The marketing communications industry has a touch of "electionitis" about it right now, too. So much is written about the proliferation of communication, yet, similar to election fever, we are losing sight of the fundamentals.
Instead of working out who should see what, how often and where, or reacting to who has said what and where, and assessing that, we are left with a fug of measurements. We are over-reliant on techniques, measures and KPIs. These seem mainly to inform marketers about the journey they are undertaking, rather than the destination for which they’ve bought a ticket. That’s all fine if you see your marketing budget as a luxury cruise; less fine if it needs to get you to a destination by the fastest and most direct route possible.
Having worked in media for more years than I care to remember, and planned, bought, sold and evaluated the performance of more media channels than I can shake a stick at, what frustrates me most is that when it comes to working out effectiveness, almost nothing has changed.
Whether we’re talking about traditional or digital media in all its forms, it’s still a game based largely on how many eyeballs are reached, and how often, to generate some kind of effect based on sentiment, engagement, recall and consideration. It’s not about tangible outcomes. Yes, there are sophisticated algorithms, automated technologies and techniques to define whom these eyeballs might belong to and who might be most likely to buy a product; and yes, people’s media-consumption behaviours can be tracked so you can send them ads based on the content they’ve been looking at.
Nonetheless, these are ultimately proxies for what really matters – whether there was a sale at the end of it all. Sometimes this is easy to work out. If you’re an ecommerce site, you can at least see the winning last click to basket, and even track some of the user’s route to it. In many industry categories, however, this kind of purchasing represents, at best, 10%-15% of sales – and even this varies by country.
So we’re left with measures of performance that, in the old days, focused on reach and frequency, then impressions and uniques, and are now overlaid with various "viewability" scores, sentiment indexing, NPS (Net Promoter Score), PTAT (People Talking About This) and any other acronym that sounds sophisticated and a touch mysterious, so we’re all too scared to ask what it actually means or does.
To gain some understanding of outcomes, over the years sophisticated media-attribution models have been developed to indicate the incremental sales-effects campaigns might produce. They sometimes vary in name and methodology, but were developed in a time when media was more "mass", its consumption more predictable and people’s shopping habits more regular. These metrics are OK at providing broad indications of ROI and directional advice about communication mix.
However, they struggle to identify what key communication drivers (if any) are behind sales, let alone establish where those sales might be coming from and what the key influences were in order that strategies can be focused and optimised on those drivers. In a world of shifting consumption patterns and fewer neat ways of segmenting and targeting people (the mere mention of such terms as "millennials" is not good for my blood pressure), these models, whether called marketing mix modelling, econometrics or anything else that sounds very scientific, don’t get us close enough to the most useful outcomes.
Some may say that we’re in the kingdom of the blind in all this, so one eye is better than none, but faced with such fast-changing technologies and habits, we should expect to have at least two eyes on the job. And that’s before we’ve donned our VR headsets or augmented our reality with mystical Pokémon Go creatures.
Knowing price, measuring value
One could argue that we’ve got to this state of affairs for several reasons – some plain old commercial ones, and some technical. Channeland content-owners may lack the confidence or methodologies to demonstrate the true efficacy of their wares; marketing service partners may feel the same way about the campaigns they’re producing.
Meanwhile, marketers sometimes aren’t prepared to really push for the answers, either in terms of funding or in how they think. It was not long ago that I was with a major global advertiser and their various marketing service partners, and they reassured themselves that, although they knew what they were doing was wasteful, at least that wastage was cheap. Then, of course, multiple media channels have multiple currencies with which to measure their effectiveness.
So even when the focus is on audience and engagement, as opposed to outcomes, it’s tricky to compare like with like, what their effects might be and how efficient they are, singly or in combination.
So, are we all doomed to continue this marketing communications life of knowing the price of everything and the value of nothing forever? Well, not necessarily. It’s all about finding an outcomes-based currency that works across all communication and media, whether that communication happens in a store, outside one, on the way to one, at home, through a device, on an ecommerce site or, indeed, in pretty much any other way. It certainly seems media audience data, in all its forms, is unlikely to do the trick for all the reasons given above and regardless of what many people might say. But there are data sets that might just get us to where we want to be. It’s shopping behavioural data, whether online or in-store (preferably both). It’s the stuff that many major retailers and service-providers use – output from CRM programmes, loyalty schemes, debit/credit cards or ecommerce sites – and tracks their shoppers’ purchasing history.
The insight generated from these huge data sets is used largely to help those retailers, serviceproviders and their suppliers (sometimes) manage product ranges within categories, layouts of fixtures and store plans online and offline, as well as negotiate trade deals on the basis of shopper insight, rather than just volume.
