Gartner underhypes open source

I'm not sure who Gartner talks to when it puts together its famous "Hype Cycle" reports, but I'm finding it hard to believe that it talks with enterprises. I was recently reading through its "Hype Cycle for Open-Source Software, 2007" report, and was astounded to find out that I've been tricked by paying customers into believing that they were, well, paying.

This chart comes from Gartner's 'Hype Cycle for Open-Source Software, 2007.'

(Credit: Gartner)

To wit, Gartner suggests that we're years away from enterprise adoption of the following open-source software categories:

• Content management (5-10 years);
• Enterprise Service Bus (5-10 years);
• J2EE Application Servers (2-5 years); and
• IP Telephony (2-5 years).

Does "mainstream" mean that everyone has already bought it? Or does it mean that a wide cross-section of the market is adopting it, and not merely the proverbial "early adopters"?

If the latter, I heartily disagree with Gartner.

I work for Alfresco, an enterprise content management company. I advise 10 other open-source companies and keep in close contact with others, including Red Hat and MySQL. For the record, it's hard to find an enterprise these days that isn't actively evaluating and/or adopting open source.

I spent the day with two members of the "late adopter" class. One is a large manufacturer within the automotive industry. The other is a large retailer. Both seem to have forgotten that they should be waiting five or more years before buying open-source content management.

JBoss sells its open-source middleware solutions into the mainstream. It clearly didn't get Gartner's memo. MuleSource can't seem to stop selling beyond its solid core in the financial services market--apparently it missed the memo, too. (Disclosure: I'm an adviser to MuleSource.)

Yes, there are plenty of early adopter types buying into open source. But in many open-source software markets, we're deep into this crowd and well into the "early majority" phase of growth.

Gartner understands this, but seems to leave it out of its Hype Cycle calculation. Gartner writes:

Today, it has become impractical for mainstream IT organizations to avoid or ignore the influence of open source across a wide variety of industry market segments. Doing so will put organizations at a serious disadvantage against competitors who are leveraging mature, stable and well-supported open-source technologies for significant return on investment and total cost of ownership (TCO) opportunities. Moreover, open source is entering IT organizations embedded within market-leading products (themselves not open source) from a wide range of vendors, including IBM, Oracle, BEA Systems, SAP and many others.

Amen. Now we just need to see that reality reflected in the Hype Cycle schematic.

[satayboy]

Gartner likes to sell to both the vendors and the customers. How many open-source vendors do you think Gartner sells to?

[bryan811]

I have also been astounded by Gartner over the years and their total lack of understanding of open source, and the depth to which enterprise is using a range of OS technologies in their stacks. Most MIS/IT manages know what Gartner apparently does not; that is, the open source train left the platform some years ago. I doubt there is actually a researcher at Gartner who fully understands this and I pity those organizations that shell out an obscene amount of money for their essentially worthless publications.

[ebegoli]

Matt,

Gartner is hardly relevant for the people who really pay attention to technology trends.

Gartner's views are not accidental. Their modus operandi is to build and preserve a strong relationship between technologically uninformed and unexcited CIOs and large vendors.

Open source will start playing a role for them only when it somehow aquires the large financial power base.

[royrusso]

[I'm not sure who Gartner talks to when it puts together its famous "Hype Cycle" reports]

They talk to whomever pays them the most.

[Savio.Rodrigues]

Matt, one point to consider is that analysts in general base what they write on what they hear from clients and the vendor community.

If clients stop asking Gartner "hey, is this OSS stuff real", because the clients are already using OSS, then this could lead to an analyst concluding "nobody is asking us about OSS anymore...maybe that wave of questions last year were just that, a wave".

I can speak for the OSS JEE App Server space....sure customers are using Tomcat, JBoss, Geronimo, WAS CE and other products. But all I can say is that a lot (emphasize lot) more of our customers are not using those products than the ones who are using these products and commercial app server products. So, 2-5 yrs seems reasonable (if not aggressive)...Remember you don't talk to the customers that are using XYZ and happy with XYZ and wouldn't even consider OSS as a replacement to XYZ.

BTW...are you sure you're allowed to show the hype-cycle without asking for permission from Gartner beforehand? ;-) [http://...read the copyright notice at the bottom of the Gartner reports....don't worry, I won't tell|http://...read the copyright notice at the bottom of the Gartner reports....don't worry, I won't tell]

[markdriver]

Hi Matt,

This is Mark Driver. I'm the primary analyst for open source research at Gartner and the lead author of the hype cycle research note you mention here. The graphic you show has _alot_ of text behind it; drop me a line and we can schedule some time to discuss the note and Gartner's overall coverage of open source if you'd like. It?s an involved discussion so i wont attempt to cover the details here but a couple of quick comments.

First you can find some details on the hype cycle model at http://www.gartner.com/it/products/research/methodologies/research_hype.jsp

We definitely aren?t saying that mainstream users should avoid technologies until they reach the far right of the HC. The HC is a snapshot in time measurement of the what ?is? happening rather than what ?should? happen.

For example, related to open source content management, the research doesn?t say that mainstream IT organization should avoid products like Alfreso; instead it says that the vast majority of mainstream IT organizations that leverage content management tools have not yet adopted open source content management; but it also says that we believe the market penetration will reach widespread adoption within 2 to 5 years as well.

[kvandersluis]

I'm attending Gartner's conference on Application Architecture in Las Vegas currently, and all the buzz here is about agility and SOA. One gartner analysts has stated that SOA has emerged from the "trough of dissillusionment", and is now poised for the mainstream. I do wonder the extent vendors influence this position, as most of the big SOA vendors are Gartner customers. I've bogged about this here - http://www.xaware.org/component/option,com_myblog/show,Along-the-Hype-Curve.html/Itemid,54/. Based on my experience with customers over the last year, Open Source in general is further along the hype curve, and SOA is not quite as far along as Gartner suggests.

[jay_batson]

Having been a Forrester analyst for several years, I have a pattern recognition capability in analyst work, and I see a typical analyst flaw here. A regular thing analysts try to do is detect and describe patterns in market behaviors. Two things happen when trying to do that:
1 - they need to succeed at finding a pattern, whether there is one or not. This is the I Need To Find Something Here To Justify My Value And Price disease.
2 - in order to communicate the pattern and structure, they invent some method to communicate an idea, and force things into that method. This is the I Invented This Hammer and Now Everything Is A Nail disease.

The Gartner pattern here is a tried and true Gartner curve, but applied inappropriately. This type of curve can be effective at explaining specific products or markets. But OSS is not a market, and it's not appropriate to apply this type of curve to a generic type like OSS.

I smell an analyst with a requirement to write a report, and Finding Stuff (that isn't there.)

[dskarjune]

Best to ignore anything coming out of Gartner regarding OSS. As already commented here, they are heavily influenced by vendors, preach to the uniformed looking for a silver bullet to budget, and behave like analysts who are required to produce reports that target forgone conclusions.

At a global convention, I asked Gartner analysts direct questions about how OSS and Open Standards were affecting CMS, for example, and they publicly laughed at the questions and refused to answer them. That's either ignorance or they truly worry that they will be seriously undercut in the coming years as their over-rated influence wanes.