Friday, 16 November 2007

Rebels vs. Dinosaurs: Are new analyst business models that revolutionary?

The original IT analysis model was about selling detailed reports advising users on either products or pricing. It’s how Gideon started Gartner Inc. -something not mentioned here. One of the last surviving examples of this era, made of typing up (and sometimes translating) technology reports, and sell copies is Sievers – a quaint German boutique analyst. This model soon evolved into selling annual subscriptions. Then, value added services were added in the shape of inquiry time to make something known in the business as RAS: Research Advisory Services. The model is simple: write enough to keep subscribers hooked, make it difficult to understand to push users to use up their inquiry time and all sales then have to do is sell renewals. The 800 Borg reps still largely work this way and wake up at renewal time to negotiate the increase and go back in hibernation for 10 months.

On top of those staple offerings, analysts firms sell consulting days (SAS or Strategic Advisory Service in Borg speak) that vendors use for speaking engagements and pre-launch product messaging sessions. Users use consulting services from analysts for contract review, vendor selection and generally speaking to cover their backside when they’re making a politically charged decision. In the industry it is sometimes referred as to the BACS –Borg Ass Covering Service. Some firms also make good money from their events business (although this is more cyclical).

The clients for the services described above are both IT vendors and IT users, in variable proportion. Examples of IT analyst firms operating on a RAS model include Gartner Inc. (IT), Forrester Research, Inc. (FORR), AMR, NelsonHall, Ovum, Yankee Group Research, etc…

Those firms and some others have also sussed out that by speaking to all vendors, they could reverse-engineer their business model and work-out their market shares: this is what spreadsheet jockeys at IDC and GartnerDataquest do for a living. Note that since most journos have no clue whatsoever on the market, IDC and DQ enjoy a high SOV with their numbers being quoted by IT and business press day in and day out.

The clients for number crunchers are IT vendors (market shares tend to be quite useful for those strategic planning exercises at the end of every fiscal, and sometimes also for setting performance goals, at which point it usually goes pear shaped when a product manager calls up on AR to know why is product’s share is not doing great).

This is nice and simple, even if you add quite a large numbers of independent analysts, trying to make a living of selling white papers, consulting and speaking engagements to vendors.

However, two trends are changing the face of the IT Analysis industry: Open Source Research and Web 2.0. We’ve discussed open source research in this post and there have been great discussions on Jonny’s praise of open source analysis. Not to start over the conversation again, but we’d agree with Duncan that free at the point of delivery is not the result of a collaborative and transparent process where producers and consumers are the same. The lack of wikis (check also here) does not mean that James’ thoughts are not shaped by his interactions with the community nor than there’s not a vibrant community of independent analysts working loosely together –see The Enterprise Irregulars and the BritsGang for instance.

However, free does not mean valuable, rather plentiful and it makes sifting through bloggers and wannabe-analysts difficult for the AR professional trying to figure out who’s influential and who’s not. In that sense, the Rebels contribute to aggravate the “internet clutter” as expressed by Joseph Martin. Noise doesn’t equate with influence and this is why IT buyers like “reference libraries” –repositories of a single version of the truth. Gartner provides this and it partially explains their success.
Librairies are fine but someone to show you around is even better. Richard puts elegantly (or not?) here that the value of analysts is in the conversations more than anything else, that reports are only an entry ticket to the "inner sanctum". Egoes aside, a conversation with a good analyst is always the best way to cut through that clutter.... James would surely agree with this point on conversations but gives the research away -at no cost.

Redmoonk seems to be doing fine even though we see no Porsche in front of their offices, so one can wonder why doesn't everyone switch to open source research? The software and web markets are full of successful examples like at Linux, Wikipedia, surely the analysts can make a living from support (consulting) revenues just like Jboss, Red Hat, etc…?

There are two fundamental issues limiting the options for new entrants:

1. There’s a big difference between direct analysis (commentary) and in depth reports requiring primary research. At one end, there can only be so many James (thank God!): there’s intense competition, low barrier to entries but a given community has a finite attention span and number of relationships that can be established. On the other hand, primary research is expensive and needs to be syndicated –except if one relies on self-selecting samples and e-polls (see this honest comment from Dale here.

Few can comment eloquently on the issue of how different are sponsored research and open-source research. Dale, take this as an invitation: we’d love to hear your views on this very fine line…

2. Skip makes a really good comment when he says that monetizing new media is something we all are struggling with. Tough question, but the ones that will figure it out will make some serious dough and probably also crack the black hole of industry analysts… SMB. Ahem.

Bottom line: The industry is changing fast on the surface, yet IT analysts still have to figure out which business model will allow them to deliver something close to a collaborative “reference library”, containing one version of the truth rather than many in-compatible open-source distributions. Other factors limiting the adoption of open-source research are the delivery mechanisms (frankly, sifting through blogs to gather intelligence is as tedious as resolving Linux driver issues). On the other hand, the enthusiasm of the Rebels and their innovative use of Web 2.0 tools generate a huge deal of goodwill in some communities and some are already making a living of the brand they’ve built.

Stay tuned for the final post in the series: Rebels vs. Dinosaurs: So What Does it Mean for AR?

Read also:
Will the Borg be dis-intermediated?
The Governor, ancient Iraq and Gartner
Open source analyst business model?


Dale Vile said...
This comment has been removed by the author.
Dale Vile said...

OK, you have stirred my into creating a whole new blog.

This post esplains why, and this post is my initial stab at putting the 'changing industry' discussion into context by cheekily comparing AR with the Telco industry

Hope no one takes offence :-)

I will respond to the request for input on the sponsored and open source approaches to research separately.

(PS - sorry for previous comment deletion - was being lazy by not including the links properly - this is just a tidier version so you haven't missed anything)

BFrench said...

I can't agree with any of your assertions about primary research, desk research, and where and how they are being used. These points are perhaps just filler for your post, however you've mixed quite a bit of misinformation into this post.

All analysts perform research and then base forecasts, analysis and opinions on research. If they do not perform research, they are not analysts.

Almost all analyst tasks require a mix of primary and secondary (or desk) research. "Primary" research entails interviewing vendors, retailers, academics, inventors, politicians, associations, etc -- just as much as interviewing purchaser segments. It also entails focus groups and other methods of research beyond surveys and interviews.

And, do not equate desk research with a "spreadsheet jockey." Quality desk research requires substantial expertise and skills. Don't encourage your readers to confuse the homogenized outputs made public -- e.g. simple ranking tables -- with underlying research methods, sources or data management. I've seen the

One final point: the clients for "number crunchers" are not just "IT vendors". Clients include the financial investment community (the original market for IT industry research), governments and other economic/community developers. To understand the importance of the financial industry as a client, look at the 3rd party/indie research distribution channels through Thomson, D&B, Hoovers -- and understand that you are only seeing the tip of the iceberg.

I'm guessing this is all covered in the pile of AR books that have been published over the last 12 months. Perhaps one of those authors could chime in.