Measuring social media ROI has always been a tangled web and, consequently, a hotly debated issue.
As one CEO told Our Social Times recently, “When I’m told that ‘we got 1.4M social impressions’, what does it mean? I know what it is, but I don’t know how it’s relevant. I have no clue what it means to me.”
It’s not hard to see why this view continues to linger. PPC (Pay Per Click) and email ROI can be accessed and measured in an instant, but the impact of social activity is a comparative enigma. How did your brand’s last Instagram post effect the bottom line? How many sales did that Facebook Live product launch generate?
It takes an expert in the field to introduce some clarity into such a thorny issue. Marshall Sponder is a specialist in social media measurement, a prolific speaker, blogger and consultant. He has also written two books, Social Media Analytics: Effective Tools for Building, Interpreting, and Using Metrics, and the forthcoming Digital Analytics for Marketing.
How do we measure social media ROI?
Many CEOs have become more sophisticated in their understanding of what social media is, and the kind of outcomes it can produce. I believe they accepted the idea that ROI calculations are impossible to make based on social media (unless we’re talking about a social advertising campaign).
It’s questionable if one should bother to create an ROI calculation from social media alone. There are many who have tried to do this in the past and even now I am sure many still are. But the value of social media isn’t monetary in nature unless, as I said above, we’re talking about advertising campaigns – where it’s easier to see cause and effect (assigning a value to one’s effort and spend against the return).
However, web analytics offers a way to measure most of the marketing impact of any actions taking place on a website. The fly in the ointment is that most social media happens on social media channels, not on the stakeholder’s website.
As a result, most CEOs now understand that they have an omnichannel marketing measurement problem to solve, rather than a social media ROI issue – this is what has changed in the six years since my first book was published in 2011.
Social media data is too fragmented to collect in a comprehensive enough way to make it easy or reasonable to calculate ROI metrics. This problem arose because each network decided to make themselves a ‘walled garden’ because they need to monetise their user data.
Suppose Facebook and LinkedIn were to share their user information in an inter-operable fashion with Google (who actually tried to get them to do this). The stock value of these companies would drastically drop and they would go out of business within a year or two.
What we have learned in 2017 is that we live in a world of mixed realities that are antagonistic to actionable analytics, and are producing the problems we are trying to overcome.
As data consumers we want a unified data set, but we’re also afraid of the privacy issues that it would pose. At the same time the platforms depend on walled gardens to protect their own value and IP and are unwilling to give that up.
The result is a stalemate that forces marketers to come up with their own unique solutions for every single complex implementation – this is the price we pay for living in a world with competing needs that are at war with each other. We cannot achieve a unified vision of data without making the result commercially unworkable – at least, not in the current world we live in.
The proposed alternative is the ‘blockchain‘, where users are rewarded for actions and engagement with real or virtual currency. In theory this would create a level playing field for data analysts. In reality, though, this is an altruistic pipedream that is unlikely to materialise any time soon, except in small, isolated applications where it will have limited impact.
For many who are unfamiliar with the blockchain, it’s both a protocol and hardware/firmware/software. I first heard about it two years ago from a guest speaker in one the classes I teach at Baruch College. Until recently the blockchain seemed to be an abstract concept, but people are now beginning to understand what it is, and perhaps what it might be useful for.
Were the blockchain to be viable, it would be the death of Facebook, LinkedIn, Twitter and virtually every private fortune that was ever made in social media. The operating systems and walled gardens would have to drop, everything would end up being an application running on the blockchain, run and governed by someone else.
In that world you could have unified analytics but the cost would be nuclear – all business as we know it would be altered. I say it’s better to live with the devil you know than the one you don’t. Consequently, as marketers, we must accept that we live in a world where the data will remain incomplete, a taste of that is shown in Image 5, below.
Each social media channel has its own data analytics layer and constructs, along with its own metrics (which are usually incompatible with any other platform’s data and it is impossible to build a consistent, logical ROI calculation from it).
Another reality that CEOs are now beginning to accept is that social media analytics is a different animal to business analytics.
Most IT departments are used to treating all their data as business data and, at first, it was thought that social media could be an extension of the rest of the customer and house data they had.
But once anyone tries to work with social data it turns into a swamp. The more work you put into it the more work is required. Every question answered ends up leading to 10 more, making it very hard to create a business product out of it.
Social media data is precious but it’s mostly unstructured. Instead of being likened to big fields of oil, ready to be processed, it turned out to be closer to gold nuggets mined in the California deserts of the 1850s.
The only companies that managed to harness social data into a product proved to be our biggest corporations. Because all this data is being created by the world, it takes a ‘big data’ approach to mine it.
CEOs have begun to figure out how to deal with that; algorithms are coming of age, programmatic advertising and targeting are now usable and extensible. At the end of the day, because social media data is unstructured, it is by its very nature antagonistic to a finished business product or ROI.
I’ve no doubt brands will continue to capture and analyse social data in the search for ROI, but in my view the real value of social media isn’t in the data but in its transformation into structure. That is where the gold lies today.