Jonathan MacDonald: The Fallacy of Presumed Influence

The Fallacy of Presumed Influence

On June 25th, Stephanie Rosenbloom wrote an article in the New York Times entitled “Got Twitter? What’s Your Influence Score”. Here is the opening text:

“Imagine a world in which we are assigned a number that indicates how influential we are. This number would help determine whether you receive a job, a hotel-room upgrade or free samples at the supermarket. If your influence score is low, you don’t get the promotion, the suite or the complimentary cookies. This is not science fiction. It’s happening to millions of social network users. If you have a Facebook, Twitter or LinkedIn account, you are already being judged — or will be soon. Companies with names like Klout, PeerIndex and Twitter Grader are in the process of scoring millions, eventually billions, of people on their level of influence — or in the lingo, rating “influencers.” Yet the companies are not simply looking at the number of followers or friends you’ve amassed. Rather, they are beginning to measure influence in more nuanced ways, and posting their judgments — in the form of a score — online.”

For those of you in the advertising and marketing industry the above is relatively old news. However, to date, the commentary has been as ‘early stage’ as the topic itself. If pressed, one could cite ‘highlights’ from Hewlett Packard who, after analysing 22 million tweets last year, found that it’s not enough to attract Twitter followers if you wish to be influential — you must inspire those followers to take action. Or maybe look to Jeremiah Owyang, an analyst with Altimeter Group, who wrote earlier in 2011 that using a single metric to evaluate influence is dangerous.

At the time of writing this in November 2011 the overwhelming feeling is that the industry of monitoring and analysing social influence and credibility is exciting, growing, and full of opportunity. Personally speaking, I feel the picture is, at best, partially painted. At worst, I feel it represents a fundamental fallacy of modern world business.

Allow me to explain:

The main thing is not the main thing

From an advertising and marketing perspective there are three types of media. Owned media (company owned websites for example), bought media (Google Adwords for example), and earned media (external conversations about your brand for example). It is common thought that the public value assigned to these three types is lowest in owned media and highest in earned media, with bought media in between. This makes sense.

Conversations that people have can effect the attitudes and behaviours displayed by others, especially when people trust the opinion or advice of the person speaking with them. This is quite unlike how companies often talk at people, hence people tending to prefer personal conversations rather than companies trying to hoodwink them into buying stuff.

Due to the above, accessing those who influence others is seemingly akin to accessing the people who are having a significant impact on how prospective or existing customers are making decisions.

However, identifying some online influencers is a minority part of accessing the power of earned media. The majority part is what you do next. At this fluid world, we find the common mistakes organisations make are:

1. Treating the influencers like any other consumer and speaking at them rather than with them
2. Offering influencers incentives that are seen as bribes, doing more damage than good
3. Doing nothing other than monitoring the influencers, hoping they advocate and don’t turn negative

To properly enter into the world of earned media, an organisation needs to zoom out of the ‘social media’, ‘digital’ and ‘online’ buzz word landscape and address fundamental strategic and organisational paradigm shifts.

This includes elements such as assessing the level of porosity your organisation has, accessing and extrapolating the level of ultimate trust in public, creating and implementing engagement protocols, agreeing and testing sign-off processes, authoring and applying crisis management systems, formulating integration into sales conversion, raising staff awareness and infiltrating behaviour mandates, managing the linkage into performance indicators, and so on. This is a brief selection of modern business hygiene factors. There are many more.

My point is, you can start with monitoring and analysis, provided you understand that monitoring and analysis is less than a single percentage of the actual requirement – and even then, as we will see below – it is as far removed from the holy grail as you could possibly imagine.

What the industry is fascinated by is ultimately a warmup act. A sideshow. I attest it’s not all about finding out who the online influencers are. Not by a long chalk.

The persistent inaccuracy of incompleteness

Imagine I publish a public status update that suggests that a particular device is worth buying and someone else (unconnected with me) gives a similar recommendation at a conference on stage.

Let’s say the same recipient of influence who reads my status update is in the audience at that conference, hearing the same positively reinforced message.

Imagine further if the recipient of both stimuli then goes and buys the device, who is attributed as the influencer?

