Wednesday, May 30, 2007

WASP 0.23, release notes

What is WASP?

WASP is the Web Analytics Solution Profiler, a Firefox extension aimed at web analytics implementation specialists, web analysts and savvy web surfers who wants to understand how their behavior is being analyzed.

How useful is it?

One of the largest sources of error in analytics is the accuracy and problems with implementing Javascript tagging. While some web analytics tools use one simple tag to collect data, others requires custom tags that can become overwhelmingly complex to implement. It is therefore important that you validate that your tags are capturing exactly what they are supposed to. WASP will ease the process of detecting the web analytics solution and showing exactly which values are being sent to the web analytics data warehouse.

Using WASP

  1. Install the WASP extension in Firefox
  2. Once you have restarted Firefox, from the Firefox menu, do View/Sidebar/WASP
  3. While you surf, the WASP sidebar will refresh with available information about the current page being visited

Detected web analytics solutions

Confidence levels refers to the avoidance of true-negative and false-positive detection by WASP itself and is really not related to the features or security aspects of any of the listed products.

Enhancements and bug fixes

  • Now detects 32 web analytics solutions
  • User interface improvements
  • More accurate refresh when browsing or switching tabs

Upcoming features

  • Display information about 1st and 3rd party cookie status
  • Display HTTP headers
  • Display P3P status
  • Put supported tools in a configuration file editable trough preferences
  • Allow sorting of value-pairs, copy (or export) of values
  • Add knowledge base links allowing to get additional information on any piece of data being sent
  • Handle frames and iframes
  • Add a tab to show Alexa and Google PageRank information about the page currently being displayed

Known bugs

  • None for now...

Your feedback is important!

Like it? Or not? Would like to see a new feature? Found a bug? I'm building this tool out of my free time and my reward is your feedback!

Make a donationIf you are inclined to do so - or if you use WASP for professional purposes - a donation would be appreciated!

WASP 0.23 released, stay tuned for updates

Last night (or rather, very early this morning) I packaged my Web Analytics Solution Profiler Firefox extension and submitted it for approval on the site. Usually it took a couple of days to get approval, but they have improved the process so much that I woke up this morning and the new version was already online! I was expecting to write about it tonight...

Stay tuned for a complete review of the changes and new features in this version!

Monday, May 28, 2007

Web Analytics: An Hour a Day

Your order has shipped

Greetings from!

We thought you'd like to know we shipped "Web Analytics An Hour A Day" by Avinash Kaushik.
At last, my order of Avinash Kaushik long awaited book, "Web Analytics An Hour a Day" as shipped. If you are not one of the thousand of people who discovered Avinash's great blog, you should read it from the beginning! And then order his book!

As I posted earlier, I stumbled upon Avinash's upcoming book while searching on Amazon a couple of months ago. At the time, it seemed like nobody had noticed it and I was excited to announce it on my blog... but Avinash asked me to keep it low profile.

It was a pleasure to chit chat with him at eMetrics, he is such a cheerful and easy to approach guy. If the book is anything like the man behind it, it's sure to be very interesting, pleasing and easy to read.

Tuesday, May 22, 2007

Omniture Café - Case study of DGAG

At the recent Omniture Café in Montréal, I had the great pleasure to present a case study about my employer, Desjardins General Insurance Group. If you attended, missed it, or are just plain curious, I'm making a slightly modified version of my presentation available to the world.

(If you are reading this post trough an RSS Reader, you might not see the presentation. Visit my blog to see it.)

(click here for a larger view)

Enjoy! And as always, your feedback and comments are appreciated!

If you would like to attend future Web Analytics events in Montreal, contact me!

Thursday, May 17, 2007

Code of ethics

I was reading René Deschamps excellent post about OX2 code of ethics and I thought I should state my own position toward the Web Analytics community, my blog and my employer.

The Web Analytics community

My involvement in the Web Analytics Association and various committees, as an organizer of the local Web Analytics Wednesdays and a speaker, and my relation with several practitioners, vendors, consultants and fellow bloggers. I will respect...
  1. Vendors: not do product demos or provide product information to competitors;
  2. Clients: any information about someone's actual or intentional use of a solution will be solemnly respected;
  3. Practitioners: empower the users with their actual solution without influencing them to change unless it’s in their best interest.
  4. Consultants: try not to interfere between a vendor-consultant-practitioner relationship. If it happens, the practitioner's best interest will always be favored.


I'm borrowing from O'Reilly's proposed code of conduct:
  1. Take responsibility for my own words and for the comments I allow on my blog.
  2. Won't say anything online that I wouldn't say in person.
  3. In case of conflict, I will connect privately before I respond publicly.
  4. Will filter comments that are against my personal values, but will gladly accept constructive comments.
  5. Will post constructive comments to other's blogs.
  6. Will link & reference others contributions in my posts.

