There’s been a bit of discussion lately about influencer marketing – which is essentially the equivalent of celebrity endorsements, except the people involved usually aren’t popular enough to be considered celebrities.
The core premise is this:
- Brand wants to sell stuff
- Some people have big follower or subscriber numbers
- Brand contacts person, asks the person to post sponsored content
These ‘people with big numbers’ can be anything from someone with 5,000 or 10,000 followers through to ‘instaCelebs’ or someone ‘known for creating quirky online videos‘ with 2.5 million subscribers on YouTube.
Here comes the problem though. I haven’t mentioned ‘number of people following’ – this would imply that these followers are people, eager to consume the influencer’s quirky content.
Often, they are not.
Whether it’s on Twitter, Instagram, YouTube or anywhere else, not-quite-real followers abound. Some networks are better than others of course – in the same way that some influencers are actually influential, much the same way that some highly influential people have only 500 followers on Twitter. It all depends on how you define influence, but that’s another matter.
Let’s take the example of Twitter to start with. I would hazard a guess that 30-40% of the followers I have are businesses or brands. They don’t see my Tweets – they just pump stuff out, and I would be surprised if the person or agency actually ever browsed the brand’s home feed. Maybe some targeted lists, or a keyword search for the brand name, but not a general feed. Why would you?
Now I don’t know whether that number is real, but it’s easy to pick these types of accounts up. Automated or not, you just start following them, see if they “follow back” – often they will, because they will have automated follow-back systems in place. You could just as easily go to fiverr.com and buy 10,000 Twitter followers and you’d probably end up with a net boost of 5,000 after Twitter shuts half of them down in a year’s time.
You’re pitching one automated action against another, and everyone ‘wins’ – the social media person gets a raise, you feel better for having ‘more followers’, and the world spins round and round.
And ironically, people will then say ‘oh look, they’re a top social media influencer!!11!!’. Without realising that influence is about being able to influence: to have an effect on something or someone else. But then you can go to a conference and tell everyone you have 50,000 Twitter followers, even if nobody ever engages with your Tweets, talks to you, or even sees your stuff. In fact, a company that runs Instagram bots (fake “people”) claims that 85% of their clients are females aged 14-20, perhaps the next wave of influencers.
But that’s not what a brand wants from an influencer. They want eyeballs, they want actions. But that’s how they have to buy. A brand barely knows whether someone’s 50,000 Instagram followers are real or not – even checking the number of likes they get on photos is a farce (that’s so easy to automate it’s not funny), or comments (moderately harder without sounding obviously like a machine to automate). And given Instagram has a) no ability for people to ‘click through’ directly to a website (unless you’re running an Instagram ad) and b) barely any metrics whatsoever, there’s no choice.
In fact, Neil Patel showed how all this works in a blog post, What spending $57,000 on Instagram taught me (be warned: most of the content is highly chauvinistic – just shows you the kind of person who sells courses to ‘double traffic in 30 days’). The post demonstrates the industry that has popped up: try and generate a following on Instagram, ‘flip’ the account by selling it to someone, who will then delete all the content (usually memes and jokes trending on Reddit), and start pushing their own agenda. Neil gets a ‘UFC ring girl’ to promote stuff he’s giving away for free to get more people to mindlessly follow (regardless of whether they would ever purchase one of his products or services), or promote stuff by doing things like this:
For example, I may do a video Skype session with a female model, create a screenshot of it, and have her post it on her Instagram profile, with a mention of me giving her business or marketing advice.
Good on him, but this shows the farce: many of these followers are likely to not have anything to do with his brand, and secondly, attracting only “sweepstakers” is never going to be a winning strategy. I can’t think of any brand that actively wants to market to sweepstakers, except perhaps RetailMeNot or OzBargain. Sweepstakers are the kind of market who have almost no brand loyalty, and just chase a bargain. But they are great for pumping up your engagement numbers, because they all flood in for that one promotion, even if they’ll never see your content or use what they win.
An interesting post to read on this was from Hunter Scott who wrote a script to automatically enter Twitter sweepstakes – the kind of “RT this to win” / “Follow us to win” promotion. While he only won 1,000 out of 165,000 of these, it just shows you how easy it is to pump out this kind of rubbish that is measured as “great engagement results” on the other side.
So back to influencer marketing.
The problem here is that influencer marketing is sold as some kind of way to access the eyeballs who follow the person. Now, some will get angry, as most countries require disclosure the content is advertised, so they claim the influencer is ‘selling out’. Others will keep scrolling, some will certainly read/comment, and perhaps some might even buy.
But that effectively means you’re just doing display advertising, and leveraging it poorly by having little to no measurement in place.
You may as well run display ads, or for that matter ads on the platform the influencer is going to promote you on. At least with the ads, you get a vaguely robust reporting solution, even if a huge number of ad clicks are fake anyway – at least you get some kind of outcome, and better metrics than “Sue has 50,000 Instagram followers so if we get her to post 3 things about us we’ll reach 150,000 people!”.
The announcement that Instagram was rolling out its ads API to some of the biggest (read: most costly) social media monitoring tools is telling, and I think it will really stuff up many of these ‘influencers’ gigs. No longer will they be able to demand insane amounts of money for posting vaguely relevant stuff – the brand will just run ads instead, without Instagram’s previous $30k monthly minimum spend to do so (or so I have heard – it’s certainly a big number). They might dabble a bit in influencers, but (I hope) nobody is stupid enough to put much money into it – the return is hugely questionable.
For all of the hype you hear about how digital advertising is killing TV advertising, the big change that is coming (please, please, please let it come quickly, savvy businesses are already there) is marketers focussing on results. Whether the results come from a TV ad, a social media ad, a PR campaign or sponsoring a local art fair makes no difference. The point is that the return needs to be there. It’s getting so much easier to measure return and marketers are starting to realise the fallacy of last-click attribution models (your TV ad says go to your website –> they go to your website when they see your Facebook ad, thinking “oh yeah, I wanted to buy that dress” –> you attribute the sale to the Facebook ad) – they don’t work, because few (bigger, and increasingly smaller too) brands can isolate marketing that well. Multichannel attribution is a legitimate statistical discipline, it just costs money.
Ultimately, if you’re not willing to spend the money to identify what works and what doesn’t, you aren’t going to be in business too much longer.
If you want to try influencer marketing, do. But do it in a way that is clearly measurable to business results. Not to Likes, not to engagement, not to #hashtag #mentions, to delivering results. Anything else is just hot air, something most ‘influencers’ have enough of already.