In another, simpler scenario, imagine I am amongst several influencers but I was the one who pushed the recipient into conversion. However, the conversion happened offline, totally unconnected with online behaviour or activity.

In the first scenario I would be the prime influencer as the offline influence at the conference falls outside the remit of online influence. In the second the influence is invisible.

But let’s not fall into the trap of assuming the online influence market is destined to pause its development at the current level. Let’s give it credit and extrapolate it to our scenarios.

Ask yourself what would it take for both scenarios to be accurately attributed to the genesis of behavioural change into conversion?

Azeem Azhar, CEO of PeerIndex recognises this, calling it “the Clay Shirky problem,” referring to the writer and theorist who doesn’t use Twitter much. “He’s obviously massively influential,” Mr. Azhar said, “and right now he has a terrible PeerIndex.”

I’m sure Clay is distraught by this.

My point is: if the thing I discuss online, or Clay discusses offline, is converted offline, the only way of tracking influence is to have data sensors on absolutely all physical touch points, linked to absolutely all virtual touch points. This, in addition to CCTV and audio recording devices on every square metre on earth.

Frankly, if those who would like to exploit influence were able to act without regulation and legislation, the ultimate win would be to have sensors inside everybody. Microchips that linked our thought, word and deed. Tracking our every move to place exact accountability on everything.

Back in reality however, there are two significant threats to the effective maturation of this industry:

1. Regulation and legislation may get in the way by limiting what is allowed to be ‘mined’ for various reasons – one of which could be human rights, another could be the terms and conditions inherent to certain platforms

2. People may decide to restrict access to their information and thus dilute the completeness of the surveyed data

Either threat would significantly distort the data.

The second threat resonates strongest for me as I personally believe the noise about how all this gorgeous data contains the ‘black gold of the 21st century’, is fundamentally missing the point that the owners of the goldmines are the general public.

Not the companies.

By applying this holistic view to today, with every single influence and credibility tool around, if you make several of your online profiles private, you will have low scores and be seen as non-influential. Even if you are influential.

Can you see my concern?

Many years ago the music industry decided to run a survey to see which guitars were most popular by region in terms of distribution and sales.

To be included in the survey each manufacturer had to permissively share their unit sales.

One particular manufacturer refused to do so, and that was Fender. The problem was, without Fender in the survey the rest of the data was immediately unrepresentative as Fender had a massive impact on the trend analysis.

I feel the online influence industry is destined to be a partial component in a fuller picture. Currently the buzz is loud enough to skim over the cracks…but take either of the scenarios above and think about whether they are outrageous or exceedingly likely. Today.

The question then is: What is the maximum level of incompleteness that accuracy can suffer? After all, companies that are seeking to identify influencers surely aren’t willing to discount chunks of real-life influence?

Surely they aren’t taking what happens on a handful of channels and assuming that’s anything more than a micro-fraction of the picture?

No, I’d imagine that once the dust settles and the cool, funky technology has become yesterday’s spikey haircut, the proper companies will address this game from an un-silo’d strategic place, rather than an online-only viewpoint that tactically assumes what happens digitally is in some way more important or relevant than what happens in total.

The above two points are my main concerns but within them are many subsets. I spend a great deal of my time working with companies to find ways of solving the above challenges and I’m encouraged that the people who understand real business are not sucked in to false dawns or partial information.

Accessing the power of influencers and nurturing armies of fanatics is mission critical in today’s business environment. However the technologist methodology belies the psychological requirements of modern assessment and engagement. When you’re a hammer, everything looks like a nail – and when you make software, everything can be answered in a programme.

Life, however, does not work like that.

There is simply no brand and no system, however cool, that can beat the physiological and sociological ramifications of how people relate, decide and act. Yes, tools can provide us some information, but I implore you to see that incompleteness for what it is.

Some isn’t necessarily all.

Online influence isn’t necessarily influence.

And none of the data gathering and visualisation equals actionable insight without addressing points that are seldom related to digital environments alone.

“The worst form of inequality is to try to make unequal things equal” Aristotle

 Originally published 04 November 2011 on