#3 - Going further in analytics
This is the third and final article in the Analytics trilogy.
- The first presented the definitions, objectives, principles and steps to successfully implement your analytics strategy: #1 - How to get started with your analytics?
- The second focused on the main concepts: indicators, audience, segments, acquisitions, etc. #2 - The main concepts of analytics
- This third article details other functionalities for a more advanced use of analytics, coupled with more qualitative methods.
Setting goals is an excellent approach to measure the impact of the marketing actions that are taken. In the field of eCommerce, this concept takes on its full meaning and is quite obvious. The ultimate goal is indeed the sale of a product. But this tool should not be overlooked in non-commercial contexts. There are multiple objectives that are very relevant in the case of sites that do not offer an online store. For example:
- Order via an online form from a white paper
- Getting in contact via a form or a phone number
- Reading at least 3 resources in the knowledge base made available
- Membership request via an online form
- Request for information via an online form
Defining very concrete objectives is also a good way to think about your site, its content and the structure of each page so that the whole is more effective. Indeed, if the goal is to get new members and the current form is a tiny link at the bottom of a very long page, we can imagine that the goal will rarely be achieved.
Google Analytics only records clicks on links leading to another page of the site, to external sites or to download files. No other clicks are recorded. Custom events can be particularly useful in understanding how users interact with the site, for example, to count the submissions of a form in a page (without loading another page). Custom events involve the addition of a snippet in the tracking code, but with the arrival of tag managers, the addition of custom events can be managed by non-developers, from the analytics interface.
Conversion tunnels are the different possible paths to reach a call to action, or a goal. For example, the form to become a member can be accessed directly on the homepage, from the page presenting the organisation to which you are invited to join or from the contact page. Which of these locations generates the most membership? Defining these different funnels makes it easy to compare data directly in the analytics tool.
In your marketing strategy, especially in its digital dimension, you will probably have planned campaigns on specific topics and/or in connection with specific events. The reflex for these campaigns is to link them to one (or more) objective(s) and to ensure as much as possible that the traffic generated will be identified by the analytics tool. In purely digital campaigns, it is obviously easier with tracking codes associated with links than in the case of marketing campaigns on non-digital media.
It can be very interesting to carry out an experiment in A/B testing to, for example, compare the effectiveness of a call to action in a page according to variations on the wording and/or on its layout and in shape. At least two variants will therefore be offered to visitors, in equal percentage, and analytics will reveal which one generates the most clicks. Be careful, however, to ensure that the volume of traffic is sufficient to be significant!
To get on board and help your employees benefit from analytics, it will most likely be necessary to give them access to personalised reports that will give them regular access to the key figures they need to monitor and know how to analyse. These reports are typically sent by email each month but are also accessible online at any time to access recent data.
To go even further ... a complementary qualitative approach often proves essential to help the interpretation of quantitative data and to invalidate or confirm hypotheses.
This method involves an expert to make an analysis of the existing situation and formulate proposals for modification in the pages of the site. This approach allows all known good practices to be applied throughout the site, but this is not always enough. To go further, the UX expert may need more qualitative data on the actual behaviour of users, such as:
Tools allow you to analyse the areas of a page that are most viewed and/or clicked. These data can be of interest to optimise the content structure of a page, to identify its weaknesses.
These are protocol tests offered to a representative sample of site traffic or at least of the priority target.