Five "Ecosystem" Challenges for Web Analytics Practitioners

RoseIn early September we had covered the challenges that web analytics Vendors were facing as we look forward. The view from the other side, that of Practitioners, is very different.

This post covers the challenges that web analytics Practitioners and their employers are facing.

Quick Summary of the Challenges:

    # 5: Lack of relevant talent / skills

    # 4: Entrenched mindsets

    # 3: The web is dead (and what in God’s name is Web 2.0?)

    # 2: Its not about you anymore, its about them (Customers). Sadly. :)

    # 1: It is no longer “Web Analytics”

    # 0: One bonus recommendation (see below!)

Here are the challenges in some detail……….

# 5: Lack of relevant talent / skills.

This one has been talked about to death. Just the raw number of postings on various job sites and the age of these open jobs is an indicator of how hard it is to find relevant skills to staff a Web Analyst position. There are several causes to the dire short fall in skilled people….

Partly it is because other than the UBC course there is no formal training that is available. Partly it is because what we have been training thus far are report creation skills based on old style clickstream based programs, when the reality is that report writing and publishing has become increasingly less valuable and web analytics is no longer just clickstream analysis. Partly it is also because most job req’s posted demand five plus years of web analytics experience in a industry where some of the most sophisticated analytical thought is two or three years old. Lastly partly it is also because the space is expanding so fast.

Regardless companies will continue to suffer from the non-availability of real good honest analysts who know that it is not a sign of competence that your web analytics program implementation publishes 300 metrics, people who possess the perfect trifecta of skills:

  • Intellectual curiosity,
  • Raw business acumen 
  • Who “get the web”.

# 4: Entrenched mindsets.

Quick when was the last time you heard someone talking about hits on their website? Probably two minutes ago. And that is a problem.

Thanks to a mix of Vendors, Consultants and smarty pants bloggers (like me!) still dominating the teaching landscape there has been not enough mindset evolution in the industry and, more importantly, in the minds of common people and, most importantly, in the minds of company decision makers (CEO’s, CMO’s, VP’s et al).

Think DifferentPeople still ask for page views and hits and conversion rates and path analysis. We still have mindsets from the old world like overestimating the value of historical data (in a industry that is changing so fast, where you are probably turning things over very few months, is it really relevant what the path was last year or can we really compare time on site from three months ago to what it is now? not really). This is just one example.

The problem this poses is that Web Analytics sits at the kiddie table during board room deliberations. Most of the stuff we get asked to report is worth very little in terms of driving strategic action, which in turn means we get asked a lot less and we are starved of budget/money and valuable stuff to do for our companies.

Net net the web business suffers because it can’t be all it can be and companies can’t get as awesome as they want to be. Starting with Vendors and other Loud Voices we need to get out there and fundamentally change mindsets and show how much web analytics kicks ass so that decision makers will ask tough questions, we will measure and share real insights that drive real bottom-line value (compared to putting out “KPI’s” out there).

We have to to also cut out little bit of this vendor bashing that seems to always be on the surface every where and the FUD (Fear,Uncertainty and Doubt) that floats around as information, yes in the short term this vendor or that consultant wins but in the long terms everyone suffers.

# 3: The web is dead (and what in God’s name is Web 2.0?)

Most companies, most of the billion websites out there, still serve pages and static html experiences. Which means we have to measure the metrics that our grand-parents passed on to us from the log file analyzer days. Yet all around us spunky 12 year olds are reinventing the web and the web experience. It is not just Ajax and RIA’s (rich internet applications) and flash but more Web 2.0 type stuff like Social Media and Wiki’s and User Generated Content etc.

The web is changing from delivering pages to a platform for applications. This changes everything.

It also poses two challenges for Practitioners:

    A] Standard web analytics vendors are a bit behind the curve in terms of measuring these new experiences and making the fundamental changes to their software and computational architectures to accommodate for these new more “fluid” and developing worlds. If they don’t move beyond page based data capture and computation models it will leave Practitioners in a lurch in terms of options for solutions.

    B] It is extremely hard to measure success of these “new world experiences”, especially for now if you have to figure it all out yourself (see A). There are no pat answers, things to implement and guide books to follow. On top of that you have to ask hard business questions before you undertake some of these new ventures and define success metrics and ensure some kind of hooks are in place to capture data that can be measured post launch.

    This calls for a amazing amount of Business Acumen to be in place (not just number crunching skills) and application of Common Sense (say for identifying Critical Few). Both of these are not abundantly available in most companies. And that is a huge challenge.

# 2: Its not about you anymore, its about them (Customers). Sadly. :)

One of the amazingly positive trends on the web (and certainly a feature of Web 2.0) is that increasingly Customers control what they want to see, when they want to see is and how they want to see it.