My employer

As a web analytics practitioner and a Senior eBusiness Architect at Desjardins General Insurance Group, I will abide the following code of ethics, as described in the Desjardins Group Code of Ethics and Professional Conduct:
  • I will disclose all the details of the situation to my immediate superior if I think a conflict of interest might occur.
  • I will never accept or solicit gifts, tokens of appreciation or benefits, financial
    or otherwise, for myself, an associate or a third party, if accepting such gratuities impairs my objectivity or my judgment in the performance of my duties.
  • Will not engage into any contract without abiding by the prescriptive conditions of my employer.
  • Will not perform an activity or hold a position or a job with a competitor if doing so places me in a position that may prejudice the interests of my employer.
  • Will ensure that the activities I perform outside the scope of my job do not prejudice the interests, image or reputation of my employer.
  • When expressing opinions, I will not give the impression that I'm expressing the official position of my employer unless I receive prior authorization to do so.
  • Will never disclose confidential information or use confidential information for my own benefit.

Full disclaimer

Montreal Omniture Cafe - Wrap up

The first ever Omniture Café in Canada happened yesterday evening in Montréal. Speaking in terms of KPI, the ratio of registration to attendance was about 90%, with nearly 50 participants from current and potential customers. We were greeted with cocktails and drinks while we networked with fellow Omniture users. Then Marianne Llewellyn, principal for Omniture Client Services presented their service offering.

I had the pleasure to present a real case study about our experience with Omniture. Here’s the key takeaways from my presentation:

Our multi-disciplinary team is composed of key players from:
  • Marketing Research: core business knowledge & statistics, analytics,
  • eBusiness: interactive strategy,
  • Architecture: link between business & IT and
  • Web Development team.
On the 5 level maturity model proposed by Gartner (Bill Gasmann, "Web Analytics Expectations and Best Practices"), we identified where we stand: most of Level 1 & 2 and some elements of the other levels.
  1. Web Metrics
  2. Behavior Optimization
  3. E-Marketing
  4. CRM
  5. Corporate Performance Management
Our objectives, challenges, lessons learned and good shots were presented. On the positive side, the organization is quite mature in terms of processes, has high management support, and is able to allocate the right level of resources. Some of the risks are related to the temptation of giving away too much info without perfectly owning the data and its meaning, we need to make sure we grow from reporting to analysis and that we bring the analytical culture to all stakeholders.

The last point, reinforced by Omniture’s Scott Wells and David Mackay, was about the importance of establishing a strong user network. I also invited all attendees to have a look at the Web Analytics Association and the UBC courses, as well as get on my invitee list for upcoming Web Analytics Wednesdays in Montreal and Quebec city.

Some of the comments I’ve got were “I can relate to what you were saying”, “I didn’t know there was so many users in Montreal”, and “I look forward to get in touch with other users”. I guess I can conclude the event was a success!

Sunday, May 13, 2007

Predictive analytics

Like a cherry on top of the eMetrics Sunday Summit, I attended the course on Predictive Analytics from Eric Siegel. The slide deck of 200 pages served as a good reference for the 10 topics to be covered in just two days. The first day was easy; introduction and preparation, the most interesting stuff came in the afternoon of the second day, when modeling and deployment were addressed.


Companies need to invest in the process of doing web analytics. I will quote one of my favorite author, Tom Davenport:
At a time when firms in many industries offer similar products, business processes are among the last points of differentiation. Tom Davenport, "Competing on analytics"
Basically, predictive analytics tap on the collective experience of an organization to puts this knowledge to action. No crystal ball here, but some serious statistical modeling and business expertise put to work. The goal is basically to "predict better than the competition".

Killer apps

This topic is about the application of predictive analytics, not about predictive analytics applications (subtle distinction, but important!). Predictive analytics evolves around a problem solving approach:
  1. Set the overall strategy, outline the initiative
  2. Predictive modeling approach
    1. Determine the prediction statement
    2. Find out which data is required (segments & predictors)
    3. Deployment, how the model will be integrated or used
    4. Business case
  3. Evaluation; where KPI are determined, A/B tests are done on a control group and a baseline method of comparison is set
  4. Determine what are the challenges and bottlenecks anticipated: organizational, technical


Preparation of the data amounts to 80% of the work... The prediction goal drives the data preparation and is part of a larger process:

Univariate analysis

If I can't picture it, I can't understand it. Albert Einstein
The goal of univariate analysis is to see how well each predictor do alone. At the same time, univariate analysis provides good insights and serves as a double check for the implementation logic. Individual predictors will later be used as baselines over which modeling much improve.


For those who know about RFM (Recency, Frequency and Monetary), segmentation is the process of slicing, dicing and clustering the data. Here we talked about segmentation, clustering, OLAP, and data mining.


