As visible on your site as the penalty area on a soccer field, the conversion funnel is the critical zone to be reached. In this area, the slightest move can have a major impact, as it’s from here that most goals are made – but here, mistakes can also be very costly…
When it comes to conversion funnels, what are the most common mistakes?
The funnel’s complexity, or lack of precision, is often at the root of many difficulties that visitors may encounter.
Step by step conversions
The goal is for 100% of visitors who enter the conversion funnel to convert.
In order for this to happen, you must give them all useful information up front and in a clear manner, ensure that they won’t encounter even the slightest obstacle or doubt while in the funnel, and make the journey simple.
So, the conversion funnel should involve the least possible number of steps?
There is no absolute rule – what matters most is the user experience, conversions and, more broadly, improving your ROI, which is the ultimate goal.
In the penalty area, the player shouldn’t necessarily shoot to score a goal. Sometimes, it’s better to make a key pass to optimise the chances of success.
An additional step in your conversion funnel can be useful, if it helps bolster user trust or respond to a particular question, but it should be introduced with a certain continuity regarding your conversion process (example: delivery information), and in any case, should not create a distraction.
But beyond simple design issues, I’ve often noticed insufficiencies at the analysis level: Only studying the number and percentage of visitors who go from one step to the next is just a mere observation of performance.
To find real insights, we’ll have to look further.
How can I get actionable insights into my conversion funnel?
To improve your funnel, you need a clear view of the visitor experience in order to identify problem areas and resolve them.
This is why your AT Internet interface allows you to consult, for each defined page, the number of abandons, backs, errors and external pages, and all their associated details.
This level of detail is critical for making useful corrections to your content and improving your conversion rates… and your bottom line!
Even if you generally monitor the number of abandons, it doesn’t always suffice.
In this example, there are no abandons, so all seems well.
Everyone is happy…
… Except more than 30% of visitors do not follow the route we’ve outlined for them (=Next).
This doesn’t go unmissed by the analyst, who must therefore find solutions to a problem that s/he sometimes (or often?) is the only one to see.
Clicks on “Backs”, or views of pages external to the funnel, clearly demonstrate the shortcomings of your conversion funnel. Why would visitors, who are already engaged, feel the need to go off-track? What could be distracting or interrupting them?
This is where segments or cohorts are crucial for measuring any resulting gaps.
For the digital analyst, it’s not enough to just work on the clear-cut, obvious problems.
The analyst pushes furniture around and looks under the rug to find and eliminate all dust lurking unseen:
The digital analyst’s added value also resides in his or her capacity to find areas of improvement by unearthing these “hidden defects”…
… provided that his or her suite of analytics tools is capable of giving the information required for the analysis, and allowing him or her to create custom KPIs.
Beyond quantitative data, it is crucial to also obtain qualitative information in order to act with great precision on the conversion funnel’s cold spots.
Because even if traffic volumes are helpful for identifying cold spots and determining priorities, it’s still not enough to enable precise action.
What qualitative data must I deal with?
True insights reside in the details of abandon pages, return pages, and external pages, because they clearly indicate the pain points in the user experience – which are therefore points to be improved.
Let’s compare behaviours at the funnel entrance, and at Step 1:
Entrance (cart) Step 1
We can see the significant share of “delete product from cart” and “product information” pages in the first case, and the “login” page in the second case. The abandon rate is 57% and 10%, respectively.
Does this data make sense in the context of the site in question, or does it require corrective actions?
Viewing conversions step by step answers the questions “Where?” and “When?”, whereas the above detailed view answers the question “How?”, which will help us answer the question “Why?”.
To go even further, a very helpful double-view of the funnel is available: “Conversion paths” (visits) and “full navigation” (clicks)
How can a view of clicks or visits help me?
A uniquely fine-grained analysis
The AT Internet solution is the only tool offering this double approach; our interface enables the analyst to discover pivotal insights. Here’s an example:
By comparing both, you can identify atypical behaviours:
In this example, let’s consider the figures to the left (conversion paths).
We can see 6 visits toward step 1, then 4 visits toward the following step.
But in examining the figures to the right (full navigation), we see 40 clicks toward step 1, then 4 clicks toward the following step.
40 clicks over the course of 6 visits = There’s a problem!
What could such a large discrepancy be telling us? Simply that during these 6 visits, numerous back-and-forth clicks were made, which would seem to indicate that visitors were quite hesitant or unsure, and were looking for answers to their questions.
In short, these visitors – who were theoretically very motivated to go further – encountered a roadblock.
This fact could only be brought to light by comparing the approach in terms of clicks, and in terms of visits.
Only by going into this level of detail can the analyst provide business users with truly actionable insights allowing for an improved conversion funnel.
If the customer journey finishes in the conversion funnel, it means we did things right to convince the user to enter the funnel. The upstream part of that journey is the process of persuasion, which deserves the digital analyst’s full attention. This will be the subject of my next article.