More fiction from influencer50?

This month’s issue of The Marketer includes an article which claims that analyst influence is falling. Influencer50’s Duncan Brown has the byline on the article, which unsurprisingly includes no detail to support the claim.

As we commented in October, while this claim is bound to get headlines influencer50 doesn’t seem to have any research that back the claim up. Brown gave me this article, which supposedly appeared in a supplement to the Journal, and refers to a Santa Clara study. Neither the Journal, the journalist, Santa Clara or influencer50 were able to point us to the data themselves. Luckily, we found the data from the Santa Clara study online, and they don’t support the claim at all. In fact, the question wasn’t even asked.

As far as anyone can see, influencer50’s claim has no material basis at all. It may as well be fiction. Our CIO surveys do not show analyst influence to be falling; the percentage of firms subscribing to analysts is fairly stable, and even growing modestly, as Outsell’s data also show. For so long as influence50’s conclusion remains irreproducible, we’ll remain lukewarm about the firm.

P.S. Brown has left a useful comment on this post, to explain that while analyst influence had reduced over the last five years they make this point simply to get people’s attention: while it’s an opening point, it’s not the key one. They key point is than influence is diffuse. And indeed it is. However, we do feel it noteworthy that influencer50 has raised no data to defend its position, even though it is a company ostensibly engaged in primary research. Lighthouse has commissioned research on analyst influence over a number of years, and our data just don’t confirm influencer50’s findings. We’ve seen nothing more than a one percent rise or fall in the percentage of firms using analysts, but we have seen a big rise in the use of these firms by their subscribers: they are using analysts much more than they were, and this translates into more influence across the buying cycle: much more. We also a see a great interest in mediated analyst influence: the media, bloggers, and many other influencers use analyst research more, so analyst influence also determines much of what is said by other influencers. You can read more about that here. We’re also adding a link to bad science.

More from Duncan Chapple

Webinar: What AR delivers, and what it’s asked for

On 31 March I’ll be presenting the findings from CCgroup’s annual survey...
Read More


  • Duncan,

    Thanks for pointing your readers to my article in The Marketer. The Chartered Institute of Marketing is an important body, so we’re glad to be associated with it. The article is available from the CIM, or here

    I doubt we’re going to agree on the trend in analyst influence, and I’m as unconvinced by your data as you are by mine. However, perhaps it’s worth noting that, while the “Analyst Influence is Diminishing” theme has been a useful hook to start a discussion on Influencer Marketing, it’s never been the main focus of our proposition or marketing. It has engaged the wider AR community in debate and thus served its purpose. Readers of this blog might want to read our “The Insanity of Marketing” paper, which is another hook for engagement

    In other words, while we stand by our position that analyst influence had diminished over the past five years, it’s a side issue for us. Our clients come from the AR community, but also PR, product marketing, partner marketing, industry solutions marketing and so on.

    Our focus is on achieving a balance in emphasis. The analyst community is highly influential, but not exclusively so. Whether influence is diminishing or increasing is irrelevant to vendors, as long as they get the balance right.

  • Hi Duncan,

    I think you’re really proving my point: you make the claim that analyst influence is falling, not because you have any data to support the claim, but because it gets you attention. The word you use – “position” – seems to be exactly right, since it’s a conviction rather than anything for you are supplying any proof.

    Data, as you know, is plural. But if you had even one point of fact that supported your position, would you not have published it by now? None of your white papers seem to point to any evidence for your position.


  • Hello again,

    Re publishing data, we’ve compiled our own data over the past 3 years, which led us to our conclusion. It’s referred to in the original white paper. We’ve taken the decision not to publish the full data set, as it may compromise the competitive advantage currently being enjoyed by our clients. To try and offset this in the white paper we did also point to KCG and OutSell who had supporting views.

    However, if any of your blog readers are vendors concerned at the prospect of diminishing analyst influence then I may be able to share some data with them, as long as it doesn’t compromise our existing clients’ confidentiality and competitive advantage.

    Please note, though, that a lack of published data is not a barrier to drawing a conclusion. Gartner’s MQ and probability statements are good examples of this (just read the MQ disclaimer). Considerable evidence is available, if you ask the right questions of the right people.

    I’m also wary of assertions that equate analyst firm revenue with influence. According to Bill Hopkins’s new book, most analyst firms don’t have much influence (over decision makers, which is what we measure). So the fact that firms spend money on analyst research doesn’t mean analysts are influential at the point/time of decision.

    It also doesn’t account for the rise in influence of other types of individual, nor the dispersion of technology decisions amongst non-IT executives, into which analysts have little reach.

  • Your original paper doesn’t really indicate any data, just views. Why not publish just one tiny data point, for example, by how much has analyst influence fallen over the last three years?

    Without data, you can develop hypotheses but you cannot develop conclusions without data. Gartner’s probability forecasts are clearly presented as hypotheses about the future. Hypotheses are not evidence.

    I do some that some readers take Duncan up on his offer to share some data with them.

Comments are closed.