As social media research matures, the big question on everyone’s lips is “How can we connect this to other data?” More particularly, how can we connect it to what really matters to our business: sales?
Last week I gave a webinar on exactly this topic, sharing the results of our research study mapping social media buzz for 3 music events against ticket sales. You can find that presentation here on Slideshare, and download the full recording from this link.
In this blog, I want to put our work in context and map the wider industry thinking on this issue by summarising 5 other key social-to-sales research studies. There are a number of different ways that social media activity and sales can be compared, and I hope it’s useful to provide a summary and outline some of the key differences:
Presenting at the Advertising Research Foundation’s Re:Think 2013 conference, Coke’s Eric Schmidt reported that “We didn’t see any statistically significant relationship between our buzz and our short-term sales.” (AdAge.com)
Note that Coke are still big believers in social media’s effectiveness as part of an integrated campaign: said Wendy Clark, “It’s the combination of owned, earned, shared and paid media connections – with social playing a crucial role at the heart of our activations – that creates marketplace impact, consumer engagement, brand love and brand value.”
[image created by Coca Cola, via iEdge.eu]
But this study is evidence that overall social media buzz – the number of brand mentions – doesn’t necessarily correlate with sales. What might?
“The consulting firm initially couldn’t find any connection between social-media buzz and sales, either when looking at overall data changes or even by applying an algorithm to assign sentiment to the buzz. But McKinsey found the relationship between negative buzz and a decline in sales when it “hand tabulated” sentiment in social-media comments.” (AdAge.com)
This negative sentiment hurt signups by 8%, “offsetting their entire TV spend,” McKinsey principal Jonathan Gordan said at the Advertising Research Foundations Audience Measurement 8.0 conference in New York. Why? Because the negativity was primarily driven by complaints about the sign-up process and call-centre workers at the telecom provider.
This shows how a relationship between social and sales can become visible when you drill down into more specific aspects of social data. Brand volumes didn’t impact telecoms sign-ups - but complaints about the sign-up process did.
A different metric here – not social media volumes (aka the number of messages mentioning a brand), but the number of shares:
“The company says that Facebook is the king of all social networks when it comes to ticket sales. In the UK, if a person shares an event on Facebook, it generates an average of £2.25 ($3.60) in additional gross ticket sales. A share on Twitter, meanwhile, drives an average of £1.80 ($2.90), and an event shared on LinkedIn generates an average of £1.24 ($1.99) in additional event revenue.” (TechCrunch.com)
Eventbrite’s reason for why Facebook is bringing in more sales is good sense: “The connections we have on Facebook most closely represent the people we actually know and spend time with offline,” its researchers write.
4. Why Twitter Buzz ≠ Movie Ticket Sales
“140 Proof looked at 25 major Hollywood films released in 2012, compiling data on each movie’s social media activity (mentions and hashtags) two weeks before, and two weeks after the release. It found that the number of overall Twitter mentions is a poor predictor of box office sales (unlike tweet volume and TV ratings). What did correlate to box office success was the number of tweets from influential tastemakers” (Readwrite.com)
Again, the relationship between social and sales doesn’t show up when you just look at raw volumes – but it is still there. Pulsar’s range of influencer metrics such as visibility and Klout filters can enable deeper analysis of how influence relates to sales, going beyond “number of tweets from tastemakers” to understanding how influence levels and sales-power scales.
A totally different methodology – they’re not mapping activity in social media, or measuring clickthroughs from social channels, but rather surveying 5,657 people asking them to report whether they’d ever bought anything they’d seen on Twitter, Facebook or Pinterest.
This is worth doing because, “68% of Facebook users are “lurkers” who post only rarely, so the influence of social on their purchasing will not be visible from social media analytics alone.” It’s a good reminder to think about social media users as much as an audience as content-creators - and that the path to purchase is more complex than old-fashioned sales funnel models, or simple ‘last-click’ attribution.
[image via Digital Information World]
Five studies, two key take-aways for understanding how social and sales connect:
1. Think about your user journey. How do people make a decision to buy your product – who or what might influence them? How do people consume your product – is it particularly social, like something you would want to do with friends, or something worth boasting and sharing on Twitter and Facebook? Is it something people can purchase quickly online, or a more considered purchase?
2. Think about what aspect of social media to measure. It may not be simple volumes of brand messages that correlate with sales, but something more specific – such as influence, sentiment, or specific topics. Or perhaps it’s not messages at all but behaviours such as sharing. Most of all, remember to measure all of social media, not just owned channel activity: you’re looking for consumer behaviour, not just reactions to your own!
As these studies from a diverse range of brands show, social media does often connect to sales – not all of the time, but often with some statistical smarts & a deep knowledge of social, a link can be found.
Note that we’re saying “connect”, not “cause” - correlation can be assessed using relatively simple stats such as R-squared tests, but unpicking causation (Was it social media activity that made someone buy, or a price promotion, or TV advertising?) is a challenge for regression analysis and a bigger topic than we can discuss here.
And sometimes the relationship between social and sales can go both ways – not only “I buy a concert ticket because I saw the news on Twitter”, but also “I bought a concert ticket for my favourite band and I’m so excited, I want to tell everybody!” Perhaps brands can even hope for a virtuous circle of social driving sales, which drives further social activity, which drives even more sales… Fingers crossed!