Last week we kicked off our 2017 events programme at our
London Finance Forum, with guest speakers from Google and Mintel. We chatted data – market data, how to collect, segment and better interpret the data you already have, and finally how to use data to design smart testing programmes.
The day’s discussion was centered around a few key challenges:
- The continued focus on mobile and the need to be “mobile ready”, or even “mobile first”
- A concern that people aren’t making the most of the data they already have, even with free tools like Google Analytics
- And lastly, testing. How to design tests and how to know when they’ve worked, or when they haven’t!
After the event wrapped up I had time to think about some of the additional tips the speakers shared throughout the day. Here are my key takeaways for those of you who may not have been able to attend.
The year of mobile is here
Our first speaker was James Nesbitt, Senior Agency Development Manager from Google, who proved that the year of mobile is now here. In Q2 2016 Google saw Mobile search queries overtake Desktop queries for the first time.
And we can see this in practice in the Financial Service sector alone. Earlier this year,
Mintel reported that 45% of UK internet users have used Mobile Banking services in the last 6 months.
So, we’re in agreement that mobile is important, but how can you apply this to your activity?
1. Ensure you are segmenting and tracking your
keyword ranking performance by device.
- Jonathan, SEO Group Head at equimedia, ran a quick straw poll, asking people to raise their hands if they were tracking organic rankings. Just over 50% of the room did so.
- When the second straw poll asked whether they were also splitting their rank checking by device, the percentage of hands raised dropped to around 15%.
- This was just one example of segmentation – by device – that Jonathan talked about in the session:
A KPI is worthless unless it is segmented
2. Don’t forget about Page Speed. Use tools to check the speed of your site by device.
- Google & Kissmetrics data presented at the event stated that 40% of users in the UK will abandon a site that takes more than 3 second to load.
- James added that the average load time of sites in the UK is 13-14 seconds, so there’s some work to do!
- Here I’ve got 2 tools for the price of 1 for you:
What we often see is that while brand sites are “mobile friendly”, the page speed of the site is then poor on mobile devices (and sometimes desktops as well). This example from Ageas shows just that:
Mobile-Friendly Test tool (Source)
Google’s PageSpeed Insights tool (Source)
This tool also provides tips on how to improve the speed of your site, so it is worth spending time looking at and sharing this with the team responsible for the website.
We could all be using Google Analytics a bit more
It was no surprise to see almost all hands in the room fly into the air when asked about Google Analytics usage. However, through the questions asked and conversations in the coffee break, it was clear that attendees felt they could be using Google Analytics more.
Following on from the theme of segmentation, we shared our top, underutilised Google Analytics reports with the room. Here are just 3 for you to go away and read more about:
Custom Dimensions to segment users by key behaviours or sections of the site.
Using HSBC as an example, we gave 2 ideas for Custom Dimensions: tracking logged in vs. non-logged in users and tracking Personal vs. Business banking customers.
Event Tracking to discover user intention.
Events track user interactions that don’t cause a state change on a URL. Event tracking can be easily configured using Google Tag Manager and, when combined with advanced segments, can reveal key user journeys and interactions.
Using TSB as an example, Event Tracking can be used to determine which button, in which location, on the homepage is the most clicked (fighting for homepage real estate is a common challenge for Financials Services brands, particularly those with many products for both B2B and B2C customers).
3. Find bottlenecks in the user journey with the
Behaviour Flow report.
The Behaviour Flow report visualizes the path users travelled from one page or Event to the next. If you also have Events and Content Groupings set up you can get a clear view of user engagement.
Jonathan’s video showed the report in action, using a theoretical example:
This is a favourite of ours because it is great at identifying ‘blocks’ in the user journey which can then give way to test hypothesises. One example discussed was common testing around quote retrieval buttons and the position, colour and design of these.
Do you really know when a test has worked?
This leads us quite nicely on to our final topic around testing. The room discussed instances where the number of tests ran might be set as an objective for the marketing team and why this is a poor metric for measuring success. What we are all aiming for is
successful tests – and that’s where the question came from; how do we know a test has been successful? Short answer? Good planning.
Our advice is to:
1. Agree the KPIs for the test before starting
- For example, Click Through Rate and Click to Conversion for PPC copy testing
2. Agree a target uplift you’d like to see from the test
- Work back from your target Cost per Conversion. What do you need to see to get you closer to your target? 10% uplift?
3. Plan how long you need to run the test to get statistically significant results
- You can use tools to help with this including this one example from VWO you can find here
This approach should mean you have a target conversion rate, over a set period, to work towards. Tracking success against this should then be simpler to measure.
Hopefully in here you’ll have found 1 or 2 tips to take away and use yourself. A big thank you to those of you that may have attended! Lots was learned and new connections were made – I’m now even more excited for our next event later this year! To register your interest in our next event, email firstname.lastname@example.org.