Subscribe

Celebrity influencers not always the best fit

Lauren Kate Rawlins
By Lauren Kate Rawlins, ITWeb digital and innovation contributor.
Johannesburg, 16 Mar 2018
Tech needs to be deployed to measure influence online, so that influencers with different audience sizes can compete equally.
Tech needs to be deployed to measure influence online, so that influencers with different audience sizes can compete equally.

The term 'influencers' crept into the lexicon of digital marketers a few years ago when they realised they could use people with large followings - normally celebrities, bloggers, or sports stars - on social media to promote their brand or product.

However, new research shows traditional 'influencers' may not always provide the best return for companies, and ordinary people with smaller, more concentrated followings could prove more useful to companies looking to take advantage of this digital marketing tool.

The study, "The Power of Brand Influencers", conducted by World Wide Worx in partnership with Continuon, looked at the social audiences of 50 South African brands on Facebook, Instagram and Twitter.

The research was done over three months by the two companies. They combed through over 100 million data points from 5.25 million unique engagers and 355 000 unique influencers.

Continuon is a Cape Town-based company which builds algorithms to analyse social networks, and connects the most suited influencer with brands.

One of the main insights from the research showed influencers can be ordinary people with small followings. It shows that while a post by a celebrity influencer may garner a lot of engagement by fans and followers, it might not equate to actual influence.

The problem is that brands measure the reach of both these types of influencers the same.

Bradley Elliott, founder of Continuon, says this is the fundamental flaw in influencer marketing: celebrities and micro-influencers are qualitatively assessed, but then quantitatively measured by their reach.

"Influencer marketing is gaining massive popularity and use among brands, because it works and yields results," says Elliott. "But the big learning from this massive research study is that the old metrics for measuring social success aren't good enough."

The research shows that technologies such as algorithms and artificial intelligence need to be employed by companies to measure the real reach of influencers with large and small audiences, so that actual reach is measured and not just how many people saw or commented on the social media post.

"Not all influencers are the same," says Elliott. "This means brands will have to employ smart technologies to segment social audiences, and dig deeper into audience categories and what they mean, to get better benefit from investments in social media.

"Brands have traditionally used what metrics are at their disposal to define performance. The current metrics used for measuring social success by brands is a limited view of the power of influence on social networks."

World Wide Worx MD Arthur Goldstuck says: "This is the first time South Africa's marketing and business communities have been let into the private workings of brands and their social communities, and what we're learning will upset the notions that marketers and brands have had about social media - how to use it, how to measure returns.

"For the first time, we are able to look across multiple categories, and at brands of different sizes and budgets, from an equal perspective.

"This not only gives us a much better view of the social media dynamics in this country, but addresses the massive disconnect between big budgets for social spend and loose measures of success. Marketers need to be more rigorous and accountable for their decisions, and knowing what 'good' looks like in terms of social media will help raise the profile of digital marketers in the boardroom," says Goldstuck.

Share