Web analytics has lagged behind traditional business analysis methods for many years. IBM, a new/old player in the web analytics world, is now in the position to change that and move the two disciplines closer. Statistical modelling will significantly increase the value and prominence of web analytics.

Last week IBM announced the acquisition of web analytics and online marketing optimisation company Coremetrics. This acquisition is another milestone in a growing trend of large corporate organisations buying into digital analytics vendors. Adobe took over Omniture, Coremetrics’ biggest competitor, last September. Google acquired Urchin to create Google Analytics and similarly Yahoo purchased Index Tools rebranding it as Yahoo Web Analytics.

At the same time large data and marketing solution companies such as Unica and SAS are expanding their web analytics and optimisation capabilities through acquisition (Sane NetTracker) and partnership (Speed Trap) respectively.

Some argue that the IBM acquisition of Coremetrics will stifle web analytics innovation. Similar claims have been made regarding the Adobe Omniture tie up.

Whilst such arguments might hold in the Omniture case (Adobe's focus on measuring its own applications), the Coremetrics acquisition is different.

IBM is investing heavily in the area of business analytics and optimisation. It has spent over $10bn since 2005 on acquiring companies in this and related fields. The company recently announced its new big data information management and insight service offer Infosphere BigInsights.

Two key shortcomings of web analytics technology are the large amounts of unstructured data and the lack of any significant statistical modelling and predictive analysis capabilities.

Tools such as Omniture Insight and Webtrends Segments offer significant analytical and visualisation capabilities but leave much to be desired in terms of modelling and future trending. Google Analytics is only starting to address these issues scraping the surface with its Intelligence feature.

In comes SPSS, IBM’s acquired predictive analytics software and solutions firm. IBM is now in a position to merge Coremetrics’ web and mobile analytics collection platform with its powerful statistical modelling software. Such analytical capability is common practice in the offline CRM and Direct Marketing worlds but so lacking in the online world.

Web analysts spend significant amounts of time guesstimating which visitor segments are most important to their business and how to break down users’ behaviour in order to optimise the online customer experience. With SPSS’ decision tree algorithms analysts will no longer need to go through this trial and error process. Instead the software will tell them which visitor segments they should focus their attention on. They would then be able to approach the data in a much more constructive way gaining deeper insight faster.

This is just the tip of the iceberg.

Statistical modelling will enable analysts to identify significant outliers skewing any analysis. They could validate which data streams are significant and which should be ignored. They would be able to structure the data and apply it onto predictive models. These models would then be used to inform A/B and multivariate testing, currently an art as much as a science, and transform them into a more robust empirical process.

More than anything else the IBM Coremetrics tie up symbolises the expected transformation of the web analyst role. Currently a siloed analytical entity residing in the e-commerce or online marketing departments, the web analyst will migrate into a more traditional business analyst role within the insights and MI departments.

This next step in the web analytics revolution promises a significant improvement in the management of online assets. This process won’t be devoid of many new (and some old) challenges. However, I will leave the discussion of those challenges to another less euphoric post.

Michael Feiner

Published 21 June, 2010 by Michael Feiner

Michael Feiner is the founder of web analytics and optimisation consultancy AEP Convert and a contributor to Econsultancy. He can also be found on Twitter and LinkedIn. 

3 more posts from this author

You might be interested in

Comments (16)

Save or Cancel

Brian Clifton

Interesting thoughts Michael...

Whilst your comments are logical and make perfect sense, I am not as optimistic as to the success of this deal. Remember, in 2006 IBM *sold* Surfaid to Coremetrics in the first place (though Coremetrics only used the Websphere client base and not the technology). 

Clearly IBM did not understand the significance of web metrics in 2006 and nothing makes me feel that they do now...

For me, the success of the web metrics industry today is due to the "simplification" that Google Analytics has brought to the table with its 2005 acquisition of Urchin. That has taken web analytics out of the realm of IT and into marketing - where it belongs!

In my view, integrating web analytics with marketing is the future. Integrating web analytics with IT (the SPSS/IBM route) is the past, and flawed. Lets face it, IBM is not a marketing company - just as NetIQ are not (the previous owner's of WebTrends who failed miserably under that brand).

Best regards, Brian

Xoogler and Author of Advanced Web Metrics with Google Analytics.

about 8 years ago

Michael Feiner

Michael Feiner, Director at AEP Convert


I see your point. I would agree that this is not a straight forward easy process. 

