{{ searchResult.published_at | date:'d MMMM yyyy' }}

Loading ...
Loading ...

Enter a search term such as “mobile analytics” or browse our content using the filters above.

No_results

That’s not only a poor Scrabble score but we also couldn’t find any results matching “”.
Check your spelling or try broadening your search.

Logo_distressed

Sorry about this, there is a problem with our search at the moment.
Please try again later.

I think a common theme in companies is that there is a need for an 'expert' to come in and make use of the data that they are creating or to turn around an organisation that has managed to ignore its customers for too long. 

This is the start of the web analytics function in an organisation. Most people don't know where it belongs, let alone what it should look like. In my professional life I have been in e-commerce, IT, search, customer insight, and marketing teams, and I'm only 29 (even if the increasing grey hairs belie this fact).

What this tends to mean is that you, as the newly appointed Web Analytics expert, can decide how you want to structure the entire thing from scratch. Which is possibly one of the most daunting, yet wonderful things that you can do. This post is here to give a bit of help.

Eric Peterson talked about how you should set up your analytics function a couple of years ago in a 'hub and spoke' model (following it up in part II with completing the communication strategy), but I want to go a bit further on this.

Web Analytics Team structureThe hub and spoke model for web analytics works on the principle that you as the web analytics team are going to be the ones who are responsible for some of the things, whilst you are going to outsource, for want of a better word, responsibility to some of the teams. This is going to allow you to spend more time doing actual analysis and less time doing reporting.

What does the analyst have to do? Well they have to deal with all the management and producing reporting for the senior stakeholders to ensure that they are capable of making the informed strategic decisions.

There needs to be liaisons with the IT department to make sure that the tagging is up to date. This should be done by the central team so that all the benefits can be seen by all the users of the systems.

The Marketing team need to ensure that they have a close plug in to the analytics function so that they can work out the ROI on their campaigns.

Finally you need to ensure that the content owners (or business owners of the products depending on your business model) are on board with the whole process as they have ultimate sign off on the most important bit of the pages.

Centralised reporting and analysis

There are two ways in which this can be done.  You can take ownership and access of the tools completely away from the teams.  This will enable you as the analytics function to have full control over the data and recommendations that are produced. 

This has a couple of advantages:

  1. You are the expert, you should be able to provide the best analysis and recommendations.
  2. You have a global picture of the site and can link that analysis to overall KPIs.
  3. You know the system inside out and should be able to do it quicker than anyone else.

It also has some disadvantages:

  1. Nobody likes being told what to do.
  2. Your stakeholders may need the insight now, whereas you do not have the resource to do it now.
  3. Your stakeholder nods with fervour to your suggestions, but doesn't enact them because they don't feel engaged in the process or don't understand the implications.

This centralised model works best for websites that are set up to be slow moving entities which can't change that frequently (in my experience). This usually means nothing actually needs to be done instantaneously, so you can resource your team for a flat rate of work over time.

Just out of interest, I think this is the route that research teams have tended to follow for a long time. I can see why as one of their biggest things about analytics data is the availability of it now, whereas one of the downsides of research can frequently be that you have to go out and collect it (from various different surveys). This means that disadvantage two above becomes a problem however you set up your team.

De-centralised reporting and centralised analysis

For faster paced organisations (those that run in sprints via Agile development cycles and media based sites who upload content frequently, for example), it is sometimes not the best situation to be in where your central team has to do everything because they may not have the resource to do it. 

Without getting more resource is to outsource your reporting to the teams that need the reports, giving them access to your tools. This means you can create a series of 'superusers' of the tool whose job it is to do the basics themselves and ask the analytics team for help where they can't.

Hub And Spoke with decentralised reporting

This has some major advantages:

  1. Freeing yourself from (boring) reporting allows more time to do (exciting) analysis.
  2. Engagement with the stakeholders in the process (they've produced the data themselves, so they'll help come up with ideas).
  3. By getting the stakeholder closer to the data they are more likely to understand it.

There are some downsides to this method though:

  1. You spend more time training people on how to use your web analytics tools.
  2. You spend more time supporting people on how to use your web analytics tools.
  3. If your analytics system isn't intuitive and the implementation consistent, there will be lots of confused non-analysts.

Of course for some organisations that continue to get bigger, different parts of those functions are going to get increasingly or decreasingly important. 

As you expand it might be that the content owner (or the person who is assigned analytics responsibility in the content team) may have to progress closer to the web analytics team and start doing analysis as well. In this case, they may even devolve reporting out again.

This is effectively the point where decisions need to get made: Can your analytics team support lots of stakeholders (how many will depend on the organisation) and how big can the analytics team get to do this; or do you bring in your superusers closer in to the analytics team, giving them more access and more responsibility and have a group of 'Clark-Kent-users' who can go to the superusers as first point of call with any queries on the tool.

