Digital Consultancy
Call Us : 01270 666 448

Econsultancy Digital Cream 2011

Econsultancy Digital Cream 2011

I moderated three sessions on Web Analytics at Econsultancy’s Digital Cream event on Tue 8th Mar 2011, and I’d like to share with you a few thoughts and common themes.

Complexity and scope

It was no surprise to hear that the scope of what web analytics actually covers, and what it means to a business is continuing to expand each year.  This brings a strain on resource and also adds extra layers of complexity to what is already a very complex discipline.
The main juggling act seemed to be getting the basics nailed (few felt they were doing this well enough), whilst needing to fulfil a heavy reporting / MI schedule and to make social measureable and prove ROI.
Quite a task, and this in turn had us talking about the resource issue for quite some time …

Resource – why are there so few good web analysts?

Everyone wants more great web analysts in-house; from people able to run reports, to project manage tagging through to delivering really good insights upon which the company can make decisions and drive change.
However, no one was finding it easy to recruit people, often with the search taking over 6 months for the right candidate or having to settle for second best.
So, what does the right candidate look like?
The hardest role to fill is the person who will drive insight.  This needs a person who is inquisitive, confident to ask lots of great questions and a good grasp of commercial reality.  So this is really someone who is a great web analyst and also has several years’ experience in a different role; such as acquisition, retention or direct marketing and they also need to possess great management and people skills.  This explains why companies are struggling to recruit because there is only a tiny pool of people able to tick all the boxes.

Data quality

Too many companies are still struggling to generate good quality data.  No one did regular (or even yearly) tagging reviews because there was insufficient time.  Also, too few companies validated their analytics data (such as orders and spend) with the offline transactional database during implementation and then to sense check this throughout the year.

Who actually owns analytics?

Most companies still had a way of working which often ignored the realities of the online environment.  Ie, they had an acquisition team separate from the web team, separate from the conversion team, separate from the marketing team, separate from the insight team etc.
Web analytics touches all these departments, yet because there was no overall cohesive plan or owner, whichever dept owned the analytics tool / web analysts / vendor contract, understandably tended to focus on their particular area at the expense of the others.  Which in reality means the expense is born by the company’s prospects and customers who do not get the best experience possible.

Cookie opt-in

A concern raised was the pending EU “Citizens Rights Directive” which may make opting-in to cookies mandatory.  This would devastate the industry with a massive impact on adserving, web analytics, multivariate testing, remarketing, recommendations, personalisation, email and so on.  I can’t think of a better resource to explain this topic in more detail than Dave Chaffey’s recent blog at Smart Insights.

Nurturing a better vendor relationship

Several delegates felt frustration with their vendors and felt they wanted more support which wasn’t forthcoming.
However during discussion it was generally agreed that to get a solution’s cost within budget something had to give and resource was often removed.  Therefore, the client didn’t pay for sufficient support, the vendor didn’t feel able to pitch sufficient support (because they would lose the sale on price), meaning that both parties ended up disappointed.
With hindsight, it was felt that much more upfront planning was required in order to include a realistic support provision and that allocating 1/3 of year 1 costs to implementation (ie people) seemed to feel about right for most companies.

Other web analytics recommendations and tips

  • If you are spending too long on report generation and too little time looking for insights, challenge reporting requests.  Try and understand the real question behind the posed question.  Ie, if someone asks for a particular report, ask why do they need it, what is their objective and what will they be able to change?  If they can’t tell you, then the report may have little value.
  • Stay in touch with key report users.  Make sure they are able to access the reports they need and interpreting the data correctly.  They will then become your best internal sales force making others aware of the benefits that your team can provide.
  • Share test results internally so people can learn by others best endeavours and mistakes.
  • Work closely with your vendor to get regular updates on new reports and white papers.
  • If you have a concern with your vendor, voice it rather than hoping it’ll go away.
  • Make sure someone properly understands your vendor contract(s).  What exactly are you paying for?  Are there any time limits on items have to be used?  What happens if you don’t use it; do you lose it or can you get a credit?  Also, make sure you are aware of any potential surcharges which you may incur, such as overage fees re server calls / page impressions, and review this each month.
  • Stay relevant to the business.  Constantly push your understanding of the business, your market and the online environment otherwise you may get left behind.
  • Make sure you have time to react to what your reports are telling you.  Too often reports and dashboards are issued with insufficient commentary and people aren’t spotting anomalies and investigating them.
  • Set SMART objectives for staff and agencies, and review them often.  For example, many people thought their agencies could work harder for them but because campaigns lack clear objectives and targets there was insufficient accountability.

It was a great day and lots of interesting opinions were shared.  I’d like to thank Coremetrics for being the table host and for all attendees for their open and candid comments.

1 Comment
  1. Hi Jonathan,

    On the data quality issue, perhaps it’s time to move beyond tags and generate the analytics directly off the web site traffic a la http://www.atomiclabs.com/pion-for-marketing/tagless-web-analytics.php

    It also puts the cookie issue firmly in the hands of the site owner, since tracking can rely on the first-party cookies generated by your domain (which the user is much more likely to opt-in to).

    In terms of “ownership” it brings control back to the analytics customer to decide where they want to store the data (e.g. in their own databases) or which third parties they want to send the data (e.g. webtrends, omniture, GA etc).

    This is a “one to many” relationship because once the page, visitor or session data is sniffed from the network traffic and converted by Pion you can send it to multiple analytics providers. Send it to Google AND Webtrends AND Omniture AND a SQL database AND log it in a flat file if you want. Each group can have the data “their way”!

    And less vendor lock-in… it make it far easier to switch analytics providers if you don’t like their service (or price) since you don’t have to re-deploy a new vendors tags.

    cheers,
    Steve

Leave a Reply