Starting from the concept that "history repeats itself", various modeling methods have been defined, some of them are better suited under specific conditions. Fancy terminology like decision trees, rote learning, naive Bayes, linear regression, neural networks, genetic programming as well as overlearning, underlearning or over-fitting were described.


This topic was all about "how well predictive analytics works". Lift curves were used to validate the model on test data.


Once the model is set and tested, predictive analytics is put to work. Deploying the predictive model brings recommendations for improvement, at the same time, new data is collected and fed back to the modeling engine, leading to further refinement of the model.


Predictive analytics is not a easy task, it requires planning and the right resources. Predictive analytics is a business activity, not an IT one. It's a wholly collaborative process driven by business needs and marketing expertise.


Predictive analytics allows for "per-customer" predictions and tactical achievement of strategic marketing objectives. At the same time, the organizational process ensures predictions are actionable and driven by business needs. Careful deployment of predictive analytics mitigate risks and ease performance tracking.

Note the words "Web" or "Internet" are nowhere here, and that's intentional. Although predictive analytics can be easily applied to the Web, it doesn't stop there. The concepts are far greater and can encompass a vast array of business activities. An interesting course that some participants found difficult, others not going deep enough. To me, the course was valuable and could have lasted another day in order to dig a bit more into the modeling, results and deployment aspects of predictive analytics.

Wednesday, May 9, 2007

eMetrics, web analysts and social behavior

While every other web analytics blogger seems to be talking about Google Analytics V2, I will post about something really different.

I already said how extraordinary my experience at eMetrics has been so far and all the great people I met. Then I started wondering what made this conference different from the others I attended in the past.

The Web Analytics Association

eMetrics is, in a way, the happy gathering of Web Analytics Association members, a not-for-profit organization. Of course the event organizers must cover their fees, and hopefully, make some money out of it. Great care is taken to have vendor-neutral speakers for the sessions, and except rare exceptions, the keynotes are genuinly informative rather than turning into sales pitches. The personality of Jim Sterne, the event organizer who happen to be the co-founder of the Web Analytics Association, certainly plays a big role in the "culture of eMetrics".

Infancy of the web analytics field

Ever since the first version of Mosaic was launched, parsing server logs and trying to understand what was going on has always been of interest. It was treated just like any other system log by IT until marketing really got interested. Other than the few gurus in the field, everyone is quite new to it. New comers and "old timers" have to rely on their peers to learn, evolve and innovate in a fast paced and mostly untamed world of opportunities.

Human side of web analytics

Web analytics is about understanding how human interact with an intangible thing called a website. Of course there are all those metrics, graphs and technicalities, but I think web analysts are naturally inclined to reach to their peers.

Social networking

Blogs, RSS feeds and the like creates social links long before people meet face to face. It makes it a lot easier to come up to renowned authors, practitioners and bloggers alike and talk to them. Going back just two to five years ago, this pre-event networking was barely possible, even impossible.

What do you think?

Tuesday, May 8, 2007

From the floor of eMetrics

What an amazing week!

In my 20 years of experience I've been to a lot of conferences, even early Internet World conferences when there was just a few hundred people, but never did I see such a truly open and sincere networking event. eMetrics is simply an amazing opportunity to learn and share about what is web analytics, how to get it right, and where it's heading.

The first time I met Jim Sterne was at Spring Internet World'96 where he was already pitching about marketing on the Web. I went to him and asked, with my very bad english, "how can I become a speaker and a consultant like you?". "Write a book!" he said.

Here I am, more than a decade later, blogging about web analytics and presenting a workshop on Website optimization in San Francisco. But it doesn't stop there, I also got to meet really intelligent and interesting people that I had only met in the virtual world. Who said blogging was only for the lonely associal freaks?

First, Kite flying with Joseph Carrabis and Jim Sterne was a genuinly fun and exciting experience, especially the part when cars and people were being chased down by the kite...

I also got to meet with he most brilliant minds in the field of web analytics. Robbin Steiff from LunaMetrics, June Li from ClickInsight, Craig Danuloff, Jim Novo, Aurélie and René from OX2, Ian Thomas from Microsoft, Jeremiah, Clint Ivy, Ian Houston, and many others, as well as practitionners from all around the world. I still have to chat with Avinash (I still want my autographed Google business card!), and Bryan Eisenberg.

In the coming days I will post about my best "take home" elements of the conference.

Friday, May 4, 2007

CMS Watch Web Analytics Report - revisited

A few days ago I blogged about the newly released report from CMS Watch: "The Web Analytics Report". Since I contributed to it, I was able to get a copy and quickly took a glimpse at it and read the first few pages. Other early readers also reported about it, all of us saying it looked great but didn't have time to read it.

Today I was on my way to eMetrics and had plenty of time to read it... plane took off one hour late, flew for 6 hours, and turned around another 45 minutes.

