete with a car. It is an exercise in failing to recognise that in the speed (digital marketing) game, a horse (advertising) really no longer has a viable role to play, no matter what technical wizardry you try and strap onto it.

A definition

Let’s start with trying to define ‘Programmatic Branding/Advertising’. Slightly unhelpfully, the report assumes that you know what this is (see earlier point about implied understanding and the automatic feelings of knowledge inferiority this can then be used to create). My definition is this:

Programmatic Branding is a word we have invented to imply we have a solution to the problem that marketing – specifically digital marketing – has now become so complex it requires a machine (a programme) to control it; this complexity being driven by proliferation of available channels (defined by channel type as well as time of usage/access) and the consequent variety of message (called inventory) that can or should be placed within these channels.

A cynic might say it is a bit like saying to Monet, “hey dude, sunflowers, paint, canvass, that’s just too restrictive man. There are a hundred different ways we can image a flower and a hundred different flowers in the flower shop and, anyway, people want to buy flowers in real-time. Real-time buying, massive inventory management, multiple formats – that’s the future man. So put down the paint brush and embrace the future, or go home.”

The assumption

Perhaps I am being a little judgemental, but my issue here in going back to basic assumptions, is that no-one has challenged the fundamental assumption that programmatic advertising or branding has a right to exist. Advertising has a right to exist and branding has a right to exist. Likewise, programmatic approaches can be incorporated into what we might call marketing. But where is the evidence that the two should go together or that the assumed future of one (i.e. programmatic approaches) can be used to assure the future of the other (i.e. branding and advertising)? The report doesn’t make this case, indeed it presents evidence of an inherent contradiction between the two, although this contradiction is disguised by usage of jargon.

Here is an example. In a section titled ‘Programmatic direct and workflow automation’ it states that, “One of the barriers to programmatic branding has been the absence of quality inventory available in programmatic channels.” Indeed this problem of inventory – sometimes called inventory inefficiency – is mentioned a number of times in the report in terms of its insufficient quantity or quality. Inventory inefficiency is thus set up as a barrier (and a nice piece of jargon) to be overcome or a problem to be solved. But this is a bit like saying, the problem with conversations is that you can’t script your responses (produce sufficient inventory) in advance. And then proceeding to find the solution to this problem by searching for the maximum number of conversations into which it might be appropriate to drop your restricted range of pre-prepared sentences, believing that your relevance in participating in these conversations is determined by the fact that you have a sentence that might fit within it. Or as the report puts it “Today, Deal ID and private marketplace functionality are enabling brands to create direct inventory connections in biddable environments.” Well quite.

But, of course, this is not the way that conversations work and you are not going to extract the value available in a conversation by adopting such an approach. Or as Michael Palin (sort of) put it, in a memorable sketch in Monty Python and the Holy Grail, “strange brands, lying around in conversations, distributing message inventory, is no basis for a system of marketing.” (If ever there was a sketch that was about challenging assumptions and possibly even about the increasingly dysfunctional relationship between brands and consumers, that particular sketch is it. For those not in the know, here is the script and the video).

The report also then touches on another favourite of mine which is the presumed existence of the ‘omnichannel’ and the need for marketers to develop ‘omnichannel competencies’. The omnichannel is another invented concept designed to deal with the difficult fact that consumers, damn them, use lots of difference devices, for different things at different times resulting in the fact that “marketers no longer have a reliable ‘reach’ channel .” Hence the invention of the omnichannel, a form of mythical channel that can revitalise marketers flagging ‘reach’. But no consumer ever uses an omnichannel. It doesn’t exist in the world of the consumer, it is a convenient figment (or is that fig leaf) of the marketing imagination.

Define the problem before reaching for a solution

At the heart of all of this is an inability to define the problem. The problems and complexity the report identifies and then luxuriates within are only complex problems because of the way they have been defined. If we define the challenge as one of channel and message, then the proliferation of channels and messages is a complex problem, requiring a programmatic solution.