Think of RSS (I just get text in my RSS reader and if you are showing ads I don’t see any), think about all the wonderful mash-up’s (not just the companies creating them but individuals hacking their own experience by pulling together the content they want), think of all the fluidity on the web that is within easy grasp.

We have been comfortable in a world where we have done path analysis to figure out how to “fix” the website experience so that visitors follow the path we think is best for them. We have had one size fit all customer experiences and web analytics measurement metrics.

In order to be truly successful, and stay ahead of the game, we have to truly get into the heads of our customers to understand their needs and where we are messing up and how we can give them what they want, when they want it in a format and medium they want it.

This means even faster acceleration away from simple clickstream analysis as the bedrock of decision making to measuring customer satisfaction and lots more experimentation and usability and integration of multiple points of data (including from customer experiences that we simply can’t measure yet) and lots more behavior targeting and knowing enough about our visitors (in near real time) so that we can react optimally to them.

This is a non trivial challenge and will require a fundamental rethink of our approach to web analytics, require super-Trinity, and a focus away to complex methodologies to measure Customer Experience.

# 1: It is no longer “Web Analytics”.

Web analytics is no longer a silo, neither is it becoming the center of the universe (much to the web analytics Vendors and our chagrin). In order to truly measure success we are having to accurately understand and measure what happens upstream (in tv ads and news papers and in blogs and in rss feeds and in magazines) and we are having to accurately understand what happens downstream (phone channel, retail stores, other websites etc).

This means data has to come into web analytics systems and data has to go out of current web analytics systems with much more fluidity. In many companies the actual determination of the success of the web channel happens outside the web analytics systems (maybe in CRM systems or BI tools on top of ERP systems).

This will mean a restructuring of the scope of the current web analytics systems, processes and team. This will mean a much larger scope, different roles and a much more complex problem to solve.

My dear friends René and Aurélie have blogged about this a couple of times: It is not web analytics any more. We don’t have a name yet for what it is but companies that have web analytics teams will not have them in the near future (I realize the irony that I lead a Web Research & Analytics team). Maybe it will all be Business Intelligence or, in the interim, Web Insights or Customer Analysis or Interactive Experience Analysis or How To Make Irrationally Loyal Customers Analytics.

As usual I do have one bonus challenge that Web Analytics Practitioners face…..

# 0: Vendors.

Perhaps one of the biggest challenge for Practitioners is Vendors. Here are some reasons (in no specific orders)….

    Web Analytics Vendors

    ~ Rough Seas: There is a new vendor in the space every week. That’s roughly fifty new vendors each year (and that’s perhaps conservative). The Big Three have been dominant for the last two to three years but with all this activity perhaps we are headed into a time of turmoil in the industry. Vendors will go away. There will be consolidation. We might lose data (not that data older than six months is very relevant). We might find truly innovative solutions. There are some rough seas ahead.

    ~ Free! Impact?: Impact of Google Analytics and Microsoft Deep Analytics (Ian: You have to give us a name we can use) free solutions. I am calling this out separately, even though it can be part of the above, because I do believe that these two guys have potential of being major disrupters for the industry and Practitioners.

    In both good and bad ways. They will put amazing pressure on existing players (which is good I think, maybe). They could become the Analog (Dr. Turner’s baby) of the future (“here’s awesome analytics bundled with every site for free”). They could over-promise and deeply under-deliver as they try to be everything to everyone and end up as other industry washouts, leaving many customers short-changed. The latter is less likely to happen, but it is a possibility.

    ~ No Standardization: Each vendor thinks they are Unique and insists in absolutely their own way of measuring even simple metrics like sessions and visits / visitors and session terminations or how to measure duplicate links or time on site or how to do sampling. For Practitioners this means data from one vendor is not comparable to another one, even for the same site. This also means that we can’t switch all that easily (or create standards or benchmarks for key metrics). I am on record imploring Vendors to embrace some standard ways of capturing and computing some basic metrics for the sake of world peace.

    ~ Progress = More Features: Our industry is nascent enough that Vendors are still on the path of showing progress by adding more features. At some point this has got to stop because many tools are already like Microsoft Word or Excel where 98% of the users use just 2% of the functionality. Practitioners will have to wait patiently until the industry realizes that progress can be shown, and next generation tool, by having less features and not more. That tools can have innovative ways of built-in base analysis features (not just more reports) and Personalization options that are simple and incorporate Right For Me metrics and computational abilities.

What do you all think? Is there a challenge you think is out there but missing in this post? What is your perspective on all this? What’s the next big thing web analytics Practitioners worry about? Is all of the above just scare tactics from a random blogger? Please share your feedback and critique via comments.

[Like this post? For more posts like this please click here.]


  1. 1

    I really want to read your posts since they seem very interesting, but my attention span really isn't long enough. Sorry if I sound like an idiot/ass – but the length of the posts is quite off-putting. Any chance of more bite-size posts?

