New website, new insight: analytics for a website relaunch

Rome wasn’t built in a day. That’s as true if you are talking about an historic city, a megalithic software project, or a web analytics implementation.

This week, I’ve had the joyous experience of pressing the button as part of a project team to launch a brand-new website and CMS platform. The possibilities for better efficiency and ways to optimise the user experience are exhilarating, not to mention the fact that the website simply looks modern and “pretty” now. These moments do not come around very often even in the lightening-paced world of digital, particularly if you work client-side.

Web and digital analytics are often overlooked components in projects, typically being briefed in last minute or – shudder – when someone needs some analysis done, only to find that nothing has been tracked. This project has been enlightening in so many ways: not only have I been involved from the beginning as a web analytics expert (allowing me to produce architecture documents and brief in well-structured data layers to my heart’s content) but I have been privileged enough to be on the core project team. This meant many decisions, from the big strategic calls to the minutiae of the page design and UX, being informed by digital data and insight. In a project now spanning well over a year, this is heartening to say the least.

I’m going to throw another cliche at you, though, and say that the real work starts now.

My web analytics implementation checklist has sixty-three items and counting, with only forty-five completed by the day after launch. These are not trivial tasks, each one requiring a mix of architectural thinking, toolset expertise, a dash of business logic, and the odd bit of custom JavaScript. That’s before we get into all the dependencies and working with a range of internal stakeholders and external agencies. We’re moving on from the bad old days of hard-coding and the general mystery around web analytics, but it’s never as simple as just throwing a tag onto a page. I’ve used much of that year-and-a-bit trying to reinforce how complex web analytics can be and keeping it on the agenda.

As I work through this checklist and feel satisfied that it’s all coming together in a modern, modular, clean, and easy-to-maintain analytics setup, the spectre of doing actual analysis hangs over me. However much effort the tagging work takes, that’s only the tip of the iceberg. The whole point is to be able to produce robust and timely insight in order to monitor and evaluate how the new website is doing. This in turn leads to recommendations and testing. The idea is to keep improving things, people!

I’ve found it very useful to break my tasks up into simple categories and manage these through Trello (other task management boards available…)

The categories I’ve used are:

  • Implementation (the bulk of the tagging and custom coding. Probably the most time-consuming part. Google Tag Manager is  your friend here)
  • Configuration (all of your tool config, Google Analytics filters, user management, etc.)
  • Testing (manual testing – on multiple devices – as well as tag crawls and back-end data validation)
  • Technical documentation (documenting exactly how everything is set up, both in the back-end CMS and in the front-end tools)
  • Stakeholder documentation (how-to on reporting and available insight, for less technical users)
  • Monitoring (automated dashboards, heatmaps, session recordings, and the like. Google Data Studio keeps winning in my eyes right now)
  • Ad-hoc reporting and deep dive analysis (answering all of the business questions and properly evaluating the site)
  • Measurement framework (what are the objectives and KPIs for the site? How are we measuring these and where is the data coming from? How does this align with other channels? The website has been informed by insight and it would be foolish to abandon that process now)

If that’s not enough work for you, start thinking about what data you may have to access via API, draft those scripts, and consider linking your digital analytics to your data warehouse or CRM system. Life’s a data dream!

Tempting as it is to bask in the moment and cherish those technical achievements, my reporting and analysis checklist only has around ten items on it, and that measurement framework needs to be dusted off. Oh, and the small matter of a global rollout lurks around the corner.

I’d better get going, as the data wolves will soon be at the door.