As you said IBM abandoned its own web analytics solution in 2006. Let's not forget the failure of SPSS, now part of IBM, in converting NetGenesis, another celebrated web analytics tool, into part of its core offer. SPSS had to abandon the product in 2008.

However, I do think that this integration is the route web analytics has to take. Digital analytics is growing up. I see a growing demand from my clients to integrate their different channels data. And it is not limited to marketing data. It is much more customer-centred.

Companies such as CACI, a traditional marketing and data analysis company, are moving into this space and looking to deploy many of the analytical methods used offline.

Of course, not all of it is relevant to web analytics data because of the unstructured nature of the data (plus data cardinality issues). However, I'm convinced we're going to overcome many of these issues in the next few years.

As for IBM, they have invested massively in developing their business analytics capabilities. Last year it announced a new business consultancy unit (I remember being very impressed with their vision at the time). So perhaps the focus they currently have on providing consultancy services simply wasn't there in 2006.

I hope IBM is successful as it will give us all a push as an industry.

Thanks for your comments.


about 8 years ago


Vincent Ammirato

I'm glad Brian brings up IBM's last few failed forays into web analytics. I'd like to add one extra bit:

Why would I switch to IBM? Their track record isn't encouraging and I've had my analytics suite running for 3+ years (exporting that data isn't exactly easy). I've paid big money training myself and our employees. We're vested.

The ability to model and predict would be nice but what good are the projections if the sample data is too small?

Still more players = more competition = more innovation so no complaints.

about 8 years ago


AK Works

I can see SOME of your arguments, but disagree with you completely. IBM's abduction of Coremetrics will NOT change web analytics nor the web analyst role.

As an SEO, I deal with web analytics a lot. And I'm siloed into just the natural search aspect. Then there's paid search. Just search alone has trouble getting enough resources to work with web analytics.

Do you know where the biggest time guzzler is with analytics? It's setting up the tracking code correctly to measure the results - this includes the time it takes to decide what metrics to measure, etc.

The roadblock to greater roles of web analyst and web analytics is NOT technology. It's office politics. You have marketers who are split -search vs. media, natural vs. paid search - and then you have sales vs. marketing. All of these things hinder greater use of web analytics. Web analytics is only the data gather platform. What you DO with that data is the issue.

If IBM gives Coremetrics all this statistical regression analysis and blah blah blah, that's all great, but who's there to turn that data into anything other than a footnote?

That kind of transformative power resides in office politics, a visionary leader who builds a cooperative team between the sales, marketing, advertising and interactive(web) people towards a common goal.

But given that most C-level execs and middle management have NO IDEA what they are looking at, what they are looking for, or what they SHOULD be looking for- IBM technology in web analytics alone will change nothing.

about 8 years ago


ddd marketing

Like your take. I've been arguing the value of using advanced data modeling of web analytics data for some time now - I think this is really where the industry is going. It might take some time, as many current wed analytics professionals are more tech / programming oriented than statistical specialists. Overall, comparing this to the Adobe Omniture tie-up is right. I don't see the value in that one (to the point where Adobe is now killing off the Omniture brand http://www.datadrivendigitalmarketing.com/main/2010/6/21/adobe-killing-omniture-brand.html) but the potential synergies between IBM and Coremtrics are massive.




about 8 years ago



Great post. Just wondering how many clients are really gong to afford and use all that technology.

about 8 years ago



Hi, Great post, thank you very much for this, Michael. Just reading through the comments, I'd like to reply to Sebastian and Vincent. I believe there will be many companies who will be very happy to switch to a web analytics tool that gives them results that are statistically correct and meaningful. Currently, as Michael states, web analytics is in between science and art. It spits out a lot of interesting data, but it struggles to prove that the data is statistically meaningful and useful. That's where SPSS will come in. The same is valid for email service providers, by the way. All the information they give to users is, in a way, meaningless because statistically not provable in a lot of cases. Big companies will have the money to afford the new Coremetrics / SPSS solution, and they will adopt it happily because it's what they need. Sure, the smaller players won't be able to afford the new web analytics solution, but so be it. They wouldn't need all that complexity anyway, as their data samples would be too small and their marketing campaigns wouldn't be as complex. For them, there's Google Analytics. Regards, Francesco

about 8 years ago

Michael Feiner

Michael Feiner, Director at AEP Convert

Vincent - You wouldn't switch to IBM/Coremetrics. But IBM having these capabilities would put pressure on your vendor to provide similar services. It won't happen overnight but it will push the industry forward.

