Monday, July 27, 2009

Web analytics– where should it sit in the organization: in business or in IT?

Where do you think that Web Analytics should sit in your organization?
In this post, I would like to talk about organizing Web Analytics in a company and the important question : who should own and lead Web Analytics competencies? where should these competencies sit in the organization? And what if Web Analytics all started from the IT side?

As a practitioner and based on my personal experience so far, my answer would be: not in IT.

But why do I believe that IT is not the best option for managing and leading Web Analytics competencies within an organization?

Web Analytics relates to business
Basically, Web Analytics is about improving business performances of online channels. This is done by understanding what are the business objectives, by defining appropriate “business-driven” KPI’s, by analyzing and interpreting the figures and taking necessary actions & business decisions.

But do not get me wrong, IT is key player as it is needed for all technical aspects such as hosting and maintenance of the Web Analytics infrastructure (if you are hosting your solution – not the most common set-up), definition of the technical/tagging specification based on business requirements, implementation of the measurement code in your CMS, websites and online applications. And if you integrate your Web data with other systems (BI, CRM...), IT is the one that defines and implements the required architecture & technology.

But in the end this is the business that drives the needs and that uses the data to gain insights. That is the business who takes decisions. That’s the business that normally defines the analytical strategy & long term vision (as it relates to what business will do).

Can IT have “business” mindset”?
Web Analytics requires a strong business mindsetAn important trait of a good Web analyst (or expert) is the “business acumen”:
“Good web analysts know the business and the current business climate. They understand the business purpose and can identify and focus on the business problems and disciplines apart from analytics. They have both a great sense of the data and the business.”
Basically a good Web analytics expert needs to have a strong “business mindset”.

Having “business mindset”, is it something that IT can have? I mean genuinely. I don’t want to be harsh on IT departments but let’s face it, in many cases, IT department mission is to implement technologies, tools, systems, to develop & deliver applications, to make sure everything is up & running and all.

I am talking about personal experience here as I am working on the IT side (no one is perfect :-)). I started as business analyst and I took over the requirements analysis, the reports creation and the coordination of measurements implementation when my company insourced Web Analytics. From there I stepped into the world of Web Analytics and I started developing competencies & expertise while working closely with business.

Explaining to IT why having Web Analytics competencies sometimes feels like this Dilbert cartoonAs we moved from Web reporting to Web Analytics, business demand grew – requiring to make Web Analytics competencies more structured and organized. But doing this in IT is far from being easy. You need to spend a lot more time in order to explain the purpose and necessity of Web Analytics (and why having a dedicated centre of expertise). Already a challenging task! And if you succeed, the next question you may get: “a large part of the competencies are purely business so why should we offer such expertise in IT?”.

Centralization is key!
The main advantage of IT is that it usually has a central view of the needs from the various departments – especially in large organizations where business departments are often structured in silos. I believe that Web Analytics expertise & competencies should be centralized in order that standardized solutions & technology areused by the different departments and to share best practices (all will reduce costs in a significant way while improving efficiency).

Web Analytics should be lead centrally by a team that supports the different business departments (who need and use the data) and that makes the link with the technical side (i.e. IT, vendors & external agencies). A team that would assist & coach business key users in how to better use Web Analytics. I quite like the “Hub’n’Spokes” model proposed by Eric T. Peterson. This is something similar that I wish to develop.

But building such central team in IT is difficult as it mixes business expertise and practices with technological expertise. In most of the examples I have seen from other large companies, Web Analytics competencies are regrouped in a dedicated team that sit on...the business side.

So it is no surprise that I feel like sitting on the wrong side. “So why don’t you move to the other side then?” you may ask. Well, it is easier said than done.

Nevertheless, I am doing my best to improve organization in order to better serve the company growing needs. But changing organization and mindset in a large company is quite a challenge and takes time. Lot of time.

So what is your experience? How is Web Analytics organized in your company? Centralized or decentralized? Do you know examples where IT ownership of analytics does work? I am very curious to hear your experience and it would be more than welcome.

Other resources:


  1. We have a Web analytics team in-house and we coordinate with IT on getting things done. The marketing part is outsourced to various agencies. We work with business hand in hand to get the reporting and analysis done.

