denise: Image: Me, facing away from camera, on top of the Castel Sant'Angelo in Rome (Default)
Denise ([staff profile] denise) wrote in [site community profile] dw_biz2012-02-13 03:45 pm

feature design brainstorming: icon add-ons

We said last month that [staff profile] mark's next big project is going to be icon add-on packs to let people buy more icon slots if they want, and this week he and I have been brainstorming ideas to make it work in the quickest, easiest, and most usable fashion. This is what we're currently thinking the system will look like, for you guys to mull over and point out all the things we've no doubt forgotten to think about. ;)

Goal: To let people buy as many icon slots as they want (up to whatever limit we impose for overall performance reasons), as simply and easily as possible.

We started out with the idea that this should be something the user can decide (how many slots you want), not sold in pre-specified numbers of icon packs that stack on top of each other. We kicked around a few ideas for a while of how to make that work (such as paying per icon slot per month, etc), but everything we tried to come up with got really confusing very quickly: we would have had to track a lot of different things, and explaining pro-rating to people is really hard, and it would've been really hard to add more icons later if you decided you didn't have enough.

So, our current working theory: we will charge you up front for each icon slot you want to add, and paying for another icon slot will give you that slot permanently, for whenever you have a paid account. (We have a vague idea of what each slot will cost, but it's not set in stone yet, so I don't want to commit to anything; I'll use the variable $amount while I'm explaining, in order to avoid making any promises.) If your paid account expires, you'll go back to the number of icons a free account gets; if you renew your paid account, you'll go back up to the paid account icons + your add-ons.

It's probably easiest to talk through some practical examples of common scenarios, so everyone's on the same page: let's say that I have a premium paid account, so I have 250 icons. I want 270 icons. I pay $amount to permanently buy those 20 icons; my icon slots go from 250 to 270. In a year, my premium paid time is about to expire, so I renew it for another year; I only have to pay the $50 to renew the premium paid time, and my icon slots stays at 270, not the 250 a premium paid account usually gets, because I bought those 20 icons permanently.

Next year, my premium paid account expires (back down to 15 icons, curses!), and I'm kind of low on cash, so I decide to renew it as a regular paid account ($35 for the year; 100 icons), not a premium paid account. But! I previously bought those 20 extra icon slots. Those still exist, but they're applied to the paid account icon limit (100 icons), not the premium paid account icon limit (250 icons): I'd have the 100 icon slots a paid account usually gets, plus the 20 I permanently bought, for 120 icons.

After a few months, though, I decide I can't live with only 120 icons, and I decide to buy some more. I pay $amount to permanently buy another 50 icon slots. My new icon count is now 170: the 100 for a paid account, the 20 I previously bought, and the 50 I just bought.

When account renewal comes around, I decide I miss the premium paid account benefits, so I renew as premium paid ($50 for the year; 250 icons). I now have the 250 icon slots that come standard with a premium paid account, plus the 20 I bought a long time ago, plus the 50 I bought recently, for a total of 320 icons.

So, you're only buying the icon slots once, and they last forever -- but, you have to have a paid account to access them. (This is to avoid people buying just icon slots, which is bad for us from a business standpoint based on how we set our account limits. For an explanation of why you won't be able to just buy icons without a paid account, see two old mailing list messages I wrote back when we were in development: #1, which explains why you can't buy paid features a-la-carte, and #2, which specifically gets into icons.)

We'll be pricing icon slots based on the cost to support them over time, so you'd pay more up-front than you would in a yearly, expiring type deal. You'll never have to pay again, though, so it will be cheaper in the long run.

What if you want to switch to using a different account, though, the way we know roleplayers like to do? You'd be stuck paying the up-front cost over and over again for each account, which would not be very fair! So, instead we make it possible for you to switch icon slots from account to account.

Let's say I have two accounts, [profile] x and [profile] y. [profile] x is a premium paid account (250 icons) and I bought 50 extra icon slots for it over time (total of 300 icons). [profile] y is a paid account (100 icons). I decide I want to stop using [profile] x and switch to using [profile] y instead: I can go to the icon slot mover tool and say "switch my extra icon slots", and move the 50 extra slots from [profile] x to [profile] y. Now [profile] x has 250 icon slots (the standard with the premium paid account), and [profile] y has 150 icon slots (the standard 100 with the paid account + my 50 extra slots that I bought).

(We may charge a small amount to move icon slots from one account to another, especially if it's been a while since you bought them, like the way we charge for a rename token. But we haven't decided that yet; it will depend on what the numbers look like when we diagram the costs of all this out more fully.)

There will be a limit on how many slots you can buy at first -- this is because the system isn't very optimized for large numbers of icons, either for resource usage or for the user interface of displaying and selecting large numbers of icons. (We can fix that over time, and we will! But that will take time, and we'd rather release the feature with a lower limit now than wait. Whatever limit we pick when we release it will almost certainly be raised later once we can do the work.) It's also possible that we might have two limits, and charge $amount for each slot up to limit #1, and $amount*2 for each slot from limit #1 to limit #2, but that, too, will be up in the air until we can really plot out the technical and business details of this way of doing things.

So, if this is all still up in the air, why am I posting about it now? Simple: We know that we can't know everything about how people use their accounts and how people want to use their icons. So, consider this the open invitation to pick holes in this plan: what kind of usage are we forgetting to think about/account for? What problems do you see?

(Also, because I know a lot of people are really sweet about worrying what this will mean for us-as-a-business: we already did all the back-of-the-envelope feasability tests, and this should remain feasable over time. We're gambling that the cost of disk space, bandwidth, and processor power will continue to go down over time the way it's been going, historically, so the pay-once model for icons should work fine for us -- and because it will be tied to paid accounts we won't be promising future services without any additional income the way we would for seed accounts.)
stormerider: (Default)

[personal profile] stormerider 2012-02-14 06:31 am (UTC)(link)
So, we'll still be getting revenue from that account, and in fact might get more revenue from the account (people continuing to pay for their paid account because they don't want to lose access their icon slots, since they're more invested in having those icon slots -- enough to have paid for them -- where they might otherwise let the account lapse).