AK Works - I don't disagree that office politics are a major issue. Heck, didn't that unknown web analytics guru at Google coin a term for it. ;-)
And rubbish in - rubbish out applies here as well. If the tracking and measurements are flawed then a statistical modelling tool won't help.
But ultimately integrating measurement with statistics could help resolve some of these issues. It will give analysts more power and confidence. They would be able to identify commercial & financial opportunities faster and quantify them. A statement like "Segment XYZ is worth £25m in future revenue" is likely to get you a lot more management attention then "We are seeing a basket abandonment rate of 65% from 2nd time visitors".

Sebastian - How many companies can afford using Google Analytics? :-)
Ironically it is Google Analytics that is leading the way in web analytics innovation these days IMO.
The Intelligence feature is a very simple statistical modelling programme. Watch this space - Google are going to enhance this feature and provide additional functionality.

about 8 years ago



Hi Michael - that's a good post, and I agree wholheartedly.

I think a lot of people are catching up in terms of understanding what analytics are about (essentially, giving actionable insight on current stuff, and steering for the future).

I've thought for a while that the job did not sit in IT or marketing or customer experience (all of this is old world thinking, I'm afraid, and leverages only one of Analytics' benefits), but was at the confluence of all these functions.

I was reasonnably happy to see Coremetrics bought by IBM (I'm a user), precisely for the reasons you raise (better be a deluxe satellite of Websphere than a number cruncher for adobe). Sure IBM seemed to have missed the point in 2006, but that was 4 years ago, and there have been leaps and bounds in these 4 years. Just look at what Google achieved in these 4 years: Funnels, on the cuff Segmentation, Insights, API release which enables third parties to come with fresh ways of visualising data... During that peiod, Coremetrics changed from a reporting tool with a bit of a query studio to an awesome flexible analytics. It feels like the tools have improved so much that businesses are finally forced to notice the potential - the limit is now with the number of people using these tools at super proficient level.

Making SPSS work with Coremetrics will be tricky (integrating tools developped separately usually is), but making it easy to answer critical business decisions through forecasting and modelling will take us one step further in getting the function recognised for what it is - Business Intelligence.

In all cases, it's an interesting time in which to be an analyst - when they shoot a mad men set in the analyst community, they'll probably choose 2010...

about 8 years ago

Ian Tester

Ian Tester, Senior Product Manager at brightsolid online publishing

90% of decent-sized websites aren't even using 20% of the functionality of Google Analytics. It's a resource'n'politics thing. 90% of clients with high end solutions don't even get them configured well to collect the best insights (integration with backend databases, external platforms) and barely anyone in the organisation looks at the outputs. Discuss.

It's good to know that I can now get a new bunch of expensive consultants to go with my expensive software, but I'm not waiting for IBM to revolutionise this space. It'll be the Marketing Directors who realise the benefits of having a (technical) analyst on the team and start building out from there.

about 8 years ago


Charles Nicholls

Hi Michael

I agree with your overall hypothesis, but as you noted I’m on record as skeptical that this acquisition will change web analytics, beyond bringing it to mainstream acceptability. Unfortunately the vast majority of acquisitions fail to live up to their potential, and while I think Coremetrics will thrive in the short term, innovation will be stifled.

Coremetrics will now be focused on integration with IBM’s websphere platform – great if you happen to use both and were looking for more integration – but for the majority, largely irrelevant.

Acquisitions also have another unfortunate side effect – the talent that goes to make the acquired company great often leaves, compounding the impact on innovation. And Integrating SPSS with Coremetrics in a meaningful way is non-trivial.

Overall I’m positive on the acquisition for both Coremetrics and IBM, though much more skeptical about the longer term impact. There’d more detail in my blog from last week here


IBM’s track record in Business Intelligence is spotty: acquiring companies is the easy bit. Making all the disparate parts work well together where the whole is greater than the sum of the parts is altogether more tricky.


about 8 years ago


Vincent Ammirato

Michael -  I wholeheartedly hope that IBM's second act/entrance will innovate the field in general.

IBM doesn't just sell software, they sell advice. Using Coremetrics as an IBM product is more like working with a Google Analytics Certified Partner as opposed to going at it alone with GA.

Even more so, I agree with Mr. Ian Tester: Most people only scratch the surface of what GA offers. I've been using it for 4+ years and still feel like I learn something new each month.

