Influencers, image recognition and the future of research
Between all the rumbles about oversubscribed “silent disco”-style talks and calls to “come visit our stand!”, the biggest buzz at this year’s Marketing Week Live and Insight Show proved social must be a focus for marketers in 2018.
Talks from Tribe founder Jules Lund and Social Chain’s Oliver Yonchev got everyone’s minds a-flutter, while the theme of social listening and research ran throughout presentations from vendors as diverse as neuroscience mapping, insights platforms and behavioural economists.
But it was the aforementioned two companies that really had everyone talking. Perform a quick search of the official hashtag #MWL18 and you’ll be confronted with Social Chain’s cunning stunt - giving away shares in bitcoin via an old skill tester procured from Blackpool Pleasure Beach - but you’ll also see the proliferation of chatter around influencers. This continues to be THE topic in marketing.
— Vickie Graham (@Vickie_CIPP) March 8, 2018
Tribe enables brands to work with influencers and micro-influencers in an easy and intuitive way. Founder Jules Lund’s session on the results every marketer should demand from their influencer campaigns had people spilling far out of the stage and into surrounding stands and aisles. Lund told us that since Facebook changed its algorithm in January, advertising on the channel has increased 43% as brands struggle to be seen. This, he says, makes influencer marketing even more important.
“People-powered marketing will be a multi-million dollar industry in a few years,” Lund said. User-generated content is often cheaper and less time-consuming to get, and 6.9 times more engaging than brand content. Lund said influencer marketing will move from engagement to performance marketing as soon as Instagram rolls out their paid partnership feature globally. Variety, volume and creative formats become key, and with micro-influencers having higher engagement rates than celebrities, brands should be looking at how you can identify and engage them.
He's here for round 2 today, talking about results every marketer should demand from their influencer campaigns.
— T R I B E (@tribe) March 8, 2018
Elsewhere, Social Chain created a stir with a look at “How to win the hearts of the generation that expects it all”. Combining the youth and digital is a sure-fire way to get interest at an event like Marketing Week Live, and these guys know how to maximise that attention.
Yonchev’s answer to his question? “Make them feel.” “The truth is this: the act of making your audience feel goes beyond crushing a football with a hydraulic press or blowing up a vintage VW Beetle (both Social Chain live streams),” he writes this in a blog to accompany his talk. “It’s all about managing expectations. We do the same for our partners, clients and their CMOs – so why not our audiences?”
— CrowdControlHQ (@CrowdControlHQ) March 7, 2018
In 2018, three seconds is considered an attention span on social media, and counts as a video view, Yonchev told his audience - then reminded them a goldfish has an attention span three times that. How can you get attention in just three seconds?
But it wasn’t just social agencies banging the social drum; TUI’s Jeremy Ellis told his audience that 75% of marketers are allocating investment to influencer marketing, while Sarah Cunningham from AdRoll reminded everyone that you can pay for likes, but you can’t buy love, so you need to get your advocates talking about you for true social ROI.
— Safeera Sarjoo (@SafeeraSarjoo) March 8, 2018
The image as social data
Talking about true insights and true ROI, multiple talks over at the Insight Show talked about the emergence of image recognition as an essential social insight. How we analyse social conversations has moved beyond pure text; we can now look at images, down to logo and facial recognition, which is bringing new possibilities for insights teams. But the growth of image recognition does come with one caveat: the need to consider context. Zoning in on one aspect of an image can ignore a whole story in the background.
The use cases for image recognition continue to grow: brand health, marketing optimisation, revenue generation, operational efficiency, customer experience, innovation. The potential is as limitless as the images we create and share; one speaker asserted images will be the most important data sources of the next five years. Just remember that context really is key - both in image recognition, and in the language used. As @ResearchMediaLS puts it on Twitter: Content and context go hand in hand.
To get that context question clear, smart marketers will look carefully at the questions they ask of social insights. Once data is seen in context, the story emerges more clearly. And once that story emerges, you should go back and ask further questions, listen more, go deeper. Dig into the underlying reasons behind the answers you see in your social insights, because what people say they do and what they actually do don’t always match up. Social insights can help us to see what’s really going on - to see what people are doing when they’re not being directly asked or watched.
Technology + curiosity = future of research
So where does that leave the future of research? A panel session on day two asked that very question. In a tribute to International Women’s Day, the panel was all young females, and they talked about the changes they’re seeing in market research. Clients are increasingly looking to ROI; there is a lot of change, a need for greater speed, and cost pressures mean researchers must deliver value.
That idea of going deeper, or asking the uncomfortable questions, and not taking things at face value came up here again. Technology plus curiosity is informing the researchers of tomorrow. Today’s market research project can have its results triangulated with social insights to prove or disprove a theory.
There will still always be a need for some human interaction and creativity to interpret and analyse results. But, once again, the need to be aware of context and culture when analysing your results was the key takeaway. Language is data; consider how language and discussion change over time.