Planning for Analytics & Measurement Strategy During a Site Migration

Scott Strang / 8th October 2020 / Comment / Analytics

At Wheelhouse DMG, we are committed to being good stewards of our clients by looking out for their best interests throughout our engagement and beyond. We see this in action during site redesigns and migrations. This is frequently a high-stress and complex time for our clients as they navigate site designs, integrations, hosting implications and more, and unfortunately this can sometimes mean that Analytics is left for last, or even forgotten altogether.

Recently, I had the opportunity to chat with our head of Analytics, Anthony Coraggio. During a wide-ranging conversation, we dug into some detail around how to ensure that Analytics support is properly planned into site redesign and/or migration projects. I was curious to understand this aspect of Anthony’s work better, and we thought that the conversation that ensued could be helpful to anyone looking ahead to a redesign or migration in late 2020 or early 2021.

You can read a portion of our conversation below. This transcript has been edited for clarity.

Our Director of Analytics on Managing Analytics Support During a Site Redesign or Migration

I know that oftentimes we run into a situation where a lot of energy goes into design, site architecture, all of that, yet analytics feels like an afterthought. Can you talk a little bit about why that happens and why it’s important to try to avoid that situation?

Absolutely. I think there’s loads of opportunity in having analytics integrated throughout that entire process, because it really makes everything else that much smoother.

Even for something that might seem to be separate from analytics, such as design, it’s actually really powerful to integrate analytics into that as well. Because when you’re making those decisions, when you’re choosing what’s going to look different on the site, you’re also, whether you’re doing it consciously or not, you’re coming up with hypotheses about user behavior: you are saying that if I do this – if I do X, Y will happen.

And to actually be conscious and intentional about those hypotheses, and to ensure you pay attention to whether data supports them or not, that’s where you bring the analytics team in to really be able to prove your work out. And when you’re able to do that, for one, you’re able to much better prove the value of the work.

So that’s obviously important for doing a clear retrospective on a project, but how else does it add value?

You’re also much better able to understand what comes next. Nothing’s going to be perfect – not 100% of the choices made are going to be the best ones. But of course, that’s why we iterate and optimize the build over time and getting that data driven connection to the work and to the hypotheses behind the work is the first step in that direction.

And what’s the risk of waiting too long to incorporate analytics?

It can create a situation where the redesign is already done, and it needs to go live at this date, and the analytics work becomes “hurry up” work.

Oftentimes, as a marketer, your goal is to get something up quick for a campaign, and you’re not asking the question “how can I do this in the most complete way?”. Without asking that question, you encounter a lost opportunity to be able to actually judge the performance of the new site.

And we’ve seen clients encounter a lot of inefficiency or have to scramble or throw resources (and money) at a measurement or tracking problem last-minute because they realize a little too late that they’re not setting themselves up for clear, actionable data with the way they’ve planned their site migration or redesign. There’s no plan.

So what does it look like when it’s done well?

So what we want to do in a site migration is to make that plan ahead of time for you: these are all the hypotheses and choices and business decisions that we’ve made about how we’re going to change the site and here’s how we are going to judge if that’s effective.

Of course, to judge if that’s effective, you need to be able to compare pre-versus-post. And if things are changing really substantially, you may need an entirely different kind of schema and approach for how to share those that data into two pieces that you can actually compare. And you might just be missing the data to do so as well.

I’d imagine at this point that the expertise of the folks involved with the project becomes important when you’re trying to determine those finer points, right?

It takes a skilled analyst to fragment seasonality, trends, noise, etc. Testing your hypotheses is rarely going to be as simple as pulling up period-over-period data without any nuance or filtering. Even if things are going “up and to the right”, you need to have a precise understanding of what that really means functionally, and you need to be able to determine what your ROI was for all the discrete elements of what you paid for with a redesign or migration.

It’s also important that whoever is running the project has a good fundamental understanding of how tracking can be “broken” by deceptively simple seeming changes, or at least that they’re able to consult with someone who has that knowledge and are disciplined about doing so. Changing site elements on the front end is usually looked at by the project leads as something that’s pretty lightweight. But from an analytics perspective, you’re extremely dependent on front end code: any kind of Custom Event Tracking, maybe even some of your fundamental tracking on site is going to be dependent on triggers, based on the current CSS ID values – attributes on the front end of the site.

The Wheelhouse Way

If you’re thinking about a site re-design or migration, Wheelhouse can help. From planning and consulting to architecture, tagging and reporting – we’re here to support. Feel free to reach out with any questions you might be having!

By Scott Strang