Interview: Exploring Data Science at Brandwatch with Hamish Morgan
By Olivia SwainAug 23
Published August 31st 2016
If you work with social data, you’ll likely have a frustrating background exposure to those that just don’t get it, for want of a better phrase.
There remains something of a social and professional stigma around social media in general. It’s a discipline that’s still not managed to shake off the suspicion from some that it’s all just a bit … fluffy.
Whether it’s explaining your role to executives or relatives, people in the industry are weary of having to justify the merit of their projects and for many organizations it can feel like a constant battle for resource and vindication.
That’s not to say things are bad. It’s been a long road from isolated social media managers ‘sitting on Facebook all day’, and those working with social data now enjoy larger budgets and team sizes as the field grows ever more sophisticated and valuable to the operations of a business – but it can still seem like an uphill struggle for many.
A social data leader, depending on circumstance and organization type, can include anyone from social media managers to CMOs. A common challenge sticks out among these leaders and that’s the very nature of proving the worth of their programs. It’s the struggle to establish internal credibility, gain recognition of successes with social, and about securing resource for future initiatives.
Wrapped up in this topic is the issue of demonstrating return on investment. Proving ROI holds the keys to unlock the justification challenge – especially as social continues to carve out more budget against traditional forms of media.
There is plenty of broad data that demonstrates the value of many social endeavors in general. It’s a smattering of micro case studies, but both Unilever and Coca-Cola have publically stated that they have observed a direct relationship between social activity and sales.
The trouble is that these case studies mean nothing when faced with a disapproving executive. Management teams aren’t typically interested in such generalizations, and walking this path is unlikely to get many social leaders very far in achieving their objectives. So what else can be done?
Many are moving towards using social to represent the value of other programs, and are extending the tendrils of social data across their organizations.
Measuring the performance of, say, a TV campaign or product launch, is a means of associating social data with something bigger than just social media marketing. It also distinguishes the value of social programs as something inherently and continuously useful, not just something tied to the success of a social campaign.
Whatever your ultimate goal and specific strategy for getting management buy-in for your social initiatives, there are a number of techniques that can be employed to convince and convert those at the top. These principles can provide a helpful process to follow for those seeking to get a bigger chunk of attention from the individuals that matter.
Setting out the expectations as early as possible with social data projects is critical. Very often the cause of friction between management and social teams rests upon a disparity between what is expected and what is delivered.
Take ownership of this from the start. Make the terms and desired results of the work as clear as possible. Ensure they are documented and agreed upon by the relevant stakeholders, as it can act as a stabilising point of reference as further decisions and reflections are considered in the future.
Ask yourself – what do the management expect from this? Is what she wants from the investment the same as what you want? Don’t do anything until you’ve spent some serious tame making sure that those expectations are aligned.
A vital part of this process includes listening and making sure the goals of the social leader meet the goals of the organization overall. When setting out the objectives for a particular initiative, the most successful leaders will spend time understanding what executives actually want – vital information when planning the program.
Listening will inform the process and provide the justification for why then we need to expand the analytics, why we need to engage, why we need to be there.
A common problem in this area is that, like all other disciplines inside an organization, social data is inherently flawed. There are drawbacks to social, and fundamentally it’s important to acknowledge the parts that aren’t going to work.
Dark data and the unseen offline effects of social will never find their way into monthly reports, yet social analytics teams regularly find themselves forced to defend the integrity of their data. The answer here is to confront those weaknesses, those biases and blind spots so closely entwined with social media, and present them as valid caveats to management before projects are undertaken and deliverables agreed.
You can use all the quantitative data you can get, but you still have to distrust it and use your own intelligence and judgment.
This mind set should continue as programs are rolled out and enacted. Admit failed endeavours and accept shortcomings as they happen.
While agencies may feel the need to desperately defend the success of their campaigns, and will seek to distort the data in such a way to paint the most positive picture they can, someone that is brand-side will be in a position to assess each piece of work in a more honest way.
Once again, exposing the real disappointments as they happen will serve to create much more robust foundations for the moments when you really need to underline the value of your work.
Quite soon after establishing the purpose of a project or investment comes the inevitable demands for proof. Proof that it was a good idea. Proof that it delivers results.
Cintas are probably the largest organization you’ve never heard of. With over 30,000 employees, Cintas are behind the unmistakable employee uniforms of brands as instantly recognisable as Starbucks and UPS. Troy Pfeffer started the enterprise’s intelligence department eight years ago, and discovered that achieving executive buy-in was the biggest obstacle he faced in making the program a success.
“Establish a new program with no precedent immediately puts you in risky territory. Winning the trust of the executive team was essential if I wanted this thing to work.”
Typically it’s from the very first moments after the initial sign off takes place when pressure begins to mount on those proof points. Troy believes this is one of the most crucial periods in any project involving new budgets or resource: “Until you do anything worthwhile, the jury is very much out. Your first priority should be to start delivering value as soon as you can, and not to get distracted by the big payoff in a few years’ time”.
Long-term decision-making should, of course, not go neglected, but there is genuine merit in the idea behind front-loading your communication of results.
