The Future of Web Analytics

This week’s write-up comes courtesy of the loquacious and erudite @danieljtruman, a legend amongst wolf cubs and regular contributor to the glory that is #EcomChat. Over to Daniel…

The #EcomChat session on ‘Future of Web Analytics’ was based around 3 loose conversation starter questions. They were:

1. What are the big challenges surfacing around ecommerce analytics?
2. What do analytics teams need to do to react to this?
3. What interesting tools and technology are springing up?

This week we discussed the current issues surrounding web analytics within ecommerce. In any digital space it’s key to get measurement right, but in ecommerce it’s imperative. Lots of interest in this topic, with some key contributions from @Eoin_Kenneally, @mcmillanstu, @MonseCano, @OptimiseOrDie, @therustybear, @cjnmathews, @Magique83, @haydens30, @AlWightman, @AndrewBarker72, @hwjgage, @ACChaudre and @trbaldwin.

Q1) What are the big challenges surfacing around ecommerce analytics?

• @OptimiseOrDie Poor skills on top of poor instrumentation/integrity kills trust and utility of analytics for the ‘end user’ audience.
• @mcmillanstu I also think many businesses don’t care about statistical rigour “post hoc, ergo propter hoc”.
• @JamesGurd Agreed – not enough scrutiny, often due to lack of experience/knowledge.
• @danbarker A challenge for me: most analytics packages aren’t designed for ecommerce. You therefore have to shoehorn lots of basic stuff in.
• @danieljtruman eCom considered as strategy, but then analytics used as an add-on. Need to flip that mindset to get best results.
• @therustybear Internal Understanding of impact, Implementation (incorrect initially), perceived value, Incorrect Stakeholders, Incorrect KPIs.
• @AlWightman Significant shortage of skills in actually getting the meaning out of data and transforming it into useful insights.
• @Eoin_Kenneally Lack of investment in skills in the industry.
• @AlWightman Also related to the continued over-investment in tools compared to the under-investment in human beings.
• @JamesGurd Curse of ‘we don’t need analytics help, we’ve installed’.
• @danbarker ‘local carousel’ prompted, missing ios/android search, not provided, all challenges.
• @Magique83 Gaining sufficient buy in and allocated dev resource to implement more complex Analytics solutions (Universal, event etc).
• @cjnmathews Accountability – content producers think job done when content goes live. Agree with @therustybear on understanding of impact.
• @MontseCano That happens a lot when content producers are not marketers.
• @therustybear Can’t always blame the content producers – they have not been educated by marketers!
• @OptimiseOrDie That most people have broken configurations to a greater degree – very few people invest or improve analytics continually.
• @mcmillanstu I’m currently spending a lot of time looking at data integrity/credibility.
• @AndrewBarker72 always the problem with GA > ‘last click’ attribution, although they have started to provide other models.
• @ACChaudre web analytics has a lot to do with evangelising internal teams.
• @trbaldwin Resource is the biggest issue I see. I’ve worked in a e-com department of ~100 people with just one overworked FT Analyst.

BONUS Pop Quiz Question by @mcmillanstu (he did apologise…!): “pop quiz (sorry James & Dan) who has any analytics tabs open in their browser RIGHT NOW?”.

• @Magique83 Always have one open, all day, every day. Have an app on my phone & tablet to keep track too
• @danbarker (3 open here – quiet day!)
• @danieljtruman Only have 6 analytics tabs open right now
• @Eion_Kenneally Have it on a screen in the office so everyone can see what’s going on, they get a real feel for it then
• @OptimiseOrDie Good idea – @danbarker showed me ways to integrate maps with social and conversion data. Flags on map in office!
• @danbarker I fully agree with that – and one of the best benefits of ‘realtime’
• @jamesgurd Visible info can provoke important discussion that drives learning

Q2) Based on all of the challenges emerging around ecommerce analytics, what should teams/businesses do to react?

• @danieljtruman Plan. Work out the gaps in their data and understand the reasons why they’re missing data. Have a clear goal to fix.
• @jamesgurd Only let an experienced analyst go near config and set-up + if people need to play & learn, set-up test profiles.
• @Eoin_Kenneally Make a sizeable investment into training and development of their analytic teams and pull back reliance on agencies.
• @hwjgage Make it personal. Find out how each stake holder’s [job] performance is evaluated and work from there.
• @jamesgurd Make sure measurement is an objective for all non-analytical staff – part of their core role & linked to remuneration.
• @AlWightman I think the sharing of data, targets, KPIs etc is just as much about management/leadership/team work as it is about tools. Make it become second nature rather than occasional exception.
• @AlWightman Invest in more analysts (yes plural, I don’t believe 1 analyst has all the skills to offer a full analytics service in house).
• @Haydens30 Doesn’t this depend upon size or organisation? #ecomchat – not every company has the head-count to employ multi analysts.
• @AlWightman totally agree. I suppose it’s being aware when ur analyst can’t do certain things and then outsourcing appropriately.
• @jamesgurd key challenge raised earlier is getting people motivated to buy-in to analytics and using the data available.

Q3) What interesting tools & technology are springing up that make this easier (or is it really 100% about the people)?

• @hwjgage I reckon its c. 90% about the people. Doesn’t matter how clever the tools are, low level of interest = low level of uptake.
• @danbarker I’m with you. Surprising there aren’t more tools designed specifically to drive interest & ‘retain the cohort’.
• @mcmillanstu TBH, I think the most important areas are the people and the attitudes.
• @danbarker agreed again. & as @alwightman said, it then becomes about leadership/behavioural change/etc.
• @cjnmathews APIs are good for report automation. Also GA is very user friendly, particularly if you set up custom reports / dashboards.
• @danbarker Ometria is worth looking at (and out of London too.

A big thanks as always to everyone who took part and if you don’t see your name in lights up here even though you made a telling contribution, send someone round to kneecap Daniel:)

Please keep an eye out for next week’s chat announcement – can you contain the excitement?!

Thanks, James & Dan.