Wednesday, June 16, 2010

IBM-Coremetrics: why it matters & my take

Disclaimer: I do not have any insider info and those comments are my personal opinion and point of view. In legal terms... those are not forward looking statements.
Another shockwave in the small web analytics world: IBM Advances Analytics with Acquisition of Coremetrics.

The history

This shouldn't come as a huge surprise. When Adobe announced the acquisition of Omniture few people understood the strategy and complementarity between the three players. In "Adobe + Omniture + WPP = game changer" I commented that "There seems to be some concensus this acquisition is awkward. However, it is in the same vein as IBM's Websphere partnership to integrate Coremetrics within the development platform and production cycle."

In some ways, IBM fix a mistake: according to some people, they should not have sold Surf'Aid in 2006. Remember this was at a time when Google had a much smaller market penetration and Micrsoft purchased a small web analytics player named DeepMetrix, based in Gatineau, Canada... the rest is history. But the re-purchase has not much to do with Surfaid.

Today's picture

According to a study of the top 500 US retail sites study, Google Analytics enjoys a market share of 62%  (up 19% from a year ago). Compare that to Omniture 40% (up 8%) and Coremetrics mere 18% (up 2% from last year). Google 60% figure is consistent with other confidential studies I conducted for the European market.

"Double dipping" (using more than one tag), is up to 33% and significantly in favor of Google.

Coremetrics offering was traditionally considered to be very strong for retail sites - a position that is obviously eroding when looking at the market share results.

What's the deal?

Google dominates the market
It is not even a matter of price, it's a matter of features. The usual claim that Google Analytics was used on small insignificant sites and amateur blogs doesn't hold anymore. Google Analytics has reached the "enterprise-class" level, and businesses are massively adopting Google.

There are a couple of perverse effects however (more on that bellow)

Adobe + Omniture
The goal is to integrate analytics into the development life cycle. There were also some speculations on some type of ad network enhancements trough Flash and better measurement but I don't believe it... Flash long term survival is seriously threatened by HTML5.

Adobe culture and historical offering is focused on marketing and rich media.

IBM + Coremetrics
The relationship between IBM WebSphere and Coremetrics is already in place. I recently mentioned IBM is pushing for analytics front and center in a massive, worldwide, multi-channel marketing campaign (I really like it and went back to it several times!).

IBM isn't about ad placement, marketing or rich media, IBM is about running core business processes. It's about business analysis and business intelligence contributing to bottom line core business optimization. We are in a whole different ballgame.

My take

IBM move is wise and certainly an influx of fresh air for Coremetrics. IBM's culture and experience with acquisitions leads us to think Coremetrics might become closely embedded with WebSphere and maybe marginalized as a standalone product.

In my Roadmap to Online Analytics Workshop I conclude with the slide bellow:

What it means:

  • IT & Logs: Early websites started in IT, and like any system, websites had logs (circa 1993). IT admins were looking at those logs for traffic source, usage and errors.
  • Marketing: Eventually, marketing found out IT people were talking to customers - bad idea! So Marketing took over the web (circa 1998)
  • Tags: Marketing still had to ask IT for logs - complex, slow, not sexy (think old WebTrends) - tags solved the issue! Google democratized this approach (circa 2005)
  • Business: Modern websites are not only about marketing - they transcend all business functions.
  • Online: Measuring the web isn't enough - think about the growing complexity of the online ecosystem, including social media & mobile.
  • BI: As we evolve our web analytics practice, we seek to integrate external data sources: call center, voice of customer, core systems data... we're going toward business intelligence (not the technology, the concepts)
  • Business: Marketing optimization isn't enough. You can do a great job at marketing, but if you bring all those visitors to a broken process you are wasting your money. Eventually, to optimize the online channel you have to look at the internals - you need to become business analysts.
In the IBM-Coremetrics context, the path toward business analyst and business intelligence is pretty clear.

What's next?

There have been speculations for some time about WebTrends. The company bumpy road in the past several years is now more stable and the product was improved (although it's a strange coincidence to see the VP of Marketing leaving WebTrends the same day IBM-Coremetrics is announced). Microsoft-WebTrends would certainly be a good fit. Microsoft is already a heavy user of WebTrends and since the debacle of Gatineau, Microsoft is left with no web analytics offering. Culturally, WebTrends remains one of the few solution that can be hosted internally. WebTrends would be a natural fit with other server-based offering from Microsoft, be it SQL Server, Sharepoint, IIS and other MS Office "enablers". Some people think Oracle-WebTrends is a possibility but I don't see a good fit.

Firing at Google
Anyone in the online world saying Google isn't a threat is either stupid or lying (and for that matter, even in the offline world!). Since a little controversy is always interesting, and will all due respect for the benefits and strengths of Google Analytics, I think there is some unfair market competition and clear subsidizing of GA from AdWords.
  • There is clear market dominance of Google, and it's grown from about 40% to 60%. At this rate, Google Analytics will own over 80% of the market within a couple of years - often the threshold beyond which "monopoly" becomes a risk.
  • Even for organizations standardizing on a competing product, Google Analytics sneaks in trough the "path of least resistance": agencies are largely using it and the fact there is no cost allows it to go under the radar. When time will come to renew paid contracts, and when companies will look at their analytics maturity, the choice will often go toward GA.
  • There is a legitimate concern the entity providing marketing effectiveness metrics is also the same who handles the marketing campaigns. It's a clear judge and jury risk. (In the same way I see it as a business risk when the web agency creating the site is also the one doing the web analytics analysis - not illegal, not really unethical, but certainly a risk for abuse).
  • But more importantly, most people now make the connection that "if I do AdWords, therefore I need to use Google Analytics". And who doesn't do AdWords as part of their marketing strategy?
The market dominance of Goole makes it very risky for anyone to sue them for anticompetitive practices. Most of the time, smaller web analytics players would much rather be purchased by Google than risk alienating them by venturing into a long and expensive legal process. But... Microsoft or IBM would probably not be as afraid to do so...

So, what's your take? Any radically different point of view? Am I off track or do you agree?