User resistance to product redesign

How to design for “change aversion”

5 min readFeb 24, 2020

Image of the word ‘change’ as neon signage

A couple of years ago I designed a methodology to migrate millions of customers over to a new loyalty program, using specific experience principles to mitigate something called Change Aversion.

In the process of designing this methodology, I discovered an awful lot about Change Aversion and how to design user and customer experiences and journeys that cause the minimal disruption to everyone’s experience, and the minimal fallout on social media.

What is change aversion?

First up, a definition for context. Change aversion is similar to loss aversion (Kahneman & Tversky 1979) — we value something more if we already have it, and we don’t like to experience loss. From this comes the assumption that we also don’t like something we have or use to change significantly. In short, humans don’t like change.

Hang on a minute…

Well it turns out that’s not entirely true. Humans do like change. We’re programmed to seek novelty — this is why our ancestors found food and why we’re all addicted to instagram/facebook/anything with a “pull for more” functionality.

People do like change… if.

People like change when they feel that they are still in control of that change. They like changes they choose for themselves — not changes thrust upon them. Consider for a second the experience of “pull to refresh” versus a constantly updating stream of content.

Image of instagram feed, highlighting “new posts” CTA to pull in new content

If your tool is remotely complex, or stores a user’s personal data for retrieval, or they have in any way learned how to use and access it then they have achieved a sense of Mastery (Morgan, 1990).

If you therefore change the experience (interface, tool, process or interaction) significantly with no warning, you may make them feel stupid. People do not like feeling stupid. It often makes them angry. Do you want to make your users angry?

Change Aversion is relatively short-lived. In fact pretty much everywhere you look you see the same phrase — it is a “negative short-term reaction to changes in a product or service”. [wiktionary]

Basically users experience shock or confusion, feel stupid, get angry, abandon a service, shout a bit on social media, and ultimately (hopefully) come back to the product or service. But you’ve done damage both in terms of that user’s experience of your product, service or brand and in terms of whatever they shouted at you about on social media.

Change aversion in action

Example 1 — Loyalty programs

In the case of loyalty programs, you have at least a portion of customers who have a high sense of mastery. They have learned how to use your program, have accrued points, maybe have “gamed” the system and have a sense of control, self-efficacy and as in the case of say airmiles, perceived higher status.

Changing the program changes their sense of status. This triggers loss aversion — either in the explicit loss of points or the implicit loss of understanding. This triggers anxiety, loss of control and ultimately modifies their perceived self-status. You probably don’t want to do this to your most valuable customers.

Example 2 — Social networks

In 2007 I was working on social networks at a large media company when we made a major update to forum design, going from the old style forum design with cascading arrows (a design pattern so old I couldn’t find examples for this post) to a more contemporary design, such as you might find somewhere today. We did this with zero warning. The users kicked off, abandoned the site, screamed on other forums (twitter didn’t exist) and within about 4 weeks, came back. But there was an awful lot of ill feeling which remained in the forum for months if not years.

Image of a small sign with the words “please stay on the path”

Mitigating change aversion

Here are some principles to consider when designing methodologies:

Understand how current users use your product/service — what do they value most, what do they use most, what do they not value or use (of course you know this, because you wouldn’t dream of changing an experience without research…)
Engage internal stakeholders from the beginning — whether their jobs integrate with tool usage or whether they are on the social media or contact centre teams who will be dealing with any fallout.
Identify any high-value customers and bring them in as close as possible. Set up steering committees, conduct co-creation sessions, make them feel heard and engaged.
Test any new propositions, features or significant changes with many different user or customer groups. And with internal stakeholders if appropriate
Alpha test with friendly users
Allow wider groups to opt in to beta test — but make it clear it is beta and allow them to turn it on/off and go back to primary live versions.
Start comms to wider audiences (including publicly) once your final migration time is coming
Have a dedicated channel or channels to take feedback so that users don’t have to scream into the void of social media
Ensure that those channels are equipped to respond to feedback quickly
Communicate any ongoing changes or improvements which relate to user feedback.

Basically, all of the above is a hygiene layer to include in any user-centred process to prevent unnecessary problems. So long as you have a client or employer who cares about their end users, what could possibly go wrong…