Notes from Measurecamp 7, London

Last weekend I went along to Measurecamp, an unconference for folks interested in digital analytics.

Very good it was too. There were about 250 people there, all unfailingly friendly, polite and interesting. It was my first time there, and I found that even the old hands were up for meeting new people and giving useful pointers.

Everything seemed to go very smoothly indeed, which can only mean that there was a lot of hard work going on behind the scenes. So thanks to the organisers for wrangling everything so well.

Anyway, enough of that. I went to six sessions over the day, with a mix of presentation and discussions. I sat a couple of rounds out, mostly because you can have too much of a good thing – no sense in frazzling the old brain. Here are the notes that I jotted.

Overcoming barriers to the use of analytics – Peter O’Neill

Quite a nice opener this. Relevant to me because most of my clients are really just starting out with digital analytics. The slides are quite text-y, so you can read through them and get the gist quite easily.

Get round to reading Lean Analytics. Key takeaway is to pick one metric at at time and fix that before moving on.

If you’ve got lots of business units to deal with then it’s best to focus on one, make mistakes, get things right, then use them as a case study then expand out. Rather than spreading your efforts too thinly across all units at once.

https://prezi.com/6b_2ir64smcm/analytics-process-management-from-hypothesis-identification-to-organizational-learning/

See the slide about the diagnostic reports. Look at something like this every 3 months. If you can’t find 1 or 2 things to improve you’re not looking hard enough. Identify underperforming areas. Find places to dig deeper. Ask what’s causing issues. Useful for regular strategy reviews.

Use views vs add to basket to find underperforming product pages.

Do set up in phases, not all at once. Especially if giving instructions to devs.

The Data Layer – Simo Ahava

I didn’t take many notes in this one. Mostly, it was an explanation of what a data layer is and does. The slides are pretty self-explanatory though (if you like code snippets and technical stuff like that), and are here:

Bear in mind that the definition of a data layer’s purpose on slide 7 was prefaced with a warning that English isn’t Simo’s native language. Sheesh.

I actually found this one more useful than I expected. I keep running across data layers at the moment, so a higher level explanation of the various facets was handy, even if I was just at the edge of my understanding when going through the code.

Automating reporting from GA to Google Sheets or Excel – Gerry White, Sean Burton, Tony Lu

This is something else I’ve been doing quite a bit lately. Mostly to automate/speed up reporting, or find ways to get data to clients that they’d never access if they had to wade through the Google Analytics interface.

I ended up with a bunch of links to check out:

Measuring PR Barcelona Principles – Michael Blowers

This was an interesting discussion about the Barcelona Declaration of Measurement Principles. It’s an initiative from the International Association for Measurement and Evaluation of Communication, aiming to establish some elements of best practice.

I’d not heard of it before and there’s some interesting stuff in there. We had a chat around the ‘measuring outcomes, not outputs’ bit about how tricky attribution can be (in the sense of, knowing how much credit you can take for those outcomes when PR activity isn’t working in isolation).

Analytics for Product Listing Pages – Kelly Mclean, Alec Cochrane

This was interesting too. Especially because a lot of the people I work with have big lists of events or collections, with lots of ways of navigating around them.

It started with a big list of things you could be tracking:

With the caveat that it’s not worth tracking something if you’ve no intention of changing it (words to live by).

The UI limitations of web analytics tools was raised. General advice was to get the info into R (which free), SPSS or SAS to make sense of it.

Ask Us Anything (with Beer!) – Russell McAthy

This was a good session for decompression. A panel of people took questions from the floor, with plenty of discussion around things like ad blockers. No slides from this one. No notes either.