Social media influence is a fallacy
Following yesterday’s opinion piece on the power of social media ambassadors yesterday Eaon Pritchard argues the idea of influencers is actually flawed.
The idea that brands can pick out and target a small group of social media users with large ‘followings’ and then imagine that they will reach everyone else with their message is still prevalent however this influencer theory is a myth and its protagonists have got things the wrong way round.
There are a couple of reasons marketers still like to believe in this idea of the ‘influencers’.
Firstly, a little bit of laziness. It’s a lot easier to believe that a message can spread by the brand tapping apparently popular individuals – those special few to whom we all turn to in order to make decisions as Gladwellian rhetoric would have it – rather than get down with the messy business of continually reaching a mass of distracted, disinterested consumers.
Secondly, just by implementing these ‘influencer’ strategies it’s actually the brands themselves who appear to be the ultra influentials!
Ka-chow!
They, after all, are now the ones who influence the influencers.
Sadly neither of these things are true.
If they were our jobs as advertisers would be so much easier and predictable.
What is true, is that you’re just as likely to spread a message or product by targeting a mass market of random consumers as you would by going after so-called influencers, as long as the conditions are right.
If people are ready to adopt a product, message or trend, then just about anybody can start one, but if the conditions aren’t right, then no one can.
Indeed, most of what we should call real influence is much more accidental and principally involves easily influenced people influencing other easily influenced people, without either party being particularly cognisant of the influence.
There’s bad news and good news.
The bad news is that the specific conditions in which any given trend might emerge are very hard to predict and success only looks like success in hindsight.
The good news is that the psychology literature explains the general conditions for copying behaviour pretty well.
All day long people unconsciously mimic the behaviours of others they interact with, including facial expressions, accents, postures, gestures, mannerisms and emotions.
And the simple act of observing others’ behaviour can induce behavioural mimicry, particularly the behaviour of others who appear similar to us, and all of the above are unconscious automatic processes.
Likewise, simply observing others’ choices induces choice mimicry – just like behavioural mimicry it occurs automatically – and collectively when we are uncertain about which behaviours or choices are acceptable or accurate, then we use the ‘social proof ‘ heuristic to be on the safe side.
Or in more simple terms, ordinary people copy other ordinary people without really noticing they are doing it.
Speaking of hindsight, we’ve never held much truck with the old Gladwell ‘Hush Puppies’ story.
The legend goes along these lines; some East Village hipsters began wearing Hush Puppies in 1994 and then suddenly everyone else started wearing them, too.
What Gladwell failed to notice is that Hush Puppies were a staple of just about every UK subculture from the early sixties onwards, worn by mods, skins, hippies, punks, soul-boys and ravers right through to 3rd generation mod brit-poppers in ermm.. about 1994.
Even if Gladwell’s theory were true, it still doesn’t mean that if East Village hipsters did wear a specific product then it would automatically be popular.
Hipsters in the East Village presumably wear all kinds of other clobber that never becomes particularly popular anywhere else, or even in the East Village.
It depends on whether anyone else was open to copying at that time.
This belief in ‘influencers’ can be simply explained using a particular logical fallacy.
Rosenweig’s ‘delusion of the wrong end of the stick’.
This is the tendency to get causes the wrong way round.
For instance, in observing that successful companies tend to have a corporate social responsibility policy, should one infer that these pro-social activities are contributing factors to their success, or is it simply that that profitable companies tend to have money to spend on CSR?
The former makes for a better story – and is therefore lapped up by the purpose-before-profit lobby and more recently proponents of the so-called ‘sharing economy’ – however the latter explanation is much closer to the truth, if somewhat less sexy.
Similarly, ‘influencer’ theory makes for a better story than random copying of each other by ordinary people.
The final irony is, of course, that the so-called ‘traditional’ mass marketing that ‘influencer’ type strategies seeks to discredit is actually far more effective at reaching accidental influencers than activity focused on reaching those with some sort of perceived influence.
Therefore smart marketers could, in effect, have their influencer cake and eat it, too.
As it is impossible to know which person, if any, is going to start any given cascade of influence, then activities should be aimed at as broad a market as possible to give it the best possible chance.
And then if something does catch on they can correctly say ‘we got the influencers’ because the random nature of accidental influence means that ‘influencers’ can only really be identified after the fact.
Eaon Pritchard, strategic planning director at Red Jelly, Australia
i.e. “Please don’t take my ineffectual advertising budget away” – I can prove that the blip of a second you saw my Flash banner on SMH is truly, deeply effective.
User ID not verified.
Applauding.
I wish more agencies and clients would get across the debunking of influencer marketing.
So often I sit in meetings trying politely to stop my head exploding while this malarky is trotted out. And the dagger-stares when you question it…
Cannes case studies and the rest keep propagating the nonsense, too.
More like this please.
User ID not verified.
Why do advertisers hire celebrities to be in their ads? Does it work?
Why do people pay attention to analyst firms or management consulting firms?
Why do people read Mumbrella, or MarketingWeek, or this piece by yourself?
Am I being influenced right now, or not?
Precious little hard data to back up the feels here. If only there was some sort of method for determining which way causation worked.
User ID not verified.
Cracking piece. Good job Eaon.
User ID not verified.
At last some sense, great work Eaon.
User ID not verified.
Great article, it’s among the topics few people dare question these days.
Brands must have symbolic meaning in order to become popular and eventually copied. But a brand only takes on meaning if this meaning is widely known and accepted. To achieve this the brand must be made famous through mass-marketing. The presumed effects of one-on-one dialogues with so-called key influencers etc are, as you say, based on conveniently flawed logic.
(In some cases meaning is created by a subculture and spreads, but this process is extremely sensitive to attempts at influence by marketers)
User ID not verified.
The most famous people on earth and unknown by Gen X. They are Youtube stars and believe you me they pack pounds of influence. My guess is that you are part of the traditional mob, still working to secure budgets from the old school because a) they are big and you receive a nice slice of commission from the publisher / media house and b) you can charge a fortune in production costs. Influence is massive. It is not everything, however it is massive and can be extremely efficient for clients.
User ID not verified.
are unknown*
User ID not verified.
I agree with Are You Kidding – The digital generation can insulate themselves in a world of just One Direction or whoever they like, and be oblivious to politics, other “celebrities” and other issues. Media fragmentation has been a total gamechanger, and in many cases people considered to be “superstars” are not even known by 50 percent of the population, and will more than likely fade into obscurity quicker than you can say “Gangnam Style”
User ID not verified.
Cheers for the response Eoan.
Agree with you that there are certain ‘conditions’ that need to be in place to catalyse the up-take of a product or brand. I’d view this influencer marketing as one of them. My point was not to push one strategy over the other – i believe the most effective approaches would look to encompass both (and others) where appropriate.
However, I find it hard not to look at the Instagram identities I mentioned, as well as YouTubers including Sam Pepper, Fun for Louis (just back from a sponsored Contiki tour) and countless others, and not see the huge influence they have over their devotees.
Their followers, subscribers etc are in contact with them on a daily, even hourly basis. What they do, what they say and what they like (or rather what they say they like) – matters to a huge amount of people. Brands can, and have been, taking advantage.
Cheers
Richie
User ID not verified.
A couple of thoughts for consideration:
‘Influencers’ have not only influence by nature, but attention.
The vast majority of marketing is delivered into channels and forms that people are trying, pretty effectively, to avoid.
Most marketing is not attempting to start trends from point zero.
If an influential celebrity wears a dress that causes rush buying to the point it sells out and creates a fashion trend, should we discount this to a meaningless coincidence?
User ID not verified.
I have just read this article twice & I’m sitting here shaking my head thinking, it makes no sense at all.
“As it is impossible to know which person, if any, is going to start any given cascade of influence, then activities should be aimed at as broad a market as possible to give it the best possible chance.”
So targeting 23 million Australians is going to be the best way to sell nappies?
Influencer campaigns in my opinion should be evaluated in the same way that any digital or social channel should be when forming a brand strategy. If they are going to truly add value, provide performance that will mean the client hits KPIs and generate positive online discussion and recommendation amongst the target market then they are definitely worth utilising.
User ID not verified.
@ Deb, if you haven;t already done so have a read of ‘How Brands Grow’. It shows how mass marketing works with actual/lots of results to back it up.
RE the nappy example, I’ve just had the 3 year ‘pleasure’ of my 2 kids going through (a lot of ) nappys. For the 3 years I have bought nappys every week and the only way I would of been targeted was via mass media advertising as I don’t go on any mummy blogs or follow any nappy influencers (whoever they may be). Appreciate I could be an exception to the rule but valid all the same.
User ID not verified.
It’s all about fit for purpose and multichannel strategy to reach a fragmented audience, the message here in my view is to know your audience, not just simply jump on the hype that is ‘Social Influencers’.
The harder you work to know your audience, how and where they engage with your brand yield the results.
User ID not verified.
For the record, celebrity marketing is not influencer marketing… at least not the way Gladwell and many ‘social media gurus’ define it.
Celebrity endorsements, etc, clearly work.
User ID not verified.
I have to agree with @Richie, @are you kidding me and I think @Julian Ward makes the best point re: fashion selling out after it’s been seen on the right person.
I think proximity plays a huge part in this. Never have we (or they – young people) been so close to our idols. The possibility of being acknowledged for your comment/retweet/likesignal is so tantalisingly close for this audience, it’s a greater spur for mimicry.
