Dissatisfaction amplification—or “How users bear a grudge.”

Some design improvements to a client’s order journey went live recently and they show some interesting statistics.*

The first is that there was a 12% like-for-like increase in conversion, which totals about £7 million of extra sales per year. I’m obviously proud to have worked on a project that has results like that and it’s great to have solid numbers to show the value that good design can have to business. (It feels a bit like I now have my own mini version of the $300 million button story.)

The second and more interesting result was that the level of drop out continued to fall on the pages in the order journey after the section were the changes had been made.

If the order journey had six steps and we made improvements to Step 3 that increased conversion for that section by 8% we might have expected the change in conversion for each section to be:

Step Change in conversion
1 0%
2 0%
3 +8%
4 0%
5 0%

with a total increase of 8%. But instead we saw results like this:

Step Change in conversion
1 0%
2 0%
3 +8%
4 +11%
5 +13%

So by improving the experience in one section of the order journey conversion increased  in the subsequent sections even though no changes were made to those sections. This is interesting.

Dissatisfaction amplification

I can’t remember where I read it, but several years ago I came across a metaphor that suggested imagining each user comes to a site with a bucket of satisfaction. Each usability problem** or bad experience would mean a bit of satisfaction was spilled from the bucket. When the bucket is empty the customer leaves the site in frustration. This has always seemed very apt for order journeys—like the ones in this project—where a customer has to go through a set, linear, journey.

The bucket of satisfaction metaphor would suggest that the scenario that we had initially been expecting (with no change in conversion in the sections after our changes). But our statistics suggest that something more is happening. There is a kind of dissatisfaction multiplication effect—usability problems early in a user journey are amplifying the dissatisfaction caused by usability problems later in the journey. So by removing some significant usability problems in Step 3 we have reduced the dissatisfaction caused by the usability problems in Steps 4 and 5.

Now that we can see it’s pretty obvious that once a user has been irritated they will be more likely to be increasingly irritated by even a small problem on the next page. Put simply, users bear grudges, if you piss them off it’ll affect how they view the rest of their experience with your site.

So what does this mean for designing user journeys?

Well for one thing, fix your usability problems. Especially the ones in at the start of the journey. But I’m guessing you I don’t need to tell you that.

What’s more interesting is what to do when you have an experience problem that you can’t design away. Let’s say you have a complicated form that user’s need to fill in or you need to try to get your customer’s to give you consent to use their details for marketing.*** Dissatisfaction amplification suggests that you’ll want to design known problems to be as late  as possible in the user journey so they don’t multiply the effect of smaler usability problems in the later pages. 

* The numbers have been changed to protect the innocent but they are all in the correct ballpark
** I’m being liberal with the idea of a usability problem here. Anything short of telepathy is counted as a problem, all pages a user needs to go to or tasks they must complete should be seen, in this context, as a problem.

***Yes, I know it’d be best not to ask for this, but marketing consent can be worth a lot of money to businesses.


Comments (1)

  1. David Hamill

    It really is dependent on what you changed. I know you can’t share but without talking about it these are really just numbers. Good ones though.

    However I agree that if you’ve got something you know will annoy/frustrate people then you’re better to leave it until the end because they are more likely to feel they have invested too much time to back out at the end.