    Apologies for my terrible grammar/writing style.

  2. 2

    Hi Avinash,

    Great post as usual ;-)

    The only thing I would add is patience. The practitioners need to have LOTS OF PATIENCE. I remember when we started WA at OX2, Aurélie was sometimes arguing with some of our clients because they didn't go the 'right way' (eg. requesting only some monthly reporting completly useless to take any action). She was frustrated and sometimes I had to go behind to extinguish the fire (if you see what I mean). With the years and experience Aurélie has learned patience and I think that it's one of the most important changes I've seen on her this past years.

    So if you're a practitioner sick and tired that nobody listens to you, don't give up! Time will prove you right.

    There are still too many corporations that are very far away from a 'data driven company', so I guess that many frustrated Web analysts are there trying to make a point within their organisations. Go little by little and remember that the force (insights) is on your side ;-) You'll see that with a little bit of patience and some internal lobbying you'll finish to get yourself heard.

    Avinash, don't get me wrong, this comment was for your audience. I know that your voice is very well heard at intuit ;-)

    Take care,


  3. 3

    At some point, web analytics folks need to become better integrated with the rest of the employees working in a business. It will be very important for web analytics folks to work more closely with finance, marketing, creative, merchandising, product development, and executives.

    A weakness that I perceive in web analytics is the view of customer behavior over time. Web analytics and practitioners do a great job of describing what happens within a session or visit.

    Web analytics and practitioners do not do a good job of explaining customer behavior across six or seven visits. We need for our tools and analysts to do a good job of explaining how four website visits cause a trip to a retail store, a trip that results in a purchase. This isn't the fault of a web analytics tool, it is simply the result of business systems not integrating data properly.

    When web analytics are fully integrated with the other systems within a company, things will get very interesting.

    Good job on your blog!

    Kevin Hillstrom

  4. 4

    Sina: Thanks so much for the feedback on length of the posts. There are perhaps two core reasons for the length:

    1) I am mostly trying to teach and like lectures and books I am trying to provide as much information and detail as I can so that it gives the readers all they need to form their own opinion.

    2) Some of the length is deliberate because there are too many blogs with pithy posts and sound bites that don't give all the context and hence add little value. I wanted to avoid that. It is quite possible that a good editor could probably chop the length of the post by half quickly. :)

    Couple random thoughts….

    1) I will try to add summaries on top of the post as often as I can. (I have added one to this post after your feedback.)

    2) There is a printer friendly link at the end of each post that will allow you to print and read later.

    3) I welcome feedback from other readers, if the posts are long and rambling then you all have to tell me so that I know this is a big problem. Also please give me suggestions on how I could break a post into multiple posts while preserving some integrity of the message (like for this post).

    Thanks again and I very much welcome your feedback and critique and help.

    PS: Your grammar and writing style were not terrible at all, they were prefect and bite-sized! :)

  5. 5
    Jaimie Scott says

    Hi Avinash,

    I don't see where this one additional item fits into your six, but maybe it does. I certainly hope that it doesn't become a major challenge for web analytics practitioners, but with all the blog and newsgroup posts in the past week I'm beginning to wonder. I'm talking about the FTC (or any other organization for that matter) regulating PII and internet data. In the current climate I've taken it for granted that I'll have good job security in the future and have never given a thought to the possibility of that changing dramatically — at least not until last week.

    Jaimie Scott

  6. 6

    Dear Mr. Avinash,

    Everytime I feel to thank you whenever I read your blogs, ideas, metrics and more. Really Terrific..

    This post has a great value as usual

    after joining as a web marketing analyst & passing Google adwords exam, I felt as I know more but after reading your blogs regularly, I rate myself at 2/10. You influence me a lot thru your blog.. Pls keep doing it.

    Thank you

    Anand K

  7. 7

    Hi Avinash,

    Thank you for contributing to the thought of world peace without participating in a pageant.:)

    As always, excellent post. And I'm glad there's a summary now as well. Looking into the future my concerns stems from lack of standardization and hence how the evolving web would be measured(2.0).

    I am still in an environment where numbers are sought without action. But the thought of analytics numbers plugging into BI tools so they affect decision making like they always should have, made my day.

    What are your thoughts on analytics as a career?


  8. 8

    Aman: I am perhaps less concerned about the lack of standardization and how we would measure the evolution. Mostly because we are in such early stages of the process and it promises to be a lot of fun figuring all this out. :)

    What are your thoughts on analytics as a career?

    MBA school answer: It depends!

    Seriously though I am very bullish on a career in analytics (regardless of where). I am especially bullish on a career in analytics on the web. Couple reasons:

    1) Imagine the power in your hands to fundamentally influence the direction of your business / company. Other than C level folks no one has that much power (and maybe product developers).

    You have data, you have brains, you have a passion for customers, you my friend have permanent job security.

