Many brands use social media to seek to get their message across in a pro-active and engaging fashion. Yet famously, Wetherspoon’s announced last week that it had stopped using social media (http://www.bbc.co.uk/news/business-43781281) because, as Chairman Tim Martin told the BBC, its decision had also been influenced by concerns regarding the “misuse of personal data” and “the addictive nature of social media”.
“We are going against conventional wisdom that these platforms are a vital component of a successful business,” he said. Furthermore, “He had always thought the idea that social media was essential for advertising was untrue” adding that “and we were concerned that pub managers were being side-tracked from the real job of serving customers.”
So, after consulting their pub managers, of whom “90-to-95% felt using social media was not helping the business”, they decided to drop it.
This comes, of course, on the back of the controversies surrounding Cambridge Analytica’s use of harvested Facebook data which led to Tesla and SpaceX boss, Elon Musk, having the official Facebook pages for his Tesla and SpaceX companies deleted.
But are they right?
Engagement versus Presence
A key belief amongst those who advocate the use of social media as a marketing tool is the idea that it can create a deeper and more long-lasting connection with customers, shoppers and consumers. Whilst that is undoubtedly true, the question is about how much effort you put into achieving it – and whether that generates tangible financial benefit for the business.
As evidence to support their decision, Whetherspoon’s gave one example that “A tweet pushing fish and chips on Good Friday got just three re-tweets.” And they said the pub chain had certainly put plenty of effort into it, creating hundreds of different Facebook, Twitter and Instagram accounts. But the truth was that none had won much of a mass following.
In the view of management – those running the accounts were not doing a very good job.
Our view, though, is that whilst social media provides the opportunity to create engagement, and that doing will, inevitably, have a deeper impact on the people directly touched by it, by far the biggest benefit comes for the “publicity / advertising” effect of social media.
This benefit is far harder to quantify because it is not direct. It requires using econometric modelling of sales data to disentangle the myriad of influences upon sales performance and net out the contribution social media has added to that. It may not be as big or immediately noticeable as a price cut or other short-term promotion, but that does not mean it is any less profitable. Indeed, our measurements show the benefits to be highly effective and profitable.
The real benefit comes from presence, not engagement; the ability for your brand to be noticed by millions of people at next to zero cost – and often in an “invited-in” manner.
So, what’s the secret to being able to put a value on the benefit of your social media?
In June 2017 Binet & Field published their latest review of the IPA Effectiveness case studies (Marketing Effectiveness in the Digital Era, Binet & Field, IPA, June 2017) and one of the interesting things they highlighted was how marketing effectiveness was affected by the mix of Brand Building marketing investment and Brand Activation that one chooses to use.
It has been known for a long time that actual sales performance is driven by two elements, historically referred by as “Push” and “Pull”, the former being short-term sales-driving initiatives like distribution gains, shopper marketing, and direct response activity; and the latter being long-term ‘brand building’ initiatives like advertising and experiential marketing.
The question is, of course, where does social media fit into this?
Is Social Media Brand Building or Brand Activation?
Brand Building investment tends to be equated with traditional ‘non-digital’ media whilst Brand Activation tends to be associated with in-store activity, direct marketing and promotions. The former being “image building” activities with no immediate expectation of a sales response and the latter being all about short-term sales response.
The problem with social media is that it can do both – and the short-term response is easiest to see. That makes it measurable but almost certainly under-values it. Social does more than just provide an opportunity to drive short term sales uplifts and is worth more than the value of the sales measured via ‘last touch attribution’ measurement systems.
Like both advertising and experiential marketing, when used effectively, social is a strategic marketing tool that influences long-term brand preference – and for strong emotional-association reasons, not just because it’s convenient or provides an easy-way-to-use the brand (important though those things undoubtedly are in creating short term sales response).
To measure the full value of your social media to your business, you need to be able to quantify its effect on sales that comes from its influence on underlying brand preference in addition to the short-term or direct sales attributed with physically using the channel.
The way to do that is to think about evaluating social media in the same way that marketers do when measuring the sales benefits of advertising – econometric modelling.
Using exposure to social media as one of the data inputs to the econometric model alongside the directly-measured short-term sales responses via actual channel usage, you can measure both the long-term benefit of social as well as the net incremental benefit of any (short term) sales attributed to the channel. Remember, the latter may have been diverted from other channels or be brought-forward sales etc so they can never just be taken at face value.
How Econometric Modelling Works
Explaining to the world at large how econometric modelling works is one of the perennial challenges of the marketing world! Not everyone has a maths or stats degree!
One simple way to think about it is to use a real-world analogy.
Imagine you have a set of weighing scales and a wide selection of differing kinds of fruit. Now imagine weighing loads of times (pun intended!) different random collections of fruit just the one shown in the picture below. Each time you weigh you make a note of the weight and what the mix of fruit was in the bowl.
After a while you have created a set of data you can analyse. Six oranges plus three bananas and two apples weighs this amount; two oranges, three bananas and one apple weighs that; and so on. Make enough measurements and you are soon able to use the answers to work out that, therefore, one orange must weigh this, one apple must weigh that, and one banana must weigh something else.
In the real world, of course, it is slightly fuzzier than that because random events come along all the time which means the measurements you make are never precise. Granny falls over, breaks her leg, and you have to take her to hospital instead of going to the shops. Lightning strikes the store you were going to visit, burning it down and stopping you buy. It’s like measuring the fruit but every time you do so a unseen Imp throws in a small bag of grapes as well but you cannot undo the bag to know how many grapes are in it. It is just a hassle factor.
