Directional influence and the Obama question

A few days ago I read an interesting post by Auren Hoffman on homophily – the phenomenon of being affected by one's friends and close associates. Intuitively this makes sense – we all make decisions influenced by those around us.

But there’s an important distinction to make between the existence of influence and its direction. What I mean is this: you might be influenced in the purchase of a new digital camera by a friend who has bought one recently. But are you more or less likely to buy the same model as your friend? You might be inclined not to buy that model, even though it might be the best model for you, precisely because your friend just bought one.

I was reminded of this case while reading Dan Ariely’s excellent Predictably Irrational. His example of ordering beer demonstrates the phenomenon at work. It turns out that when ordering out loud people in a group opt for more variety, not less. Ariely suggests that this is because people need to choose something different to show they have a mind of their own, that their order conveys individuality, or perhaps that they are trying to impress.

This might mean that people order beer they don’t actually want to drink. Irrational maybe, but experimentally validated.

The really interesting part is that when people are allowed to order in private, by writing down their order, they order what they want.

Understanding this, from an influence viewpoint, is important:

1. People are influenced by others, but that influence may cause a decision contrary to the choices made by others;
2. People may make better (or at least more truthful) decisions by being protected from the influence of others and making their decisions in private.

This is very pertinent today of all days, as the US goes to the polls. The well-documented Bradley effect is an example of how some people will state their voting intentions in public, but vote differently when in the privacy of the polling booth. Will people who said they'll vote for Obama really vote for him?

We’ll see shortly in which direction the US public has truly been influenced.

The influence of online product reviewers

Rubicon Consulting has written a white paper based on research conducted on US-based web users. Rubicon is run by Nilofer Merchant, with whom I worked in compiling case studies for the book.

There are some important points to pull out from the study. It finds that those people that regularly post reviews and comments are not your average customer, but enthusiasts (or enthusiastic detractors). Some firms may decide that these folk exist at the extreme ends of the customer spectrum, are not typical of general customer, and can therefore be ignored.

This is a mistake: although average customers don’t post reviews they do read them. Importantly, product reviews drive product purchases, so ignoring the review posters is dangerous. As the paper concludes:

“The most frequent contributors are the influencers, and they have a strong influence on purchase decisions because they write most of the online recommendations and reviews.”

This means that firms can’t ignore frequent contributors, but they have to talk to them in a different way to ‘normal’ customers. This is music to my ears, echoing Influencer50’s own mantra of “Don’t pitch to influencers.”

Other findings I picked out include:

• Approaches that work well in one type of community may fail utterly in another. Confirmation of the ‘horses for courses’ guide to influence ecosystems.
• Confirmation of the 90-9-1 rule: 90% of users are lurkers, 9% of users contribute from time to time, and 1% of users participate a lot and account for most contributions.
• Influence of product reviews varies by category. You’re more likely to use an online review to buy a digital camera than you are to choose a doctor. (I’m relieved to hear this!)
• Online discussion is theatre: “Web discussion is a performance in which a small group of people interact with each other, and with companies, for the benefit, education, and amusement of everyone else.” Understand this and it shapes your entire approach to online communities.

There is a ton of other information on web usage in the US, which makes interesting reading. For example, the research finds that web users are more likely to vote Democratic. That should be an interesting theory to check in the coming week…

Influence in SMBs

Barbara has picked up on the new CMB Sage Market Pulse study, which shows who SMBs use in making IT decisions. A high dependence on independent consultants and peers, followed by (whisper it) vendors' sales reps.

The influence of vendors on their own markets is typically understated. A typical buying pattern of an SMB is: get quotes from three vendors. A consultant or systems integrator or VAR may provide this quote gathering and assessment service by proxy. But that's pretty much all there is to it in the supply chain.

Little analyst penetration at this price point. But what's stranger is the absence of journalists and bloggers, much of whose information and opinion is widely available and free. And where are the other sources of advice, such as industry associations, government agencies and other influencers not in the supply chain. Was this an omission in the survey?

Anyway, the survey supports Forrester's own study last year, which found similar sources of influence, though in a slightly different order.

The main point is that SMBs are influenced by different folk than larger organisations. Indeed, SMBs are not a contiguous group, and there are many variations in influence dynamics within segments of the broad SMB space. So watch out if you're targeting firms other than enterprise size - you may be surprised who pop up in the influence ecosystem.

More lists of influencers

You know how I like lists of influencers. Or rather, how I like to rant at their general pointlessness.

So I got all excited about another list source, called Most Public. It’s an index of the most influential public figures in some predefined news community. Like New York.

I'm troubled by any information source that claims "a teenage Twitterer may have as powerful a voice as the New York Times editorial board". What nonsense, at least without qualification (such as, influential on whom?).

Oh, and it has to be online influence. The measurement criteria, which are published, make it clear that it’s the online world that is being indexed.

In fact, I actually quite like the idea of this type of list, especially when there’s an obvious methodology in play. Disagree with the method, but you can’t claim that the list is made up randomly (unlike most compilations of influencers).

