Skip to content

Carter and The Devil in the Detail: a mammal’s eye view of industry analysts

Today there are the established firms (e.g., AMR, Forrester, Gartner and IDC) that have been around since the 1980’s that on the surface have not changed all that much. Then there are the insurgents like Gartner alumni Vinnie Mirchandani at Deal Architect and James Governor of Redmonk with his Monkchips blog who claim that that they are becoming just as influential as the dinosaur firms. As a consequence, we in the AR community should pay just as much attention to them as old firms.

 - Carter Lusher, head of analyst relations, HP.

I would prefer to see myself as a mammal than an insurgent, given I have no interest in taking up arms. I have been making a conscious effort not to play the David and Goliath card lately, which is one reason Carter’s choice of words surprised me a bit. A bit of context: I recently received an email from Carter pointing to his blog, which I assume means he wants to open a dialogue on the subject of the changing nature of IT advisory and influence. Vinnie has already responded, as has Stephen, who refers to Carter’s argument as a straw man.

I don’t think RedMonk is “just as influential” as legacy firms (no pejorative intended; I am a known as a fan of “legacy” technology platforms: they do however create shackles of success.) RedMonk is playing a different game, with different influence dynamics. In some communities we’re far more influential than any of the Big 4 named above. We’re resolutely committed to lowering the barriers to participation inherent in the traditional industry analyst model and a lot of people give us credit for that. Open source people tend to see echoes of their approach in what we do, which is one reason many commercial open source vendors become customers. We also have enterprise fans (a fact Stephen chooses to underplay here).

I think influence in deployment and purchasing is a far more complex phenomenon than simply asking who talked to the CIO or director that signed a deal. Dale Vile of Freeform Dynamics, who is doing his bit to shake up the quantitative IT research business, recently carried out some interesting research into buying behaviour which showed that many IT Directors and Managers appear to trust their lieutenants to make their own deployment and purchase decisions. Budgets are evidently not as centralised as many people claim. Jonny writes up the numbers here.

Note that above I say deployment and purchase. With open source software, deployment and purchase are loosely, rather than tightly-coupled. MuleSource (client), Terracotta (client) and Alfresco can all be deployed for free. People that download and use Eclipse and want a fully supported “enterprise” version can buy a $29.99 subscription to MyEclipse from Genuitec (client).

My favourite MuleSource story comes from the head of market risk at one of the world’s largest financial services firms, who told me over a glass of wine at Bedales one day that during the company’s 18 month procurement process for SOA all of the major vendors had done a poor job of the dog and pony show, while in the mean time the firm had downloaded and deployed an open source ESB for point to point integration, and so purchased a support contract. Shadow IT is not seen as an evil-its simply part of getting the job done.

Some of the distinctions between sell-side and buy-side won’t stand. There is just too much information and rich community interaction in the middle, too many extant relationships that punch right through what Hugh MacLeod calls The Porous Membrane. My guy in market risk said he would love to be able to follow the lead Mule developers on Twitter. He didn’t need an analyst to tell him the product worked – he downloaded it and kicked the tyres.

I recently helped to broker a conversation between a software vendor and a major Spanish bank. The buyer in question is an important part of the RedMonk community. He is as much a contributor to us as a consumer or customer of our thinking. At RedMonk research is a conversation, rather than a broadcast. Rather wonderfully, before I got a chance to post this blog, Ferran wrote his own treatise on influence. I am going to quote him at length.

A few years ago, when we needed to look for a solution on the SDLC space, the first thing we did was to look at the reports from Gartner, Forrester, Meta Group and/or Ovum. For two reasons: first, because there was not so much information on the Web, except from the vendors Websites, and if you wanted to analyze vendors or products in depth the effort and time required was too big, so in most cases we were forced to turn to some independent analysis; second, because when we showed our decision to our CIO, it was a guarantee of approval if these analysts firms placed the selected product in a good position. CIO’s trusted these reports more than our own analysis.

Nowadays, the scenario is considerably different. First, because CIO’s no longer rely on these analysts in the same way that they did before. I am not saying that they don’t take them into account, but they don’t give them the same importance as before. Secondly, because management rules have changed and now organizations give more confidence to the bottom of the command chain, including developers (a survival strategy if they want to innovate). Third, because a new actor has appeared on stage: open source. A subject that traditional analysts often do not cover so good (if they do). And fourth, because now you can find and share lots of information on the web, in form of white papers, articles, opinions, tips or successful and unsuccessful implementations, and not only from analysts or vendors, but also from the users of these products.

My opinion is that the last argument is the biggest change in the scenario, and forces the rest of the arguments I mentioned before. CIO’s now know that there are more sources of information than the traditional analyst’s reports. CIO’s now knows that their workers can access to these sources of information (they become knowledge workers), it is no longer information reserved for the upper management. CIO’s now knows that some information is only available or it is better described or analyzed in other sources. You can call it Web 2.0, social computing, or whatever you want, but the reality is that now there are lots of information flowing free on the web, inside and outside the firewall. And most importantly, now, I, you and the CIO can share this information with other colleagues.