With any project, there is likely to be an air of apprehension that can be swiftly blown away by some real, value-laden results.
Look for low-hanging fruits and go after those first, focusing on the elements that are easier to demonstrate the value of. With social data, a key part of this will be to distance the results from socially native metrics, such as followers and mention counts, and move towards universal business language such as sales, retention or net promoter score.
Compartmentalization can be an effective strategy. Rather than attempt to justify the entire social operation, sometimes isolating the ROI for each project, or even hire, can make it easier to communicate the value upwards.
This is echoed by comments from the VP of Analytics at a popular teleshopping retail brand in the US, who says that she isolated the specific benefits associated with new hires not in terms of input, but in output.
Speaking on the topic of weighing in on securing new headcount, she revealed she was able to add people to her team one person at a time in a piecemeal fashion – with associated increases in committed metrics for each new person added to the team.
We were looking at two potential new roles. For me, I knew they were going to make a difference, but I had to get management to see that too. In the end, I wrote down all the ways in which these people would affect the team’s performance – not just in the amount of work we could get done, but in the results of the department overall. And I committed to those results. That was enough to seal it.
This partitioned concept can be taken further. One of the most potent ways in which social resources are appropriated is through innovation programs or, more specifically, pilots. As a disruptive and novel discipline in general, social lends itself well to small-scale, focused pilot schemes to explore the value of new ideas.
Senior management like them because they are cost-efficient and can be side-lined relatively cheaply if need be, helping risk-averse executives feel more comfortable with what may be seen as a risky investment. See hesitant, limited backing of projects as opportunities for trial and experimentation. Then, when projects go well, it’s easier to build a case for evolving them into something more serious, well armed with data and first-hand practice.
Speaking to Tracy Bell, Enterprise Media Monitoring Executive at Bank of America, she revealed that in the early stages of developing a social intelligence program, social data leaders should target forward thinkers that will be receptive to social insights. As she points out, they should “go after business lines that will actually take actions on it.”
After social media intelligence has proven its ability to inform business decisions, it’s much easier to carry the torch for social and build a case around future insights.
As well as simply proving the worth of your initiatives, it should be considered best practice to maintain constant communication with project sponsors. Having secured early wins and starting to build more trust among senior management, social leaders should develop strategies for disseminating updates about the work in general.
Momentum can be a powerful force, and for meaningful long-term success it is important to occupy as much of those sponsors’ headspace as possible.
A tale about fighting for this headspace comes from Anthony, leader of social analytics at one of the big west coast technology firms.
Speaking to Anthony at an intelligence event in Madrid in 2015, he revealed he initially found it frustrating that his team’s monthly reports were being sent to board members yet were so often left unopened in inboxes or otherwise not acted upon. He needed to change something.
Recognizing the need to grab the attention of those with little spare to give, Anthony recruited the help of an internal designer to style the reports in a more visually engaging way. Fewer words, more images.
He also took the time to prepare highlight emails for those still unwilling to open the PDF itself, which included bold callouts of what he likes to call ‘key insights’.
Anthony jokes that his seniors didn’t respond well when they were named ‘recommended actions’, and these suggestions ended up landing in the wrong way. So he changed the name to ‘key insights’, and, somewhat amusingly, they are now received much more positively.
Many organizations are providing guidance to data scientists on how to make results easier for managers to digest: storytelling, for example, is now becoming a standard part of analytics training curricula.
Things had improved, but something was still missing.
Anthony spoke privately to some stakeholders on the project and a few colleagues close to his seniors to understand more about what might be lacking. He soon discovered that there was a perception among those above him that his reports didn’t have much in the way of credibility – they were packaged in a way that relied on the reader’s faith that the insights within were true.
Anthony again adjusted quickly, and began attaching raw data files along with the highlights email and report. These attachments included the verbatim comments from real people gathered by social listening tools, and exposed the data behind all the calculations and graphs featured in the report.
It felt a bit like data overload, but it didn’t take too long for this to make a noticeable difference to the reaction – and adoption – of the ‘key insights’ generated by his team.
Anthony’s experiences point to two considerations that often seem in conflict with one another. Social leaders must strike a balance between distributing coherent, digestible findings and still maintain the trust and credibility that comes with data-driven decision making.
Social media data, initially unstructured, can be difficult for those unfamiliar with it to understand. Before a social data leader can expect social insights to actually inform business decisions, “creating a common language across the organization” and “educating the consumers of data” is a valuable prerequisite, according to Bank of America’s Tracy Bell.
Troy Pfeffer at Cintas puts it nicely with his heavy repetition of the phrase “what’s in it for me?”
It’s a phrase he regularly drills into anyone responsible for sharing findings with other people, and has been shortened among his team as WIIFM. Constantly framing the WIIFM factor for the recipient of any distributed intelligence can fundamentally change the way insights are shared.
Creating such reports with what may (or may not) be in worthwhile for the recipient is an incredibly helpful method for making sure work gets noticed.
Perhaps it’s this phrase that unifies the many techniques that get talked about in this area. Considering the WIIFM factor is a helpful reminder to bear in mind across any activity you might be undertaking in generating more credibility for your program.