But what if, in the digital space, the challenge is no longer one of channel and message? What if the challenge is one of behaviour identification and response? If we look at things this way all of the problems programmatic advertising is designed to solve simply melt away. You don’t have to go through the complicated business of hunting down the individual consumers for whom your ‘inventory’ might be considered relevant. If you sit still and wait (listen) the consumers will find you. Consumers will identify themselves and the nature of the situations in which they will do this define that an inventory response will almost always be inappropriate. If one wishes to still use the term programmatic, one could say that it is not programmatic advertising or branding that is required, it is programmatic listening and response.

And this is why the technology challenge (and I prefer now to use technology rather than programmatic) in marketing is not about creating ‘direct inventory connections in biddable environments’, it is about how you use technology to manage a process which allows you to listen and respond to individual consumers. The output is a process, not a piece of inventory and the technology is used to facilitate that process – a process which inevitably will involve a level of human intervention. And the value is created, not through the volume of connections (reach) but through the value of the connection or experience we now have the opportunity to create. This is why I foresee a battle between facilitation platforms, such as Sprinklr (the car makers), and automation players, such as Quantcast (the horse modifiers). And I think the facilitation players will win. However, this won’t happen without a fight. Remember the words of Henry Ford when he said that if he had asked people what they wanted, he would have built a faster horse.

Digital advertising: its works, but does it work enough?

A channel and message approach, measured by reach and frequency, is a high reach / low engagement game. But the opportunities within the digital space are characterised by the possibility of creating much higher value connections, albeit at much lower volume. Advertising, especially if defined as purely an ‘inventory connection opportunity’ is never going to create enough engagement. That word ‘enough’ is the key to it. You may be able to add programming to your digital advertising to make it more effective, but will it ever make it effective enough to overcome the manifest other difficulties and forms of resistance advertising encounters in the digital space? It is currently regarded as an unthinkable, but possibly, just possibly, digital advertising may not actually work in the digital space. It may look like it is working and we may be discovering techniques to make it work better, but a final judgement on its effectiveness can only occur when we have the opportunity to compare it with alternative approaches to using the digital space that might work better – approaches which we may not yet have discovered or, as is more likely, have been identified but not embraced, simply because they lack the institutionalised familiarity of advertising.

And let’s just face it, you don’t have to be a creative director to realise that if we have got to the point where what we produce is called inventory, what we are doing no longer deserves to be called advertising.

This is not to say that advertising no longer has a role: this is a bit like telling Monet to put the paint brushes down. Rather it is more about telling Monet to apply his creative juices to producing something relevant to the day. Perhaps he may decide to put a shark in a tank of preservative – who knows? What advertising, or rather creativity, has to do is adapt itself to the world where the channel no longer defines the message. It is not about targeting audiences, it is about convening audiences. It is about creating ideas that draw audience-sized groups of people around a brand: ideas that define the channels they sit within rather than being defined by the channels they are required to sit within.

Neither is this to say that programmatic approaches won’t work. Algorithmic (as distinct from programmatic) marketing is undoubtedly the next big thing, albeit this could also be a concept that is so powerful it won’t need to be strapped to something called marketing to make it work. Algorithms hold out the possibility of doing away with the need for marketing all together. There is another Python sketch which illustrates this (there are in fact Monty Python sketches available to illustrate almost all aspects of existence). This is from Life of Brian, where Brian is in desperate need of a beard to disguise himself from pursuing Roman soldiers. He goes to the beard seller, asks him the price, gives him the money and is about to take the beard when the seller stops him and demands that he haggles. The haggling is equivalent to marketing – a sort of warm-up process required to reach a sale and agree a price. But in this instance it is not necessary – Brian just needs the bloody beard, he needs it now and is prepared to pay over the odds to get it. That is what algorithms do. Why broker messages in biddable environments when you can broker the product (or the need) in biddable environments: the message bit is just a distraction and delay in much the same way that ads in the digital environment may well become seen as just a distraction and delay – a relic feature that we are reluctant to do away with.

Perhaps I am wrong and this stuff is genuinely complex stuff, which I struggle to understand rather than stuff that has been made artificially complex in order to panic marketers into hiring experts.  Afterall, strapping an engine to a horse is not something you should try at home, but before you do it, just ask the question: why?



  1. Pingback: TechCrunch Disrupt: putting disruption in front of success - Richard StacyRichard Stacy

Post a comment

You may use the following HTML:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>