Still I can’t help wondering, who are these people influential on? My first guess is, other people on the list. There’s a tendency in the online blogosphere twitterverse web2.0 world to refer to other people in the same community. This is, of course, natural since we gravitate to others like us. Fair enough. Except don’t assume that the online community is a proxy for the rest of the world. It isn't. In a recent survey we conducted for a client, the most popular answer to "Which blogs do you read?" was "What's a blog?"

I’m left with the impression that the New York Most Public list is interesting, like a top ten list of marching band music is interesting, if you’re interested in that sort of thing. From a practical point of view, I’d rather see a top 50 list of the most influential restaurant critics in NY. Or who’s influencing advertising trends in print media. Or who’s influencing the economy. Or who’s influencing voting intentions.

Lists are all very well, but they beg the question, what are they for. Too often, this question is left unasked.

A reminder

When considering influence, please remember to pose the question: influential on whom?

It saves a lot of bother if you have this question in mind before embarking on any Influencer program, or start composing a list of “influencers”, or even dismissing the concept of influencers.

It also helps if you have a decent understanding of the answer. Hopefully it will be something useful like “Influential on our customers and prospects”.

What constitutes quality in an influencer?

One of Influencer50’s criteria by which we score and rank influencers is Quality of Impact. But what is 'quality' in terms of influence?

Firstly, quality is not an all-or-nothing concept. It might be obvious if an individual has a lot of it, or none at all, but what about the large grey area in the middle?

We use Robert Cialdini’s book (bible?) on influence to provide the framework for our Quality measure. In his discussion on authority, Cialdini poses two questions:

1. Is this person truly an expert (measured by the person’s credentials and the relevance of those credentials to the matter in hand)?
2. How truthful can we expect the expert to be?

We simplify these to 'Expertise' and 'Independence.'

In the book we discuss the relative importance of these two dimensions, and offered the following diagram:



The diagram implies that 100% expertise and 0% independence means that influence is severely constrained. Similarly, 0% expertise and 100% independence also represents low influence. But is this true in the real world?

There are many cases where people have 100% independence and 0% expertise. My own influence on the wine trade is a good example of this (never trust my wine suggestions), and it translates (intuitively) to low influence. But is the converse true? I think I’d rather buy wine from an expert who works for a wine producer, even if they recommend their own wine. In fact, I’d surprised if they didn’t. As long as they’ve declared their interest I know that the advice I’m getting is trustworthy, if qualified.

Back in the real world, can vendors be influence on the market? Absolutely. No-one expects them to be independent, but they can demonstrate their ample expertise, and be influential.

So the relationship between expertise and independence looks more like this:



In other words, you’ve got to have some minimum level of relevant expertise to be influential at all. Independence increases influence, but it is not a pre-requisite for it.

Why you can’t guess your influencers

We often play a game with our clients. Write down the names of the top ten influencers on your market (or segment). If you guess correctly you don’t pay us.

It’s a safe bet – we’ve never encountered a close guess. But why?

I think there are two possible reasons. Firstly, most people have never thought about the question before. Although intuitively they know that their prospects are being influenced by a range of individuals, they’ve never considered who these people might be.

The second reason is that when considering influence, they use one, or maybe two, dimensions to measure influence. The most common ones used are frequency and market reach/awareness. Sometime they’ll use connectedness, especially if they’re considering the influence of bloggers.

The problem is that influence is multi-dimensional. Currently we use four dimensions of influence, and are piloting a further four (from which we expect two to be practical and consistently measurable).

It also explains why Bill Gates and Steve Jobs rarely turn up on our influencer rankings, along with the other obvious CEO of top companies. These individuals may influence industry trends and directions, but they rarely affect real decisions at the coal face.

Analysts influence, as measured by HP

HP's AR blog announces its research findings of its survey of what influences customers’ decisions to place a vendor on the short-list. Interesting stuff.

But note that while 55% of respondents say analysts do influence the decision, this doesn't not equate to share of influence. It might be that 80% of respondents think peer customers have influence, or that 100% think consulting firms have influence, or that journalists have no influence at all.

Unfortunately, they don't say who else also influences customers, or where in the ranking analysts come. I assume it wasn't top...

The only tidbit they offer is that bloggers and social network sites rank 11th of the 14 types. Go on, HP. Tell us the order of influencers. We'd all love to know.

Influence, independence and impact

Good post from Jon on the influence, independence and impact of analysts. It comes spookily hot on the heels of Alan PS's note on the independence of analysts. That this issue still pops up decades after it was first raised undermines the credibility of the analyst industry. When I was at Ovum, we declined to take money from vendors for white papers, but changed our minds several times during my time there (1995-2004). At IDC (2004-2006) we happily and regularly took money from vendors for white papers. At both organisations we claimed independence.

At Influencer50 we are, not surprisingly, focused on influence. But it's always important to say what we mean when we try to identify and measure it. We think independence and impact are constituents of influence. A lack of independence erodes influence, but doesn't eliminated it. There are plenty of folks working at vendor firms that are themselves influential, but you wouldn't expect them to be independent. Vested interest, as long as it's declared, is the key issue.

The main issue is, then, transparency. So, come on analysts. Why not declare the extent of revenue from vendors. Name your clients and the proportion of revenues they contribute.

Perhaps the IIAR could define (or rate?) analyst firms on their independence...