It was good to attend Econsultancy’s Digital Cream event last week. It was my first due to diary clashes, although it’s now in its fourth year. I moderated a roundtable on web analytics, which is one of my main digital passions, so it was good to see analytics was one of the most popular topics.

This post summarises the main challenges and gives tips on approaches the managers present are using to overcome them. The Econsultancy peer summits operate according to “Chatham House rules”, so there is no attribution to companies.

Challenge 1. From reporting to actionable insights

There was much discussion around how best to cope with “Information overload” and “analysis paralysis”. There was a common need to make subsets of information available to marketers who are non-analytics specialists who don’t use the systems regularly.

Practical solutions discussed included creating custom profiles, dashboards and alerts within the analytics system which related specifically to a market or category the marketer was involved with. We identified a 1 in 10 rule, that for every 10 marketers or digital marketers there needs to be one specialist analyst (or part of) to devote time to producing the tailored reports, optimisation projects and education. There needs to be a way to avoid this person being sucked into reporting only. Identifying the right culture for a rolling programme of optimisation of site areas, page types or marketing activities was another common approach.

Education was also seen as key with many trying to build in sharing of insights into the campaign process from making identifying learnings part of the briefing process through to always making some time for discussing learning in post-campaign reviews. Several of the attendees had produced a step-by-step guide to applying the software to their individual business. If you’re using Google Analytics, my guide to customising Google Analytics to your business may be useful.

Challenge 2. Selection of multiple web analytics systems

When we reviewed the main web analytics systems companies used, the paid systems (i.e. Coremetrics, Omniture, Site Intelligence, Unica and Webtrends ) dominated as you would expect from companies investing substantial amounts in Ebusiness.

But many were also using Google Analytics as a “free” system, so we discussed the need to standardise on a single tool. Some were taking this route to avoid the issue of resolving discrepancies in reporting methods and lower cost of training and support. Others felt that Google Analytics provided a useful tool for less-skilled marketers to review how well their campaigns or landing pages were performing whereas the power users .

This recent report from iPerceptions on Who Runs Analytics? iPerceptions shows that amongst retailers there is a trend to using more than one tool as also suggested by the Econsultancy report on measurement strategy.

Another trend suggested by the largest companies was the creation of a data warehouse for integrating data exported from web analytics with that for company systems. Visualisations were then created using business intelligence system like Business Objects or SAS.

Challenge 3. Evaluating social media 

There was a separate roundtable on social media, so we generally kept this brief discussing the tools that were most popular and selection criteria. Not much to add here, other than keep an eye out for the Econsultancy Buyer’s Guide to Buzz monitoring tools which is scheduled to be published in March or April. In the meantime, you may find this compilation by Michael Brewer and myself listing 25 social listening tools useful.

Another trend is the integration of reporting of social media into analytics systems – I mentioned the recent announcement from Webtrends about Facebook tracking and someone mentioned that AT Internet had an established approach for this.

Another interesting tool which was mentioned that I was unaware of was Rapleaf which appends social network data to your customer database to help you discover where your customers are across the social web.

Challenge 4. Improving search engine marketing

Again, there were separate tables on this, but many attendees had particular concerns on measuring and improving SEO effectiveness in particular.

We didn’t really get into the details of this, but several mentioned that additional analysis was required outside the main features offered by web analytics tools. In order, for example to evaluate multiple touch-points from search engines and integration of SEO and Pay-per-click through a gap analysis.

One tool which is interesting for visualising searcher behaviour is the free keyphrase visualisation tool from Juice Analytics (part of their more sophisticated Concentrate tool). Worth a look if you haven’t tried it, although I’m still making my mind up as to whether it’s a toy or an actionable tool – it does highlight problem areas though.

Challenge 5. Attribution of sale to individual digital channel and multichannel attribution

I’ve left this to last, since it’s certainly not a new challenge. I’d say there was a mix between those using the last-click win model and weighted attribution approaches. There was interest in the relatively new Tagman approach for measuring multiple touch points

There was also discussion on how to model and assess the impact of TV on sales. Some attendees had used econometric modelling which help develops a rule of thumb on the optimal mix of online and offline, but only really as a one-off analysis rather than an ongoing approach.

Other challenges

So what wasn’t covered that we expected to be? Well there was surprisingly little discussion on AB and multivariate testing, probably because there was an adjacent table hosted by Craig Sullivan covering this, although it was mentioned by several participants as a worthwhile activity.

If you’re completing a vendor selection I have just been alerted by @JimSterne of this new specialist AB or MVT tool comparison site.

One of the areas of analytics that excites me most is tools that elicit feedback from customers. These systems close the loop from most web analytics systems which show what customers, but not why. For example on my site and at Econsultancy we use Kampyle to gain feedback and this gives great granular feedback, for example, feedback on individual reports or problems with the site and this is integrated with Google Analytics. I notice that Kampyle now has a two-way integration with Omniture which enables marketers to be automatically alerted about on-site issues affecting customer service and conversion.

I’m also seeing more companies replicating the idea from Dell Ideastorm by creating open or closed customer communities where they give feedback which feeds through into new product ideas. Ideascale offers an increasingly popular solution for this.

If you’ve found some effective solutions to some of these challenges or have other challenges to discuss, do let us know.