September 3, 2024

The importance of vetting influencers

Here's how AI can help save money (and future headaches).
Communications
TABLE OF CONTENTS

Today, as traditional marketing shifts increasingly toward the creator economy, vetting content creators and influencers is more vital than ever.

Consider the case of Vodka Beluga, a spirits brand that began researching influencers for an upcoming campaign in 2021.

Vodka Beluga carried out what seemed like solid due diligence. 

But then it took one extra step that proved pivotal. It used the SaaS platform Influencer Marketing.AI (IMAI) to evaluate the list of names it had come up with—only to find that 90% of them had followings that were overwhelmingly fake.

“Thanks to IMAI’s thorough data analysis, we were able to prevent Vodka Beluga from wasting its marketing budget and potentially damaging its brand reputation by associating with fraudulent influencers,” says Eran Nizri, co-founder and CEO of IMAI (recently acquired by Stagwell, and now incorporated into Stagwell’s PRophet Comms Tech product suite).

“It's an easy trap for brands to fall into if they don't thoroughly vet influencers,” Nizri adds, “as superficial metrics like follower counts can be misleading without deeper analysis of engagement quality and authenticity.”

Don’t freak out, but a lot can go wrong

It’s not just fake followers that brands have to worry about.

When brands engage with content creators, the results can be unpredictable. A company enlists an influencer only for an old, racist tweet to resurface; or a content creator makes money with sponsored content, but is routinely mocked for it by their followers in the comments.

Platforms like PRophet’s Influence can help prevent public-relations crises—for instance, using AI to comprehensively scan a creator’s past years of content and surface any objectionable, red-flag moments. 

But the tech can also ensure that creators and their content deliver the most bang for a brand’s buck in other critical ways.

Considering engagement quality

All engagement with a brand is not created equally.

Say a creator posts the agreed-upon content, but it doesn’t seem to be getting the hoped-for traction. How does a brand know what went wrong?

One way is to use PRophet’s AI-powered Branded Engagement feature that evaluates all the comments on an influencer’s sponsored post. It then assigns a positive, neutral or negative sentiment to each—and can even detect jargon and slang in multiple languages.

“Branded Engagement will give a brand an immediate indication of audience reaction to the content, and if the audience reaction is highly positive, then they can decide to put money behind it and boost that post. If it's negative, then they can decide not to boost it or adjust the content accordingly,” says PRophet CMO Jason Brandt.

Leveraging unique access to an audience

PRophet can also determine if the content is reaching the right audience in terms of geography and demographics, and if it’s getting appropriate numbers of shares and saves. It can even issue a report card on imagery, both static and moving.

“The great thing about social media content creators for brands is the audience it opens up,” Brandt says. 

“Audiences choose content creators they want to be associated with and receive information from—often fun and informative content, sometimes complicated and sensitive. It’s a pure ‘pull’ relationship based on respect and trust in a creator's unique voice - very few media channels work that way. Our platform respects this relationship and excels at ensuring audience quality and brand’s briefs take audience expectations into consideration.”

Meanwhile, PRophet has developed a feature to help brands understand which influencers convert most effectively.

For example, it has been shown that influencers who post at least 10 stories a day tend to convert more business—a metric that the feature tracks, helping brands identify the most sales-effective influencers.

The shift toward influencer marketing

All of this technology comes as the industry finds itself relying more and more on social-media content creators as the voices of brands.

Brandt attributes the shift to a long-simmering distrust of big business and major brands, which has led consumers to seek information from elsewhere—often individuals they trust or admire (or who at least make them laugh).

“As a consumer, I'm going to decide on my terms who I want to follow, who I want to receive information from, who I want to take a moment out of my day and stop and view a piece of content that served to me,” Brandt says of these audiences. “That's why so much money is now being shifted from traditional marketing spend to putting money behind social media content creators.”

 Contracting with the right influencer can get trickier when the content creator is smaller-scale—those who may have less formidable follower counts, but a more targeted reach. 

Increasingly, it’s becoming important for brands to incorporate these nano- and micro-influencers—even those with as few as 1,000 followers—into their portfolios, Brandt says.

“Are this person's audience numbers growing or declining? Is the content quality up to the standard that I need to represent my brand?” he says. “Or am I willing to lower my expectations about content quality knowing that the creator has an audience I really want to get to?”

The role of tech and GenAI

Since launching, both PRophet and IMAI have evolved technologies to power vetting algorithms.

The latter developed its tool by purchasing fake followers in earnest, and then analyzing this assortment of bot accounts to identify common characteristics and tip-offs: a lack of profile pictures, inconsistent content, and a suspicious ratio of ‘accounts followed’ to ‘followers.’ 

“These patterns helped us define what fake followers typically look like, enabling our algorithm to detect and flag such accounts accurately,” Nizri says.

PRophet, meanwhile, partly based its algorithm on technology used during the 2020 election to detect fake Facebook accounts from foreign countries.

It modified these foundations to detect a wide variety of fraudulent followers, bots, and dormant profiles.

Ultimately, PRophet’s ability to determine fraudulent followings has proved not just crucial to ensuring a brand reaches its desired audience—but it’s also provided a powerful negotiating tool.

A PRophet credibility score that finds, say, only 75% of an influencer’s one-million-strong following is legit, doesn’t necessarily mean a brand shouldn't engage the creator. But it does mean that it can negotiate “on the 750,000 number, not the million,” Brandt explains.

An end-to-end influencer marketing solution

While vetting is key, Brandt stresses that this is just one step of a complex process—one which PRophet Influence is designed to support from end to end.

The platform “is really a one-stop shop” that can help brands put together media plans, send contracts to influencers, deliver their briefs, and help them create the content—and, of course, evaluate the results afterward. 

Returning to the example of Vodka Beluga: Thanks to IMAI, the campaign ultimately engaged 11 influencers and achieved an audience reach of 6.3 million—doubling its initial estimate. 

We’ll drink to that.

Stagwell Marketing Cloud

Take five minutes to elevate your marketing POV
Twice monthly, get the latest from Into the Cloud in your inbox.
Related articles
How QuestDIY uses AI to simplify survey writing
Consider it a drafting & research assistant that never sleeps.
Research
10 mistakes to avoid when drafting your market research survey
From leading questions to inconsistent answer scales, here’s what to keep in mind.
Research