    2) Specific to the web, the space is expanding so fast and everything is up in the air and promises to be a great ride. Through it all one thing will never go away: The desire to measure impact and if you are indeed solving for Customers (first) and your company (second).

    3) You can have a lot of fun. Of course in the right company and the right level of empowerment.

    Honestly I have a blast opening clicktracks with absolutely no goal except to "look" for stuff and as I slice and dice the data looking for patterns in the data it can be all subsuming. There is such a high when you find something interesting. It is addictive.

    Does that answer your question? :) [Full disclosure: As you can imagine I can't give a unbiased answer to your question.]

    Thanks for the thoughtful comment.


  9. 9
    Andrew Orsini says

    Hello Avinash,

    As a neophyte in the realm of analytics, I would like to say I favor the in-depth nature of your posts. I set aside at least an hour to read what you have to say, reflect and research the information you are presenting.

    I hope in time as I progress in my learning about web analytics, actionable insights, or whatever the future name will be, that I will be able to post more meaningful responses, but for now I thank you for what you have done in this community.

    Best Regards,
    Andrew Orsini

  10. 10

    Avinash – as always…fab post. Totally agree there is going to be a quantum shift in web analytics in the near future when urls & pages don't really mean anything. Its all about content and getting customers to where they want to be. So we'll see who steps up to the plate to deliver a way to measure the effectiveness of all this.

    Also, i agree with your piece on the trifecta of skills, which reminds me I need to ask for a raise.


  11. 11


    I know – the wheels of the Microsoft machine are currently turning, and will spit out a new name in due course. In the meantime, the project is codenamed Gatineau, so you can call it that (Gatineau is the city in Canada where DeepMetrix was based for some years, if you're wondering).


  12. 12

    Hi Avinash,
    If this page is still ‘open for comments’ … I have one

    [In many companies the actual determination of the success of the web channel happens outside the web analytics systems (maybe in CRM systems or BI tools …]

    You raised the issue of multichannel data integration. There is a lot of talk about the ‘customer holistic view’ or the 360 degree view. Is it possible to integrate weblog data into this view ?
    I guess it is feasible to integrate signed web interactions to the customer’s record.
    However integrating anonymous web interactions to the rest of Customer interactions, should be hard.
    Is there a practice to associate an anonymous cookie to a signed interaction and store this info for future use ?

    [Web analytics is no longer a silo]

    Does this mean that Web analytics should produce some aggregate metrics that can be incorporated into a multichannel BI dashboard or integrate all web interactions at the Customer level ?

    You mention in another post of yours that ‘data quality on the web sucks’. Are there any ways to overcome this low info quality and achieve web channel data integration ?



  1. Five “Ecosystem” Challenges for Web Analytics Practitioners…

    Business intelligence is a broad category of applications and technologies for gathering, providing access to, and analyzing data for the purpose of helping enterprise users make better business decisions. Some companies are making high end artificial…

  2. […] I came across this interesting post about the challenges web analytics practitioners are facing. Lack of relevant talent/skills was fifth on the list. That might very well have been first, though. I’ve found that few truly understand how to interpret this data – or what to do with it once it’s interpreted. I believe web analytics consulting will become bigger business in the years ahead as more companies realize they need what these stats have to offer – actionable information.

    Click here to read the original post about the 5 ecosystem challenges. […]

  3. […] According to their post, Microsoft Gatineau is the code name for Microsoft’s new free analytics tool. If you attempt to login with your Live ID, the system automatically kicks you out.

    I Googled “Microsoft Gatineau,” and the only result of any interest was a comment on Occam’s Razor: […]

  4. […] (click on the image to see the enlarged view) Anyway there is one comment from Ian Thomas – Director of Microsoft's Digital Advertising Solutions – at Avinash Kaushik's Occam's Razor holds the hint about this. And his Blog has announced about this on January 08th, you can read it here. […]

  5. […] Ian writes in his post —”it doesn't come as any great surprise to me that blog posts are starting to appear that have spotted that there is now something live on the Internet, as the observant amongst you will know, Gatineau. Earlier this month we took an early version of the code live onto the Internet as part of a closed Alpha program for a very limited number of our existing customers.” […]

  6. […] As the observant amongst you will know, Gatineau is the code-name for our forthcoming web analytics tool. Earlier this month we took an early version of the code live onto the Internet as part of a closed Alpha program for a very limited number of our existing customers – so no, I'm afraid you can't get a login ID to take a look at this stage. […]

  7. […] On a related note, last week I ran across a blog that Avinash wrote over a year ago – Five Ecosystem Challenges for Web Practitioners – still as relevant today as when it was written. In it he talks about the fact that web analytics is not a silo – web visitor data is tightly related to the “upstream” (tv, magazine, newspaper, radio ads), and the “downstream” (phone sales, retail outlets, other sites). […]

Add your Perspective