Fortunately, you can use statistics to work out, for example, that you can be “95% confident” that the Apple weighs this, the orange weighs that, and the banana weights the other.
In a nutshell (or rather a weighing-pan), that is what econometrics does.
Instead of fruit we are weighing the impact of £1 spent on advertising, £1 spent on social media, £1 spent on price promotion, and so on. And if you are concerned that these things all interact then do not worry, the econometric modelling copes with any interactions as well.
So, if you have your sales data on the one hand, and data on the things that may have influenced sales on the other, then econometric modelling will crunch the numbers for you and quantify how much impact each influence has had on your final sales outcome.
Measuring the Long-Term benefit of Social Media
However, having a technique to tease apart the differing influences on sales is one thing, knowing exactly what data to feed into the model is quite another.
With social media having, (and potentially, having at one and the same time) a long-term “presence” benefit and the inclusion of sales-driving mechanics (like AR features, QR codes, website names, telephone numbers and other devices) to elicit immediate sales (or other relevant trackable) responses, we need the means to quantify the amount of each of these elements it has generated.
In her article for Marketing magazine some time ago, Nicola Kemp, head of features for Haymarket Publications, highlighted how data from ComScore revealed that visitors to news-media websites spend three times longer on those sites as visitors that get taken to them by links on social media. Moreover, she made the point that brands will only succeed in achieving tangible returns on their social media investments if they do not just divert attention but also then engage people effectively. Not rocket science, I hear you say, but what is the acid test of whether you are achieving it?
The immediate sales (or other behavioural) impacts can be directly tracked and so are known precisely. But in addition, estimates (or counts) of the number of people who have seen the piece are also known, in many cases quite accurately. These ‘numbers of views’ are then used in sales modelling in exactly the same way as data on the number of views of traditional media – and we’ve found that the way they affect sales is much the same.
So, the answer to evaluating the full benefit of your social media is to look at not just the immediately visible sales impacts, but also the “advertising effect” of social media. In other words, not to just measure communication delivery, but also the tangible impact – i.e. what happens to sales performance?
That is, and always has been, difficult when there and many multiple influences on sales, and today’s multi-channel, multi-device, multi-tasking world has made the situation ever more complex. As we mentioned earlier though, if you are able to track the inputs and communication outputs from social media – which now you can, both readily and cheaply – then you can use those variables as inputs to the econometric model to see if they correlate with actual sales performance net of all the other factors affecting sales at each point in time.
The expected time horizon for getting a payback is the key consideration.
So, how does that work exactly?
The real secret is about knowing the time horizon over which you are expecting a return on investment. Brand Building investment is usually defined as being spend that is designed to payback in the ‘long term’ – which typically means within 3-5 years – whilst Brand Activation should be considered as spend which needs to payback within the next 12 months.
Applying this distinction enables spend to be aligned with the purpose of the activity – creating awareness and changing attitudes but not necessarily with the objective of stimulating an immediate response, versus activities designed to drive sales or other tangible behaviours in the short term.
Like other media, social media can do either task – and often both – so, as with any other marketing activity, one of the first things to be clear about is whether your social media campaign is designed to primarily generate an immediate response or primarily generate awareness and positive perceptions about the brand, service or product being promoted.
The vast majority of social media activity is, we would contend, more about the latter than the former, but every activity is unique so you to consider your own situation and decide accordingly – with potentially some campaigns being mainly brand building and some being mainly brand activation.
Brand Activation campaigns are, of course, the easiest ones to take a judgement on. There would be a specific action or actions they are seeking to prompt, and we only need to count the number of actions taken to see what the payback is.
Judging the return from Brand Building campaigns is much more difficult, however, and that is why people have written many books on how to do it – most advocating counting something in much the same way as measuring short term activation.
Econometrics, though, is a much better type of measurement technique. It is one that is focussed on measuring the net outcomes and, as explained earlier, one which allocates importance objectively to each type of media and to each type of activation.
Binet & Field noted in their 2017 article that 60% of the award-winning entries to the IPA Effectiveness Awards programme supported their entry with evidence from econometric modelling of the effectiveness of their activities and, when done soundly and objectively, that evidence is reliable.
A Case Example from the Retail Industry
We have been working for some time now with various retailers and shopping centre owners using econometric modelling to quantify the impacts of differing media activities, including differing social media platforms. The table below summarises the findings, all of which have high levels of statistical confidence.
Whilst TV remains the most powerful medium in its own right, if you factor-in the cost of production you can see why TV revenues have been under pressure for a very long time. There is still a question, though, of sustainability. TV continues to support sales through the medium term. The decay rates of social media are virtually instantaneous. Unless you maintain a constant presence with a constant flow of new and engaging content the impact dies away.
Also, increasingly neglected these days, are the still influencing benefits of traditional media like Radio, Outdoor and Magazines. To be fully effective, campaigns need to be integrated, and present both offline and online. The synergies of Reach (from broadcast media) and Engagement (from social media) combine to generate action and lasting memories.
Naturally the results in the table above are not “generic” – they really relate only to the specific mix of campaigns and uses of social media for the clients concerned – but what they show is that you can move beyond measures such as “numbers of re-Tweets”, “Numbers of Impressions”, and so on, to get reliable estimates of the real financial benefit of social media to your business. And show that the value is more than just the value of any sales from the handful of people who clicked-through to your website and pushed the ‘buy’ button.
And they also align totally with the findings of the meta-data research projects conducted by Binet & Field on behalf of the IPA.
Quantify Your Own Benefit Today
If what we have outlined above sounds like it could be useful for your own marketing planning, then don’t hold back, do something about it today.