Effects of this "trade" activity are tracked against not just sales uplifts, but also whether a product is being tried by a shopper for the first time or is a repeat purchase; whether they’re switching from one brand to another or from or to retailers’ own-label products. It can show peoples’ repertoire of shopping and how loyal or disloyal they are to particular brands and categories; whether they buy at full price or how much they are swayed by promotions. It’s a wealth of information many of us know about, but seems to rarely move out of the category solutions "trade" marketing arena.
It’s employed, to varying degrees of success, within retailers’ owned, targeted media, such as CRM or direct-mail campaigns to target shoppers with special offers (sometimes co-funded by brand-owners) based on purchasing behaviours, but in the great scheme of things, it’s underused in marketing communications overall.
Yet the metrics it produces about weight and frequency of purchase, trial and repeat purchasing, changes in share of requirements or share of wallet are consistent, regardless of the changing influences on them. So suddenly a TV campaign can be measured in the same way as social media, billboards or any other media or messaging channel. It’s mediaeffectiveness nirvana.
There are a few hiccups on the way to reaching this heavenly place, though, including a couple of technical ones.
First, it’s a question of how to align peoples’ exposure to communication with the impact it has on how they shop, especially when this exposure could have occurred anywhere along their path to purchase, however circuitous that route might be. Also, the consumer of a product may not necessarily be its purchaser.
So we can’t include only the media channels in this, we must also add the influences we traditionally look at – the effect of changes in sentiment, impact of "likes", click-throughs, retweets, recall and awareness, among others. Second, it’s likely that this kind of heaven is going to be reached only in categories that have relatively short purchase cycles, so one can see the changes in behaviours over a commercially practical length of time.
So it lends itself more, perhaps, to such sectors as grocery, financial services, travel and fashion, rather than to, say, cars and property. The final hiccup is more structural. You will have heard it a thousand times before, but we all seem to work in silos – whether "trade", "commercial", "brand", "consumer", "customer experience", "traditional" or "digital" – and, when all is said and done, we struggle to join these up. So shopper analytics and all the expertise that goes with it tends to sit under "trade", while communication and media expertise, and supporting analytics, tend to sit under "brand".
Broadly speaking, marketing departments are often structured in this way, and their marketing-service partners seem to reflect it.
A new world emerging
These hiccups don’t make "achievable change" impossible, however. Methodologies have been developed for many media channels that enable the effects of exposure and engagement to be measured against shopping behaviour, either using geo-locatory techniques (at multiple levels of definition) or user-matching, and change is already afoot in this area.
Facebook, for example, has matched some of its user data to research firm Dunnhumby’s UK Tesco Clubcard shopper data, and in the US, cable network Comcast has matched viewing behaviours to various shopper data sets. The trouble is that there are some "old-school" behaviours going on in all this – it’s Facebook’s use of the data to show how good its channels are, rather than those who advertise on them. Perish the thought that Facebook would do such a thing, but one can guess it prefers to tell clients of its successes rather than failures, so its version of the truth may have a bias.
Then data-owners, such as Amazon, which are making moves to become both mediaowners and retail channel-owners, appear a little reticent to share their shopper behavioural data with brand-owners. Coming from such powerful retail propositions, this seems a surprising lack of confidence in their own channels – or is it just another example of old-school media thinking? I don’t have the answer, but as our world becomes more accountable, and with the ongoing presence of advertiser pressure, I’m sure these kinds of data sets will open up.
The issues with structure are different. There is always much talk of breaking down silos, but quite often this talk is siloed in itself. The classic example is discussions about breaking down silos, when in fact the silos being referred to are within "digital", as if other worlds don’t exist. Part of this is down to ways of working; we all tend to sit within specialisms – part of it is vested interests relating to budgets and the internal politics that go with them.
There are pioneering clients who recognise this new world of outcomes-based marketing communication and are making attempts to use it. But more people need to be prepared to take leadership, and get the bloody nose this entails, to drive through these new techniques, join up marketing disciplines, datasets and insight to enable a truly brave new communication and media world rather different from the one we currently inhabit. A world that will be about the outcomes, stupid!
What CMOs need to do
Establish the incremental sales effects of past marketing activity and establish the shopping, consumer and media behaviours that have driven these sales effects (in both the short and longer term).
Use these learnings to set future objectives, strategy, tactics and the KPIs you need to chase to realise your objectives.
Continually track performance against these KPIs and course-correct activity to meet them.
Ensure your shoppers and consumers are on a journey with you based on how they behave. Know when to say something and when to leave them alone.
Always take the shopper or consumer’s point of view – how they shop, consume and engage with communication is a joined-up experience, so your expertise needs to be, too.
Marketing communication is ultimately about outcomes. Always make sure that you are focused on a commercially worthwhile end point. (Even if it takes time to achieve this.)