This is the only reason I renewed with LJ the last few years. Which reminds me, I need to trim down my icon list there before my time runs out again.
zellieh: kitten looking shocked, openmouthed, text: WTF? (What the fuck?) (Default)

[personal profile] zellieh 2012-02-15 03:17 am (UTC)(link)
Hah, me too! ::Opens a new tab, starts swapping icons around:: I have several months to go, but getting down to 15 is a painful process, so I've been trying to do it a bit at a time.
gumbie_cat: sketch of the original iron man armour (built this thing in a cave!)

[personal profile] gumbie_cat 2012-02-14 07:30 am (UTC)(link)
Honestly the biggest cost associated with large numbers of icons isn't the disk space and transfer associated with it, it's the computational cost of loading all the icon keywords/displaying all the icons in the icon selector drop-down and the UI cost of figuring out how to display and manage them all.

Triggered by you mentioning people liking to use their icon slots as a gallery (because gallery => museum => curation => YAY! in my head) and the fact that I'm personally never likely to max out my icon slots because I find large numbers of them unmanageable to sort through every time I comment and mostly only ever use the same 5 or so anyway...

Would it be possible to set things up so that icons could be sorted into categories (so for example I'd probably have fandom, knitting, books, and the all important misc) then have the icon browser only load your ~10 most used icons by default until you selected one of those categories?

I have no idea how complicated/easy something like that would be to do, or if it would even be worth it for the difference it'd make to the costs you mentioned.
kyrielle: A photo of kyrielle, in profile, turned slightly toward the viewer (Default)

[personal profile] kyrielle 2012-02-14 04:18 pm (UTC)(link)
...besides the resource savings for the site, being able to categorize my icons and look by category would be so shiny to me as a user. SO SHINY.

Also, an upload page that didn't waste time displaying my existing icons might also be handy. If I want to upload one or more icons, I usually don't feel the need to even glance at the others. I've picked the image out and just want to get it in place.
kyrielle: A photo of kyrielle, in profile, turned slightly toward the viewer (Default)

[personal profile] kyrielle 2012-02-14 05:59 pm (UTC)(link)
Or just make it one page where you can load in the stuff to be edited if you're going to edit it, or two variants on the same page (with a parameter), or....

Or not, if it's not worth the hassle for what it saves. But I seldom edit existing icons; if I go there, it's usually to upload a new one, and loading in the old ones is unnecessary. It may not be worth it to Dreamwidth to split it, in which case as far as I know it's harmless for me (maybe not for people on limited connections, but that's not me), tho.
turlough: green origami Stegosaurus ((other) not a cuddly toy)

[personal profile] turlough 2012-02-14 06:32 pm (UTC)(link)
We'd have to split the two

Oh no! As someone who changes icons a lot but always keeps the same keywords I'm dependent on being able to see my existing icons when I upload new ones. Otherwise I won't know which one I want to delete so I can transfer its keywords to the new one I'm uploading.
kaberett: Grinning emoticon. (:D)

[personal profile] kaberett 2012-02-14 11:20 pm (UTC)(link)
turlough: green origami Stegosaurus ((other) not a cuddly toy)

[personal profile] turlough 2012-02-15 03:18 pm (UTC)(link)
Isn't it the best thing ever?!! I found it over on LJ a while ago. Many different kinds of origami dinosaurs in fact.
kaberett: Grinning emoticon. (:D)

[personal profile] kaberett 2012-02-16 12:16 am (UTC)(link)

citymusings: (Chloe - Yay!)

[personal profile] citymusings 2012-03-25 06:20 am (UTC)(link)
♥ This whole plan and discussion page is making me so happy. I'm currently using about 30 of 250 icons because I like being able to see them all easily. If I could see most common + FandomX on one post and most common + FandomM on another, that would be amazing.
sophie: A cartoon-like representation of a girl standing on a hill, with brown hair, blue eyes, a flowery top, and blue skirt. ☀ (Default)

[personal profile] sophie 2012-02-14 06:46 pm (UTC)(link)
Some people will almost certainly pay once and then continue to use that icon slot for the next ten years, but that's offset by the people who pay once and then use that icon slot for maybe a year before they get tired of using DW and wander off (and thus don't use the icon slot over the X years we counted on for amoritzation).
Except that that doesn't take into account the fact that if they've wandered off the site, then they're not going to be removing any inactive icons. And I just tested and it seems that inactive icons still display on comments - which means that the bandwidth for displaying an extra icon is still going to be used. The fact that they can't use any others or upload any more isn't really the problem here if they've wandered off anyway.

LiveJournal gets around this by removing inactive icons from the server after 6 months. I couldn't see this in DW's FAQs, but since we forked from LJ, it seems reasonable to assume we do the same thing, in which case this issue isn't as bad as it seems. We should mention it in the FAQs, though, if this is indeed what we do. (And if we don't, it might be worth saying that as it would be different from what LJ does.)
Edited 2012-02-14 18:48 (UTC)
vatine: Generated with some CL code and a hand-designed blackletter font (Default)

[personal profile] vatine 2012-02-15 06:56 am (UTC)(link)
Mostly, because disk space, bandwidth, and processor power gets cheaper all the time (Moore's Law applies pretty universally)

I wouldn't take that as the gospel it used to be. Things will most probably continue getting cheaper, but slower. I'm fairly sure "computation per dollar" has had a slow-down in the increase over the last 3-4 years.