I think this is a stage that many organisations are now getting to as the web evolves and I am looking forward to working in new and exciting teams who get to decide this!

Alec Cochrane

Published 22 October, 2010 by Alec Cochrane

Alec Cochrane is Business Consultant at Adversitement and a contributor to Econsultancy. He also blogs at whencanistop.com

5 more posts from this author

Comments (7)

Comment
No-profile-pic
Save or Cancel
Avatar-blank-50x50

Mark Jones Jr

These are some interesting perspectives on how to set up such a team. I think that we have overlooked the most important aspect, however, and it goes directly to how we view ourselves. If I am just an analyst, then the composition and organization of these teams is probably sufficient. If I am a leader, responsible for the leaders I work with and for, invested in the decisions they make, accountable for my advice, then I need to be engaged with the leadership team, equipping them with the tools they need to understand and make decisions, helping them set an attitude that will achieve the desired performance (like a ship or an airplane), and developing myself and them as a more informed leader. All of these things can be done without as much focus on structure. Personally, I think Venn diagrams would better illustrate the true relationships. Keep your analytics REAL: Relationships-Equip-Attitude-Leadership.

almost 6 years ago

Alec Cochrane

Alec Cochrane, Head of Optimisation at Blue Latitude

Thanks Mark!

I didn't go into too much depth about what an Analyst (or analytics team) does, because I think this depends on the organisation and the maturity of it. In some organisations you need to build relationships before going into deep analysis and recommendations, whereas in other organisations they may be ready and willing for those in the first place.

I'm also didn't want to get into a debate about where it should sit in the organisation because, again it does depend on many things. I don't think any set up is right or wrong, what we need to do is make sure that is the most effective for the organisation (and for you as well, because nobody wants to stagnate).

I think that as with many things relating to the web this can be trial and error to work out what works best. Definitely lots of food for though.

almost 6 years ago

Avatar-blank-50x50

Carson

As web analytics programs get easier to use (still wating for you, Omniture) , we'll likely see more movement to scenario #2. Super user idea is a great one -- I've started to train some folks on more advanced analytics reporting and it's helping a lot. While it may take them more time to pull numbers, I believe that to really understand web performance you need to use analytics software yourself. It gives you a more intuitive "feel" for the numbers that  can't be attained from static reports. Users are now asking better questions, which allows me to do more relevant analysis.

almost 6 years ago

Avatar-blank-50x50

mackaycs

Good article Alec. For me a key factor is type of your website and how dynamic it is.  I'm currently working with a highly transactional fast moving site, so we try to devolve basic reporting wherever possible and focus our efforts on deeper insight that either helps colleagues make the right decisions or finds key issues and areas for optimization.

For example, a significant majority of our effort is in tying web analytics data to back office transactional data, which is more specialist and manual.

over 5 years ago

Rob Mclaughlin

Rob Mclaughlin, VP, Digital Analytics at Barclays

The de-centralised appraoch is likely the most effective - analytics really needs to be in the hands of the person who requires insight, whether they are in marketing, HR, operations, logisitcs, UX, design etc.

I feel there may be a valueable convergence occuring as many digital natives understand the concepts and structures that surround web analytics, they do not need the 101 training that those who were not born digital require.

This is not to say they do not require guidance and consultation in order to access the insights they seek. With this group there is actually a really exciting aspects of 'needs capturing' that can be picked up by an analytics team - understanding what the wider business demands from analytics and evolving reporting and analysis to reflect it.

A combination of training, encouragement and needs analysis seem to be the combination of support that will enable most businesses to report, analyses and act on their data.

over 5 years ago

Ed Longley

Ed Longley, Head of Direct Online at Hiscox

Nice article Alec, and I hope your grey hairs are kept to a minimum - Boots have an excellent range of colouring products.

I see from your comment to Mark that you've kept it top-line in terms of what analyst does, however does a key relationship exist with the internal corporate analysis/ MI/ data warehouse team who are in possession of the data that is frequently used as the "truth"? Web analytics data, being what it is, is unlikely to marry with this data. Which comes to the next point, which is language and communication. Web analysts need to be highly capable communicators to explain their craft and de-bunk myths.

over 5 years ago

Alan Williamson

Alan Williamson, Director eCommerce at Goldmedal Travel part of the DNata and Emirates Group

Nice article which has certainly given some food for thought. I think content owners can get the info they require day to day through productionised dashboards. Whilst time consuming to set up only a novice level of experience is required to then make use of them and gain any additional simple information required. I require all my team to have a basic understanding of our web analytics package.

Taking this approach frees the analysts to spend time digging deeper into the data and determining issues and opportunities that don’t obviously present themselves.

However no one size fits all here, in my opinion.

over 5 years ago

Comment
No-profile-pic
Save or Cancel
Daily_pulse_signup_wide

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 Daily Pulse newsletter. Each weekday, you ll 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.