If people open the journey but don’t reach the chat, look for the step where the most drop-offs occur. Focus not on the overall conversion but on the transitions between steps: where the drop is sharper than others, that’s where the problem lies.
Why analyze by steps instead of overall?
The overall figure of "completed the journey" indicates a problem exists but doesn’t specify where. A five-step journey loses people unevenly, typically in one or two places, and that’s where the drop usually occurs.
So, compare adjacent steps. If almost everyone moves from the first to the second step, but half drop off from the second to the third, the third question is the culprit, not the journey as a whole.
Which step usually drives people away?
It’s often one of four:
a question requiring typing instead of selecting an option
an unnecessary step that adds no value
an unexpected off-topic question
requesting a phone number or address too early
They all share one thing: the step demands effort or trust before the person is ready for it.
What to do with the problematic step?
First, try removing it entirely and see if more people complete the journey. If the step is necessary, replace text input with multiple-choice options or move it closer to the end, when the person is more engaged. Tweak one step at a time; otherwise, you won’t know what worked.
How to check if it got better?
Compare the percentage of people reaching the chat before and after the change with a similar number of users. If it increased, you found the right step. If it stayed the same, the issue wasn’t there; revert it and check the next largest drop-off.
Frequently asked questions
Where to look for drop-off points?
In the journey statistics: Mate shows how many people passed each step, so you can see the drop-off through transitions. The final step leads to WhatsApp Business, revealing how many people actually reached the chat. No separate analytics are needed; just look for where the numbers drop sharply and address that step first.
How many people are needed to draw conclusions?
A rough picture can be formed with a couple of dozen who completed the journey, but the fewer the data, the more cautious the conclusion. With five people, one random drop-off looks like a step failure. Wait until enough people have gone through the journey, then decide what’s driving them away; otherwise, you might fix something that isn’t broken.
What if people drop off at the very first screen?
Then the issue isn’t with the steps but with the cover: the title doesn’t promise value, or the screen speaks about you instead of the client. People leave before the first question when they don’t understand why they should start at all. Rewrite the first screen to focus on the user’s benefit, then look at the steps inside.
Can the journey simply be too long?
Yes, and this is evident from the drop-off pattern: if people leave gradually at each step rather than at one point, the issue is the overall length. In that case, reduce the number of questions to three to five, keeping only those that change your answer or filter out the wrong clients. A shorter journey has a higher completion rate.
Should I ask those who left why they didn’t finish?
Usually not, because they didn’t reach the chat, and there’s no connection with them. It’s much more reliable to go through the journey yourself as a customer and honestly note at which step you wanted to close the tab. Your feelings at the problematic step often align with others more than you think, and it’s free.