I’m not sure the ripple effect is so strong though. The people being influenced are a smaller, if more deeply attached, group. By their nature, it’s less likely they will themselves be influencers. So the pyramid scheme might be relatively flat. And I also wonder whether the radioactive half-life of an influencer might be that much shorter.
What does this mean for marketers? Maintaining a much higher number of ‘touchpoints’ or influencers, which is a lot more effort in the traditional sense of ‘booking media’. But with the possibility of more tangible ROIs.
Of course, you have to be selling the right sort of product to the right type of audience.
Most interesting is that Richie received no comments for his piece, and Eoin has thus far received 13.
User ID not verified.
*Eaon
User ID not verified.
The distinction is the broadest market POSSIBLE not everyone whose arse points to the ground. It means making media buyers earn their keep to get to the largest proportion of the segment.
User ID not verified.
@are you kidding me. The most famous people on earth are not Youtube stars. These people you talk about are most famous to Tweens.
User ID not verified.
@me comment 16: Can you provide a link to a case study where the brand can prove 1) they were able to identify the ‘influencer group’, then 2) target their marketing to that group, then 3) have that group adopt the product/behaviour, then 4) their usage being copied by the broader market?
I ask not to be aggressive or confrontational, but out of genuine interest.
We hear it talked about all the time, but I’m yet to see a credible case study of it actually happening.
Everything I’ve read as been post-rationalisation/a posteriori, etc.
Any links/case studies much appreciated.
(Further, “mums” is not an ‘influencer group’ when it comes to selling nappies.)
User ID not verified.
@Tom Donald
http://www.abc.net.au/4corners.....048106.htm
I think the most persuasive case study from the episode is marketing for the ‘Hunger Games’ sequel. I hope it’s still accessible, or alternatively you might be able to find the original PBS Frontline ep online.
As I mentioned, I’m not fully convinced that usage is being adopted by the broader market.
User ID not verified.
Everyone’s an influencer
User ID not verified.
Thanks for the comments everyone.
@Justin Probably easier to think of celebrities in terms of ‘creative device’.
Re: the data question.
There’s plenty of science that validates things like social defaults, recognition validity environmental priming and such like, you are right though there is no perfect instrument.
@are you kidding me and @geoff Yeah, Im an old bastard and there is stuff I like that half the population has never heard of, too.
@deb targeting the whole of a market is not the same as targeting the whole population.
@Ritchie cheers, will do
To be clear, influence is everywhere. Of course it is.
Some people might be a little bit more influential than others.
Some girls are bigger than others (gen x joke)
Network effects exist, they are pretty complex and we won’t figure out how they really work anytime soon. What we do know is that social influence is pretty subtle, mostly unconscious and most of the time we don’t notice ourselves, not noticing ourselves being influenced.
User ID not verified.
I’d tap popular individuals.
User ID not verified.
Interesting point and happy to agree on many points, in particular that many so-called ‘influencers’ are simply riding the wave. However this is not a one-size-fits-all discussion.
Case in point: Asia.
As someone who works in a relatively sophisticated South East Asian market/agency, it’s very apparent that ‘people power’ is in fact a huge draw. Without getting into too much of a slanging match though, I will say that – just like any channel – you need to do the research. Find yourself the right influencers, because they are out there.
User ID not verified.
@Deb, My TV has nappy ads on it all the time and I have no need for them.
But a longtime ago nappies were a big part of the daily routine in my house and the mass marketing of nappies these days would still inform my recommendation to others. Even though nappies are not apart of my current day, I remember them, I know my old brand and i can offer an opinion on them to others.
To add to @bubbles burst, mass marketing of Beer Brand X will reach fractured and distracted buyers and build mental availability. Slipping 6 cases of Beer Brand X to the bar staff to drink for knock off drinks will create social proof that Beer Brand X worthy of a purchase. If you want a social influencer get into the real world and sweet talk people at the purchase situation.
User ID not verified.
Some evidence for you.
Firstly, this special influencers theory is such a seductive idea, and seems so plausible.
Compared to mass marketing it feels so much more modern.
This is why it’s hard for this business to let go.
For more on getting things the wrong way round, I’d point people to Phil Rosenzwieg’s mighty tome on business delusions ‘The Halo Effect’.
On ‘everyone’s an influencer’ the core of this idea comes from Microsoft scientist Duncan J Watts ‘Everything is Obvious’ has a whole chapter on his social influence experiments with Twitter, full of data analysis.
Also on social influence, Mark Earls is required reading for anyone in this industry who wants to have a point of view.
Similarly ex-Facebook data scientist Paul Adams has also proven that individuals sphere of influence online is only about 5 people deep. He published his findings in ‘Grouped’.