Pernod-Ricard Asia Pacific is the biggest wine firm in the world. Its a customer. Ask Ric Hayman if RedMonk influences his thinking. 

The existing analyst model is coming under increasing scrutiny.

We recently signed a deal with Atlassian. Jon Silver, Atlassian marketing director, argues in his recent post Analysts Go The Way Of The Media:

“The news and analysis that we read is churned and turned into strategy. Regarding strategy development, we do most of it internally but we have recently penned an agreement to work with RedMonk, a firm that takes a different kind of approach to the whole analyst thing, for a short period of time. We’ll take that for a test ride to see how it works. We like the people behind the RedMonk brand, they’re bright, down-to-earth, and “with it.”

I don’t think we’re know-it-alls — we are far from perfect. It does, however, seem like a heckuvalot better approach than hiring traditional analysts or paying for their expensive opinions. Blogs and bloggers are having the same effect on analysts as they have with news media.

Not being sarcastic: Am I missing something? Is there some other reason to pay for a subscription to Gartner, Forrester or other firms these days? Do big companies still pay for them because it’s cheaper to pay an analyst then try to develop those critical skills in-house?”

Back to Lusher.

“The insurgents are using blogs, but making relatively little use of other forms of new media that I can find. For instance, Redmonk as a podcast, but it has only had three episodes since May 1st. Plus, it is really difficult to find evidence that they are influential whether by directly impacting HP’s sales deals or broadly impacting market perception. The insurgent bloggers all want me and other AR professionals to take it as an article of faith that even though there is no proof that they are influential. That is nice, but they all demand hard proof from HP and other vendors about our claims. Isn’t it a little inconsistent of them to want to be exempted from providing proof?”

There are two points here which I will deal with in turn. First do we “get” or use what Carter calls “multimedia”? And might there even be some useful metrics there?

As Stephen explains on tecosystems we haven’t done the best job of centralising and marketing all our new media work, but its certainly there. RedMonkTV is I hope fun and irreverent, but also educational on key technologies and trends. Meanwhile Ferdy does a pretty good job of pushing back for me.

Just a few examples from what Redmonk is doing (and contradicting Carter’s statement “The insurgents are using blogs, but making relatively little use of other forms of new media that I can find“). The core thesis that guides Redmonk is “technology adoption is increasingly a bottom up proposition“. Michael Coté is doing a spectacular job with RedMonkTV videos, where the main argument is always a conversation between two colleagues. The links imported from that they post daily on their blogs are a gold mine, not only for the links, but also for the comments they make (this is something that I have copied from them in my blog). Twitter conversations are invaluable (yes, updates are not only about what we are eating at the moment).

Does anyone watch this stuff? I am justifiably proud to report that our most successful RedMonkTV episode was watched 26k times. What was the subject? Eclipse-based modeling tools. What a great example of the new open source and media dynamics at work. 26k views…

Its true that we haven’t been doing so much Redmonk Radio, but we have been doing plenty of Monkcasts with ZDNet. Question back at you Carter-why is podcasting more sticky that blogging? Got metrics? At the risk of appearing self-serving, Carter might also like to check out Jonny Bentwood’s Social Media Index: its far from perfect but at least begins to attempt to create engagement metrics.

So what about articles of faith?  I agree this stuff is hard to measure, but I certainly don’t expect anyone to take it as an article of faith that we’re influential. But then perhaps we’re useful even if we’re not “influential” under your current metrics. Talk to us and see the value. Our clients get benefit from working with us.. or they would stop.

Do I demand proof of vendor claims? Sure - customer references is a very useful part of the research tools at hand. Hard proof is different though- how for example is market share calculated? By asking vendors about sales is how. Numbers analysts need to parse and aggregate a number of soft proofs, and soft truths, before providing a market view.

I don’t want to say too much about influence as what Vinnie would call a “sell side phenomenon”. The fact Sun CEO Jonathan Schwartz reads both mine and Stephen’s blogs is a good indicator of industry influence. When Adobe planned to make its first moves in open source they brought in RedMonk to help with the planning. Adobe’s Jeff Whatcott has been a great evangelist.

RedMonk takes a position. 

We advocate openness. We advocate lower barriers to participation. We advocate open standards. We advocate open source.

Those are the areas in which we exert most influence. We’re just scurrying around, warm-blooded, doing things quite a bit differently.

update: bonus link from Barbara at Tekrati.

I couldn’t help but notice that word of eroding Gartner influence has not reached the editorial offices at Computerworld. Gregg Keizer’s piece on Leopard security cites Rich Mogull as “a security consultant and former Gartner Inc. analyst.” Gartner, not Securosis, got the outbound link.

It’s tempting to brush this aside, joking that Rich has formally joined the venerable ranks of used to be’s, like Prince (used to be Prince) and Al Gore (used to be the next president of the United States).

However, that one simple phrase — citing Gartner, omitting Securosis — reflects the difficulty of establishing credibility as an independent analyst or, in Rich’s case, consultant.



picture of a wood mouse courtesy of whiskymac.

Categories: Uncategorized.

Comment Feed

No Responses (yet)

Some HTML is OK

or, reply to this post via trackback.