Skip to content

Instantly share code, notes, and snippets.

Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save jpmens/49577d0a0c6eb93392b82686e2672dfa to your computer and use it in GitHub Desktop.
Save jpmens/49577d0a0c6eb93392b82686e2672dfa to your computer and use it in GitHub Desktop.
'Users hate change'

'Users hate change'

This week NN Group released a video by Jakob Nielson in which he attempts to help designers deal with the problem of customers being resistant to their new site/product redesign. The argument goes thusly:

  1. Humans naturally resist change
  2. Your change is for the better
  3. Customers should just get used to it and stop complaining

There's slightly more to it than that, he caveats his argument with requiring you to have of course followed their best practices on product design, and allows for a period of customers being able to elect to continue to use the old site, although he says this is obviously only a temporary solution as you don't want to support both.

This argument is both incredibly entitled and terribly egocentric, as well as being wrong-headed on several counts.

Firstly: humans don't resist change when it's something that they asked for, they resist things being imposed upon them against their will. There is an incredibly persistent cultural movement in product design that "we know best", this is a very parent-child style relationship: "Mother knows best", that both disempowers and disengages customers.

Let me be clear: when I buy a product I am paying for what the product can do for me now. It fulfils a need that I currently have. I am not paying money out of my own pocket for a faint hope that the product may do something in the vague and nebulous future.

So: Product does X. I find that valuable. I pay $n to buy X capability. The product probably does Y and Z too, but I don't care about that. I bought it to do X.

When you as a product manager or designer or PO or whatever decide that your product should do A, B and C too, I don't care. I don't want those features, I didn't pay for them.

When you as a product person change the way that I have to use the product in order to do X, you are asking me to spend time, effort and attention to change my habits around X in order to do something differently, which may (or may not) benefit me in the future. In all likelyhood you made it easier for new users to learn X. I don't care about new users. I care about continuing to use the product in the same way as I always do in order to do X, even if you have forced me to do it in a sub-optimal way.

Every change that you make to the product after I have bought it makes it more likely that I will leave your product and find something else that does X instead, because the cost to me to learn how to something different in your product is now not much different than the cost to learn how to do something in a different product.

The more times you force me to change my behaviour, the more badwill (being the opposite of goodwill) builds up. Eventually I'll become so pissed off that I'll move, no matter what the cost.

Secondly: Your change probably isn't for the better. Not for me, not for the majority of existing customers. As stated above, the real benefit is almost always for new customers, who will find it easier to learn to use X. That's even assuming that this isn't a 'branding' change, which actually benefits no-one other than the expensive branding consultants that you just paid.

The vast majority of the effort that designers spend on look and feel, typeography, colour palettes, image choice and placement, tone of voice, button placement, size and style and a host of other things are of marginal value at best. The really hard stuff - like ethics, accessibility and knowledge architecture are almost always neglected in favour of bike-shedding. The popular rise of apps like Pocket and browser features like Firefox's Reader View are proof that it is the functionality and the content that is important, not what colour the buttons are.

Thirdly: the idea that you can just tell your customers to suck it up is a relic of last-century marketing that relied on captive customer bases and lack of customer knowledge, awareness and community. Modern customers are, in the majority, well informed and highly vocal with other customers in their community. Unless you have a significant barrier to exit you'll find that your established customer base leave the moment your competitors make it easy enough for them to migrate. Even the most impressively built and reinforced barriers don't last forever. OpenOffice and Google Docs, coupled with a change in the way that offices work have meant that even giants like Microsoft are losing their hearlands of enterprise business software contracts.

We can no longer afford to be complacement with our customers.

The idea that it is impossible to support more than one version of a product presupposes that a) work is required to upgrade both versions simultaneously, and b) that the esxisting product isn't stable i.e. still many bugs being surfaced. We have many known solutions for the second malady (q.v. software crafting) but the first problem overlooks a simple strategy: Extensible Product Portfolios (EPP).

The idea of EPP is thus: when you have a product that works, and an existing customer base - freeze it. Instead of a major redisgn because 'Material Design is so 2014' simply leave the product the way it is, bar minor BAU and bug-fix work. Instead devote effort into building a new, next-generation product that addresses (hopefully) a new customer segment, and allow existing customers to add this new product to their portfolio for a incremental fee. This allows existing customers to self-select into a new product, protects revenue and reduces the risk of existing product customers leaving due to badwill.

In this way a team/organisation builds up a protfolio of products, all of them profitable, all of them long-lived. After the vast majority of customers leave an old product for '2.0' then when only a small minority remain you can sunset the old product, perhaps offering customers a free upgrade path, or just leave it running indefinitely as it's marginal cost of maintenance is now essentially zero.

This treats your customers like adults, gives them the freedom of choice and empowers them to use that choice in order to best satisfy their own needs.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment