The Ultimate Guide to Using Google Analytics for Cross Device Optimisation

Android Device Fragmentation 2015 — @opensignal

Optimisation is what I do for a living.

I’m always hunting for actionable, easy, quick and high revenue things I can test or just push live.

This article is one of the most surprising things I teach to people — that it can be so easy to find stuff worth so much money.

Broken device experiences are almost always — the EASIEST things to demonstrate to clients and one of the EASIEST to get them to actually EXECUTE. It’s normally pretty straightforward — if you can demonstrate the issue (with a video or screenshot) then people will take notice.

This veritable tome — is a guide to over 7 years of refining my approach to finding device experience gaps.

Here is the index, so you can browse the entire series or dive straight in. Parts 4–7 are coming shortly.

TL;DR?

If you know how to do the analysis, try parts 5–7.

If you haven’t been through a detailed analysis, waded through heaps of User Agent strings, checked device databases or questioned every metric Google Analytics serves you, then don’t assume you know what’s truthful. Read all the parts.

(1) Optimising the Device Mix

We introduce the concepts and thinking behind understanding the true picture of device usage on your site(s)

(2) Tablet Device Analysis

We analyse the composition of Tablet Devices (hands on).

(3) Desktop Browser Analysis

We analyse the composition of desktop browsers (hands on).

(4) Mobile Device Analysis

In the final device class, we finish up with the mobile devices (hands on).

(5) Google Analytics Templates

A summary of all the report configurations you’ll need to build a complete device experience flow model. A short version for the tl;dr audience.

(6) Finding the Missing Money (coming soon…)

How to read the reports you’ve produced — to draw up a small list of device experiences that are leaking money.

(7) Testing Your Shizzle (coming soon…)

Hands on explanation of what tools to use with your testing list — and how to uncover the problems that you see in the analytics data. Once you have a final list, we show you how to prioritise everything you found into an actionable plan!


Training and Mentoring

I am not an agency. I have no interest in ‘scaling’ — a euphemism AFAIK for charging the same money for delegating this to someone else more junior.

My mission is to transfer enough skills to clients, so they don’t need me to do the stuff any more. In practice, this means longitudinal mentoring, oversight and support over time. My clients then don’t need my time so much, they mature and they then ask me more strategic questions.

If you want some skills transfer to your team or a training session on the analysis or testing techniques here — Abi and I would be happy to help.

@OptimiseOrDie or

Linkedin : linkd.in/pvrg14

Acknowledgements

This was a freaking nightmare to write — because I usually did this stuff without thinking about what I was doing. It was quite hard to explain quickly or easily.

I then had to think really carefully about how data might look on different setups, so I tried it against a heap of my clients to validate the model. It took about two weeks of work to draft, iterate and check.

Putting it into a sensible set of parts and giving you a repeatable exercise is my pleasure though — I hope it’s worth every pound of missing money you find, that you never knew was waiting.

Tell me your stories!

Thanks to Abi Hough, who has been my ‘Wingman’ on many testing missions over 10 years, with superb kill rates. Abi is better at finding flaws in your stuff than anyone I know (yes, including me on a very observant day). If you want her to, she can train your team to do this anytime they like. Couple it with this guide, and you have a money sniffing machine.

Thanks also to the huge list of my analytics, cro, stats and GA friends and experts — who contributed through comments on the draft, my questioning and lots of arguments. Thank you all.

Most of you didn’t know, but my wife has had a tough rollercoaster ride with cancer since March last year. I remain positive, despite everything and with confidence that time and science will prevail. I have nothing to complain about — and as Barry says “These things recalibrate you”.

Unknown to many of you who helped, your words, gestures, drinks, small questions and chewing over problems as friends — you helped me so much.

It made me feel enmeshed, supported and somehow part of something bigger, less isolated in my worries and just back to normal again, for a fleeting moment at least.

I haven’t told many people about the story but for all of you, even if you just said hello and smiled, you rock my world.