“Side effects of developing for yourself” is an interesting piece by Marco Arment, creator of Instapaper (a simple tool to save web pages for reading later).
In it, Marco talks about how features only get developed if he wants to use them. That means a big NO to the following: unread-count icon badge, tags, full-screen reading (where you tap to temporarily show the toolbars), comments, and Graphical Mode (“It’s one of those features that people say they want until they actually use it and realize that it’s not worthwhile at all.”)
Does this mean he’s not listening to customers? No, he’s just not letting them steer the product.
I try to minimize ways for my customers to shoot themselves in the foot…If I let users steer product decisions, the result would be a massive codebase producing a bloated, cluttered product full of features that hardly anyone used at the expense of everyday usability and polish on the features that matter. Like Microsoft Word. Or Firefox.
Great to hear about Marco’s strong point of view. And I can vouch personally for the results: Instapaper is the iPhone app I use the most.
On a related note, “Feature checklist dysfunction” is another post by Marco where he rails against checklist comparisons. Here he evaluates the iPhone to see whether it’s a good product:
“Sounds like a terrible product. I bet it will fail.”
Jean Philippe Cyron 18 Feb 10
Solving something for you is often solving something for others. Do it for you, and not for no one else… and you may build a business upon it.
I cannot agree more about how Marco is developing his application. Yes, he listen to his customer, but he’s doing the application for himself first and it’s giving him a horizon to follow, a ground to return to.
Instapaper is so useful, that I will consider buying an iPhone for that application alone. I use it even more than the actual phone!
Marco Armenton 18 Feb 10
Thanks for the post and links, Matt. 37signals’ design and product philosophies have strongly influenced me, and I’m flattered that you think highly of my product.
Anonymous Cowardon 18 Feb 10
You could just buy an iPod Touch and save the phone bill?
JFon 18 Feb 10
Thanks Marco. Instapaper has become indispensable for me. I read more because of it. Thanks for making it.
Christianon 18 Feb 10
While I read Marco’s post I thought how he’s kind of the mini-Steve Jobs of Instapaper. To a certain degree it kinda proves a single, strong leader can be helpful to make successful products.
I’m curious how Marco deals with the product getting bigger and bigger. I hope he can keep saying “no”.
Pieson 18 Feb 10
“Like Microsoft Word. Or Firefox.”
Both of which, as we all know, are failed niche products, whereas Instapaper rules the world.
Seriously though, don’t argue with results.
Jeff Putzon 18 Feb 10
At issue isn’t the results of sales/downloads, it’s whether or not the product is truly usable. Features do not equate to success in those terms.
Marco Armenton 18 Feb 10
Christian: Instapaper has been on the iPhone since a few days after the App Store launched in 2008. Since then, I hope I’ve demonstrated continued restraint despite the product’s (and service’s) growth. It’s difficult to add useful features without adding complexity, but it can be done. In many ways, Instapaper today is even simpler than its 2008 versions.
Pies: My goal in this was to discuss quality and how I prioritize features and feedback. Success, as measured by installed base or revenue, doesn’t strongly corrolate to quality. A lot of mediocre products are extremely successful, and a lot of extremely successful products are mediocre.
I can’t argue that Microsoft Word or Firefox haven’t been successful, and I won’t argue that they don’t deserve their success. They both try to be everything to everyone, and they’ve largely achieved that, hence their success.
But, like most independent or small developers, I have neither the resources nor the desire to be everything to everyone, and I don’t like the experience of using most products that were designed in that way. Being everything to everyone incurs huge costs in complexity, reliability, and efficiency that I can’t afford, that I can’t tolerate in products I use, and that can’t result in a product I can be proud of.
Merleon 18 Feb 10
“A lot of mediocre products are extremely successful, and a lot of extremely successful products are mediocre.”
I can just picture Fozzy Bear saying this to Kermit on the Muppet Show. And then Kermit does his double take.