As Einstein said, "As our circle of knowlege expands, so does the circumference of darkness surrounding it."

about 8 years ago



The deal makes sense to me.  Just as Yahoo acquired IndexTools and added a Yahoo audience insights feature called Interest Groups (which shows what categories the audience to your web site is interested in for better targeting), I see IBM combining tools to provide deeper insights about a web site's audience.

Just as Yahoo is dicovering ways to monetize the gold mine of user data they are sitting on, IMB is looking to monetize data also. 

about 8 years ago


Roslyn Layton

Great points, everyone, and Michael, your posting is the most informed and insightful I have read so far.  As a former employee of Coremetrics and leader of one of the services groups (our team were the power users of the platform), I have observed the many situations to which all the respondents have mentioned...office politics, employee bandwidth, too much information etc.  Two possible benefits for customers are (1) ability to leverage SaaS pricing if they take advantage of other IBM products and services and (2) increased power and platform for evidenced-based management.  You might know the story of Harrah’s,  a text-book example of business intelligence applied to every part of the organization and a testament to brilliant CEO Gary Loveman, one-time Harvard professor who turned a a sleepy casino company into the world’s largest hotel and gaming conglomerate by embracing analytics at every level of the organization.  As Michael rightly points out, though human expertise is important and essential to decision making, applying it to the scale necessary for enterprise operations, often fails for various reasons.  If artificial intelligence can be deployed in a meaningful way, then companies have a huge upside for their balance sheets.  Hence, the IBM-Coremetrics deal could offer value to the marketplace for the right customers. You can read more here.


about 8 years ago

Michael Feiner

Michael Feiner, Director at AEP Convert

ChL - no doubt integrating two tools is a big challenge and won't happen overnight.
Several months ago I was talking to SPSS and playing with the idea of pulling web analytics data into SPSS. It was very manual and time consuming (certainly impractical as a business approach). However, the results were pretty powerful. So I believe there will be great benefit in integrating the two tools in some way.

Ian - as I mentioned previously, I don't disagree that most companies are not taking full advantage of their web analytics tool or that politics are a big hurdle. You are right.
However, I'm unconvinced that the drive for change will come entirely from the Marketing Directors as you suggest.
Some of it, yes. But ultimately the technology to facilitate that change must be available to them before hand. Only a few of the Marketing Directors are thought leaders able to articulate their needs or aspirations to potential vendors.
Google's acquisition of Urchin and the subsequent launch of Google Analytics was arguably the biggest step change in web analytics in the past few years. The wide adoption of web analytics wasn't due to the Marketing Directors realising all of the sudden the importance of web analytics data (that played a role but was/is a gradual process) but for the availability of an easy to use free tool. There are other examples.

As a side note, I strongly recommend reading Kevin Hillstrom's post the 6 steps of channel death. There is some corollary regarding how Marketing Directors perceive marketing channels.

Charles -  nice to hear from you.
Yes, there is the integration with Websphere which undoubtedly is a priority for IBM but will not benefit most of us. However, IBM are investing heavily in their business analytics proposition. This proposition is not only a technology/software offer but also a consultancy arm. So I believe they are looking beyond the Websphere integration.
Last week McKinsey & Co, the biggest management consultancy globally, and Nielsen Online, the audience measurement company, formed a joint venture to "...help leading companies harness the power of social media intelligence to drive superior business performance." This is just another indication of a growing trend where large data analysis and consultancy firms are putting web measurement more prominently on their radar screens. IBM is indicating in more than one way that it wants to be part of this change.

Rosyln - thanks for your comments. Interesting to hear the views of someone that has inside experience with Coremetrics.
Tesco, the biggest UK supermarket chain, is another great example of using analytics to grow the business. The entire Tesco Metro (small convenience stores) concept was born out of the insight collected through the Tesco (loyalty) Card scheme. 

about 8 years ago



I love the concept of applying predictive tools to web analytics and online marketing data, but I have one question. If both of these already exist, what is the 'missing link' why this isn't already commonly done today? What is it about IBMs ownership of Coremetrics that suddenly changes the feasibility? If it is simply "integration", well that would have been done a long time ago, regardless of ownership structure.

I'm not trying to poo poo the concept; I truly want to know the 'missing link' if somebody has a guess at that level of detail. Good dialogue Michael. Cheers 

about 8 years ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.