    Ankit Nagarsheth
    Toyota USA

  2. Hi Michael! Most of our larger clients with many web sites or web activities have pointed out one person (or in some cases a small group of people) to coordinate their web analytics practices. This person would then become a "power user" of the web analytics tool, have contact with the vendor and be responsible for helping others setting up reports. Ofthen this person would also be the main driver in setting up some standard reports which are distributed weekly or monthly to different (national) web teams. These temas are typically new to web analytics, but need feedback on their work with online marketing and web site editing. How is your work in Toyata? Are you also responsible for serving others with reports? Or are you mainly helping yourself optimizing "your own" web activities? Thanks for a goo post!


  3. Michael, Very well written! From start to finish I had to keep reminding myself I was reading someone elses situation & history - living the identical world. Analytics is a business function, few would put their marketing machine & intelligence in IT. Completely agree too on the centralized view of this function particularly in large organizations that often resemble mini businesses under one roof. The key for us was recognizing this situation and equally recognizing that on the other side was One Customer trying to look at us as One Company - Analytics is the voice of that customer. If done well, with a One Customer strategy in mind, it's a huge opportunity for large organizations! Equally this can be significant challenge when the function has no home! Uncanny the vertical you are in too. Thanks

  4. Hi Michael. Much food for thought there! I tend to agree "not in IT", but think there are exceptions depending on the company. Sometimes web analytics is best placed where-ever one can find an executive champion.

    Good post!

  5. @ankit
    Thanks for sharing your experience. Glad to hear you have a dedicated in-house team over there. But from an organization point of view, where does your team sit? I would be really curious to know more…

    I am actually doing all you described and more. I am responsible for not only some technical aspects such as having Web Analytics implemented based on business requirements, coordinating with external agencies & vendors, defining guidelines & in-house standards, managing vendors but also doing a lot of “business” related tasks: assist business in defining appropriate KPI’s & in their implementation, analysis of results, recommendations,… But as you stressed in your comment, larger companies need to appoint dedicated resources for coordinate Web Analytics practices in order to manage and develop these efficiently. Fully agree.

    Thanks for the comment. Always good to know that you are not alone to go through such situation :-) And thanks for the advice. I will take this into account.

    I am pretty sure that there are examples out there where WA sits in IT and it works fine. The problem is getting executive support – something I think is less straightforward when you are in the IT and where the Web is not (from IT perspective) the most important area of activities. Not impossible, just thougher in my opinion.

    Thanks all for commenting!

  6. Well, my take is that we could consider a hybrid between the 2, and needs to be supported by marketing...

  7. I tend to favor a decentralized "department", where many different roles are using "analytics" for different purposes. Especially in online news media, journalists, editors, frontpage editors, UX, IT, administration may all want to know what's going on on their website, but for different purposes. Why not make the analytics tool fit all user groups, eliminating the need for an analytics department? Dangerous words perhaps on a web analytics expert's blog :-)

  8. @web analytics & @Tom Erik:
    Thanks for sharing your thoughts. Really interesting.

    Fully centralizing all Web Analytics is not the best solution. Eric T. Peterson's model (Hub'n'Spoke) goes more in direction where there is a central hub that have a global vision, that supports other departments and provide them services. The spokes have their own resources that are "using" the output i.e. data, insights, who are making decisions, who tracking KPI's...

    The Hub would help these resources by coaching them, by analysing their requirements...

    Also you need to have a central view to avoid disseminating costs (having different solutions for same purpose) and promote best practices among the company.

    In his Web Analytics Maturity Model, Stephane Hamel promotes the idea of a Web Analytics Competence Center (WAAC), typically a multi-disciplinary team that mix analytics, IT & business skills.

    Whether this central hub/WAAC sits in IT or business is not most important. The fact it exists is the important point.

  9. The front end of our website is transactional meaning that a key focus is on the UI and UX (to use too many acronyms!). Therefore my analyst team sits directly with the product teams that make the most complex decisions such as account management etc; and work closely with design/dev to develop the user experience (UX).

    The most important thing in terms of centralization for me, is that the tagging is managed (preferably and executed too) by the analysts - learning a bit of javascript never did anyone harm.

    This is because web analytics is only as good as the data that you collect, uniquely compared to back office data for example (where a "new account" or a "purchase" is pre-defined), you can collect whatever data you want and therefore it's the analyst who should be close to this activity in order to ask and answer the right questions.

  10. @mackaycs: thanks for sharing your view and experience. Very interesting example. I fully agree that a good analyst should be a bit technical savvy to ensure quality measurement and proper interpretation.

    But this aspect is often separated from the business side - relegated to IS / dev team. Not best scenario in general...



Feedback Form