My take

If you are new to web analytics, or if you are yet to decide on which solution you will use, the first part of the report is an absolute must. But at $1,175 it's expensive for an introductory book. The biggest value is in part 6: Web Analytics Software Vendors.

Worth it?

Some people, like web analytics guru Avinash Kaushik, openly says if you don't have a product, you can start with a free one; Google Analytics or Microsoft Gatineau for example. So why spend the time and money on the report? The problem is once you're started, you are likely to stick with the free tool, or think all web analytics solutions are like Google Analytics and constantly refer to it as a base point. There are some very good reasons to look at other products right from the start. That's where the report will help you pick the two or three products that looks like the most valid fit for your business.

The product evaluation appears to be pretty good, at least for the products I know. And I learned a few things about them.


Funnily, the report says on page 3 that "it is interesting to note that a definition of Web analytics cannot be found on the Web Analytics Association website'... But right at the top of the About page of the WAA site it's there:
The Official WAA Definition of Web Analytics
Web Analytics is the measurement, collection, analysis and reporting of Internet data for the purposes of understanding and optimizing Web usage.
I think we can excuse this small glitch because in the end, if you are going to spend tens of thousand of dollars a year on a solution, you should probably start by doing your homework and gathering the most valuable data. It doesn't mean to go blindfold and use only the CMS Report, but it certainly ease the decision process, which amounts to saved time and effort worth more than the report price.

Tuesday, May 1, 2007

Q&A on web analytics

Yesterday I got a nice email from an advertising student at University of Oregon. She had just a few questions, but very good ones!

What's the most...

... important thing I look for in a web analytics product?

To me it's ease of use, which translates to segmenting data the way I want, get insightful reports that are accurate, visually appealing and easy to customize.

How do I...

... use web analytics in my job? What has it helped me do?

In one word, "understand". In 3 questions: where they come from, what they did, were they successful? (and if not, why?) Then I can communicate my insights (some would say "make a good story") and help decision makers take the best possible course of action.

Is there one...

... product I believe is best?

There is no clear winner (otherwise there would eventually be only one product on the market!). I can only talk about the products I've implemented or used. Coremetrics is great for ecommerce sites because it includes the notion of ROI in almost every reports, WebSideStory HBX Active Segmentation feature is really great for segmenting the data, Omniture SiteCatalyst is often considered to be one of the best (for a number of "small" edges over their competitors). Google Analytics best argument is obviously that it's "free". Webtrends was one of the early player and remain one of the few log-based solution. The newly released CMS Watch Web Analytics Report might help you decide which one is the best for your needs.

What sort of people or business...

... have the most potential as future web analytics customers? At what point does a business decide it needs analytics?

The field of web analytics is as old as the Web itself since logs existed from day 1. However web analytics as it stands today is still in its infancy, so there's lot of opportunities for those who want to make a career shift. I'm probably one of the few practitioners with an IT background, most people seems to be coming from marketing, and some from a statistical background. The "need" for web analytics most often comes from the marketing/communication department because they want to better understand how the money is being spent on advertising, and if the site is effective.

What has been...

... my worst experience with a web analytics product?

I wouldn't say "with a product", but rather "as an analyst". Failing to "sell" web analytics to a large organization that could really have benefited from it. They focused on the price of the solution, while I should have been a better salesperson and made them focus on the benefits of web analytics (and the risks of not doing it!).

... my best experience with a web analytics product?

Analysis of a local ecommerce site similar to Amazon.

Where do I...

... get information on web analytics?

The web analytics field is in full "ebullition", I read all the books I can (see the sidebar at the right of my blog for some recommendations), but I also track, on a daily basis, something like 50 blogs from US, Canada, Europe, vendors or consultants and practitioners, etc. I'm the organizer of the local Web Analytics Wednesdays, so I get to meet fellow practitioners from nearby, and I helping Omniture organize the first Omniture Café in Montréal, on May 16th. The eMetrics Summit I'm attending next week (and being the speaker for the Web Analytics for Site Optimization workshop) will be a great opportunity to meet "face to face" with some of the best people in the field.

Web Analytics Report by CMS Watch

A few months ago, I got an email from Phil Kemelor asking if I could take some time to answer a few questions. The result came out yesterday and I got my copy of the new CMS Watch Web Analytics Report. Nearly 300 pages that will guide you trough:
  1. The basics: what is web analytics?
  2. Sell it to your boss: business case for web analytics
  3. What to look for: Web analytics technology and features
  4. Purchasing and implementation
  5. Vendors: An in-depth review of the web analytics marketplace covering 17 solutions conveniently organized as SaaS, hoster or open-source and categorized based on the requirements, the website orientation, and internal resource profile.
Even if you are not planning on acquiring a new web analytics solution, the first five parts of the report are really worth a read. So far I just breezed trough it but I can already spot some chapters that I will point to my boss and colleagues.