User ID not verified.
Brands that tap into so called influencers do so because they offer REACH combined with a peer to peer evaluation or endorsement of a product or service.
There is in value in that when another option would be to create an ad, which is less likely to be trusted or seen and distributed en masse in environments which the brand has little control over.
I think the challenge for brands that want to tap into influencers is to do it in way that has longevity and is real as opposed to a ‘we’ll pay you to say this’, which audiences are increasingly tiring of.
I don’t disagree that everyone has influence, but everyone doesn’t have reach.
User ID not verified.
I’m still waiting on the conclusive case study that shows it works.
That Hunger Games case study doesn’t wash (for me). The book was a smash hit. The first film was a smash hit. Saying “influencer marketing” led to the second film being a smash hit is like saying influence marketing led to the iPhone5 being a hit. (It was already massive.)
The Byron Sharp analyses says it’s malarky.
The Duncan J Watts work says it’s malarky.
Both of the above use hard data.
All we have is anecdotal evidence of success, or outlier examples (flukes).
I think the pro-camp have a lot more proving to do. (And YouTube stars aren’t influencers – in the way Gladwell/gurus define them. They’re celebrities.)
User ID not verified.
@Tom nice one. And absolutely correct. Actually a significant number of case studies showing some sort of replication of effects are whats required. Im not going to hold my breath.
User ID not verified.
Ummmm… no. I didn’t say it led to the second film being a smash hit. Influencer marketing appears to be an increasingly valid component of a strategy, not the entire strategy itself. The difficulty I have with some of these articles (or comments) is that they discuss things like this in absolutist terms. If you don’t want to use this ‘channel’ as part of your toolbox, then it becomes my competitive advantage.
Btw. Watch the ep.
User ID not verified.
Looks like your not alone Eaon
http://blogs.hbr.org/2014/09/3.....marketing/
User ID not verified.
@fraser ha, yes. I actually bumped into the author on twitter at the end of last week.
Splendid article.
User ID not verified.
Despite the unfortunate premature end to the whole ‘Kony 2012’ thing, you still have to respect the messaging and goal of that organisation was clear; I sure as hell didn’t see any TV ads or newspaper full pagers.
User ID not verified.
I broadly agree Eaon
It’s interesting to actually scrutinise people’s social media brand interactions. In ethnographic research we conducted around people’s brand interactions on social media and what influenced them, there were big differences between interactions with big mass market brands and smaller specialist interest and niche brands. Whilst people were quite concerned in say fashion with what others are wearing or what cycle experts have to say about a new gearset, people were rather uninterested in other people’s comments about Coke or biscuits or dishwashing liquid. People do connect with bigger brands but on social media but in our research this was more often about accessing specials (eg travel agency specials) or freebies (tix to concerts) rather than extolling the brand or seeking others approval or opinion of it.
User ID not verified.
I have never seen an advert for Beats By Dre, or have I?
User ID not verified.
@ Fraser, best line in the hbr article. ‘The fundamental problem with influencer marketing is not that some people aren’t more influential than others, but that there is little, if any, evidence that influencer strategies—other than celebrity endorsement—are viable.’ So to all those that refuse to agree with what Eaon is saying, or accept what the data shows (see 29 above for best sources), you need to do more than simply ASSERT that influencers work because, well, because you think they do…
User ID not verified.
@brent but anything celebrities like Clooney, Oprah etc are involved in gets reported and broadcast by mass media.
User ID not verified.
Actually Duncan Watts (mentioned earlier), has plenty of data around the impact of “influencers”, particularly around the field of virology and the concept of patient zero (to use a more obvious example). For mine his stance is that influencers do exist and they serve to speed up adoption of something, even if it may well have caught on eventually regardless. Though I realise summing up a really complex set of theories in one sentence is just asking for trouble, so I’ll make it clear: I know this isn’t the only thing he has said.
Another resource stacked full of data is Ed Keller’s book “the influencers”.
In the end, for every book you read saying “influencers exist”, you can read another saying the opposite. Find the balance people…
User ID not verified.
Influencers….hmm James Dean convinced me smoking was cool…so I did. Does that count?
User ID not verified.
Nice one Eaon.
User ID not verified.
@charlie cochrane @mark@quirk
good points, ta.
User ID not verified.
@st yeah, the Keller thing is basically the same ‘special people’ idea as Gladwell’s.
User ID not verified.
Excellent points made. Continuous exposure is a necessity!
User ID not verified.
Great debate. Reminds me why i keep coming back…
User ID not verified.
Although you have to keep a keen eye on your social media strategy, many large American brands are now in fact cutting down on their social media departments as they find that the costs (man hours) hardly way up to the benefit they receive. Having said that… it still pays of if you are able to find influencers for your brand.
User ID not verified.