Learning isn’t free; re-learning is paying the price twice. Many of the people who use what we geeks make would like to re-learn less.
One of them is Patric King, interviewed in The Setup, an instructive and enjoyable publication. I’m excerpting his whole last paragraph but the rest is good too:
I would love to see a return to a longer turnaround between software packages. There’s an artificial churn happening in how quickly we need to re-learn tools, because companies are learning to move their software products to a subscription basis. I am seriously tempted to jump off that bandwagon, if I were confident I could find a workflow and OS that wouldn’t be painful to re-learn on a bi-annual basis, rather than every six months.
At Apple historically it was like this: If Steve and Jony wanted to make a big change (say, reverse scrolling in Lion), they just told the engineers to do it, and the engineers did it. People complained, but because Steve and Jony had great instincts, usually ended up ahead of where they’d been.
At Google historically, someone proposes a change (say, rework the Gmail compose window), and there’d be agreed-on metrics as to what part of the user experience they were trying to improve, and they’d run big studies, and if the metrics moved the right way, the change rolled out. People complained, but because the data mostly doesn’t lie, usually ended up ahead of where they’d been.
OK, I’m stereotyping: Apple measures things, and Google has Matias Duarte. And crucially, both strategies give the same weight to the interim pain, when people are figuring out how to do again what they could do before: Zero.
This hurts me. I’m an old guy with lots of civilian friends, and when they ping me and say “OMG why did you BREAK my Gmail?!?!” I get pissed, not because they’re ruining my day but because we partly ruined theirs.
Am I a Luddite? Am I just change-averse? Do I want to stick with “good enough” when on the Internet, tomorrow is axiomatically better than today? Do I want us to ignore all the people building better things that people will switch to if we don’t build even better things?
Well, no, mostly. It’s called “software” because it’s soft, you can change it. But I hate to ruin anyone’s day. And I’d like to measure the pain not just the gain.
Comment feed for ongoing:
From: Paul Morriss (Apr 17 2013, at 01:42)
If the data doesn't lie, yet there is pain, then maybe the metrics need to measure that pain. I guess that's hard to do from the server end.
[link]
From: ken (Apr 17 2013, at 07:48)
Is the "take me back to the old version" button, the analgesic? Each click is the result of a stab of change-pain.
[link]
From: tom jones (Apr 17 2013, at 13:22)
i think it's painfully obvious that without change, there is no progress.
and in every version of every software there is that one favourite feature that someone doesn't want changed.
the solution sounds simple: "just let users use old version", but that doesn't scale. maybe they would like other features to change, just not this one. do you fork your software for every feature change? what about security patches? how many versions can you maintain in parallel?
the problem with software is that it's insanely hard. even maintaining code in a complex project that doesn't even need to add/change features takes great effort.
most people, even people in our industry who should know better, underestimate the effort required, on a daily basis.
i think it's just the nature of life, adapt or die..
[link]
From: Karl (Apr 17 2013, at 13:24)
Difficult topic.
A "take me back"-button might cause people switch back without really trying the new interface.
Active responses are tilted because people who are happy hardly comment things.
I personally do wish for a "please do not turn off Reader"-button. But this is a different story of a different kind of (great) pain.
[link]
From: Kimberly Blessing (Apr 19 2013, at 04:24)
<q>Am I a Luddite? Am I just change-averse?</q>
I'm asking myself the same questions -- not over software, but over hardware. So long as software maintains the same keyboard shortcuts, I can typically deal with most interface changes... but hardware changes (laptop keyboard layouts, touchpad positioning and removal of mouse buttons, and screen sizes) are causing me the greatest amount of pain and I currently have a brand new machine in pieces trying to figure out how I can trim plastic and splice cables to hack a better solution together. So I suppose I'm more change-adverse, because I'm certainly not opposed to taking on the technical work of customizing my kit so that it works better for me. Still, it would be nice to have the hardware options we once had...
[link]
From: len (Apr 29 2013, at 11:10)
1. What is worth not doing?
2. What is enough?
Answer those two questions every time you make any kind of deal.
GUI migrations for the sake of style points with precisely the same functionality fail question 1.
Not making a fair deal because there is still money on the table fails question 2.
I am still using XP at home and if the web keeps punishing Internet Explorer I will only use sites that don't. Enduring pain as a means of punishing a tormentor is a fair strategy when at war or with children.
If a change doesn't make a difference to me, I will generally consider it torment. :)
[link]