Doug Reederon 18 Feb 10
I agree handheld applications should only add features when they don’t break a clean user interface, but limiting yourself to features useful to yourself can limit the appeal of the app.
ipotpalon 18 Feb 10
I completely agree with Doug Reeder. It is a good starting point to think of the features you wouls like to see in a product like this but you should be willing to hear what customers want
Mathew Pattersonon 19 Feb 10
Is there something wrong with “limiting the appeal of the app”? Trying to make apps that appeal very broadly is often a path to disastrous interfaces.
Jesseon 19 Feb 10
Marco, thanks for the post and your comments here. Interesting as always, and I basically agree with all your points. Only, I wish you’d have addressed the idea of comments with more than a flip aside. I’m glad to know you’ve at least considered the idea. But what exactly do you find so foot-shooting-ish about it? Some of the longest comment threads are the most worthwhile to read, which would seem to make them perfect Instapaper fodder. Can you be more specific than “trust me, you’d agree if you tried it”?
Marco Armenton 19 Feb 10
Comments can be beneficial, but usually aren’t. For the vast majority of comment-enabled blogs, the comments are a net loss for the author with very high rates of ad-hominem attacks, nastiness, nonsensical responses, and spam.
Therefore, since the most common case is a failure, it’s the sort of feature that people like me choose not to implement. Rather than handing the user the gun they asked for and saying “Be careful”, we just refuse to give it to them in the first place. It’s a difference of philosophies.
I recognize that having this discussion in a blog’s comments is a counterexample, but this is a very unusual setting. Most comment threads are nothing like this. (Even 37signals’ higher-traffic comment threads can get unwieldy.)
Lukeon 19 Feb 10
Perhaps the solution is to sell, in addition to a “standard” version and an elegant “pro” version, a “bloat” version, in which you give in to every last feature request, including comments, cupholders and a horn that plays “La Cucaracha.” What could go wrong?
G.Irishon 19 Feb 10
Clearly one should not blindly implement every feature users ask for.
Word definitely is more complex than it needs to be, but there’s a reason for that. In my opinion, Word has the massive mountain of features in order to justify charging people for upgrades. For the overwhelming majority of Word users they could get by on using the features from Word 97 but the problem for MS is that if they left Word alone in 1997 they wouldn’t have made much more money from it.
I don’t know how anyone can sit and call Firefox mediocre. It is a web browser, by design it has to support lots of different web technologies, otherwise it is not very useful. The core functionality is simple and easy to use and the application is stable and relatively fast.
It’s all well and good to preach the gospel of simplicity but there are good reasons for simpler apps and good reasons for more complex apps (sometimes if only to sell upgrades) .
Using one’s own tastes and desires as a way to develop the user experience and feature list is a fine way to go if you’ve got good taste. But it is also the road to ruin by hubris. At the very least usability needs to be tested and customer response analyzed to make sure that you’re not barking up the wrong tree.
Jesseon 20 Feb 10
Thanks for the response. I wouldn’t suggest saving comment threads by default; I’d just like some way of overriding the comment-clipping in those exceptional threads. Even some of the wackier threads can have a lot of value in them. I love reading Bob Cringely’s essays, but sometimes he can get carried away with the creativity that makes him so interesting to read. So I read his blog comments as a reality check, even though many of the commenters are assholes. I can skim, and the intelligent comments justify it.
What if you added an option to make the bookmarklet save any text that’s been selected in the browser window, rather than the whole page — a la the Google Reader “share this” bookmarklet and many others? Even this feature could be tucked away as a preference, disabled by default, but available for people like me who want some more flexibility over what they read. How does that sound?
Stephanon 21 Feb 10
or maybe micro-payments for the privilege to post a comment… it would make you think twice about your commitment to your comment
Steveon 22 Feb 10
I agree about the tyranny of the user/client. I have worked on and off in commercial software for many years, more than 10 as a vendor. Our biggest problem with a successfull version was all the additional functions that people wanted added in because they had seen them somewhere else (they never said that but we knew because we had looked as well). Trust your own judgement so there is nobody else to blame and you keep following your own curiosity in the product.
This discussion is closed.