By John Gruber
Work at Atoms. Make the best shoes ever.
I suspect that many Mac users — as well as some aspiring developers — have only a half-conceived notion of the economics involved with becoming a successful indie Mac developer.
The basic idea is that a developer comes up with an idea for an app, implements it, ships it, and starts selling software licenses for, typically, around $20-40 a pop. If he can sell 1,000 licenses in a year, that’s a nice hobby. Sell 2,000, and he’s getting close; at around 3,000 licenses per year, revenue is probably in the ballpark range of a full-time salary. (Keep in mind, however, that, say, $80,000 in software license revenue results in much less personal income than a job with an $80,000 salary; salaried jobs tend to come with things like health insurance.)
At that point, I think, many people assume the developer has achieved The Life. Economically freed from the need for a day job, he gets to hack all day on the software instead of being forced to squeeze development time out of nights and weekends. I.e. that once a developer gets to the point where revenue from license sales is high enough to be considered a good full-time salary, he’s done, and now he just has to keep it going.
But then what? What does it mean to keep it going?
The full story is that like most forms of popularity, software sales tend to follow a power-law distribution. Most apps languish in obscurity, but popular apps tend to become super-popular. The hardest part about selling software is just getting noticed; if your app is only known by the sort of people who check VersionTracker and MacUpdate a few times a day, or even just the sort of people who know what VersionTracker and MacUpdate are, it’s unlikely that you’ll sell enough software to earn a living.
But once you get over that initial hump, subsequent sales become much easier. You start getting media coverage, and most importantly, assuming most of your users are happy, word-of-mouth advertising really starts to bloom.
But selling software isn’t like selling books. When a book takes off and climbs the best-seller charts, that’s just money in the author’s pocket. Each software sale, on the other hand, comes with incremental support costs.
Users need help. They need help when they encounter bugs. They need help sometimes when the software is working as intended. With thousands of users, even if your software has few bugs and an easy-to-figure-out UI, you’re going to field dozens of support issues every day. And once it breaks out of the early-adopter crowd, new users are going to be even more likely to be the sort of people who need help with seemingly ridiculous issues, such as what to do with a disk image.
So the conundrum is this: once a developer gets enough paying users to consider quitting his day job so he can devote full-time effort to writing code, he’s quite possibly got so many paying users that he’ll spend much of his time helping customers in ways other than writing code.
That’s why so few developers pull this trick off. Two that come to mind are Michael “C-Command” Tsai and Gus “Flying Meat” Mueller. I’m sure there are others, but the list of one-man-show indie Mac developers is short.
Which brings us to Brent Simmons, and last week’s announcement that NewsGator has purchased the rights to Ranchero Software, including NetNewsWire, and hired Simmons as a “product architect”, which means, more or less, that he’s been hired to keep working on NetNewsWire.
I get the impression that many NetNewsWire users were surprised by this. Not that NewsGator would want to buy NetNewsWire — people had been speculating about this ever since NewsGator bought Nick Bradbury’s Windows-only FeedDemon back in May 1 — but that Simmons would want to sell.
By the “make some cool software and sell enough copies to make a nice salary for yourself” theory, Simmons was, obviously, doing quite well. On a typical weekday, Daring Fireball’s main RSS feed is accessed by NetNewsWire from about 8,000 unique IP addresses; in the month of October to date, 21,274 unique IPs hit the main DF feed using NetNewsWire. And those are only the numbers for the paid version of NetNewsWire, not the free NetNewsWire Lite — add in another 15,468 unique IP addresses if you want to count NetNewsWire Lite, too.
Unique IP addresses do not correspond, one-to-one, to unique users, of course. One kid with an iBook wandering around a Wi-Fi-saturated college campus could be hitting it from half a dozen different IP addresses in a day. But it’s a good ballpark estimate, and so it’s hard to look at these server stats and come to any conclusion other than that Ranchero has sold many thousands of NetNewsWire licenses.
RSS usage is still nascent; there are still more people out there who don’t use an RSS aggregator but who will, soon, than there are people already using one. And NetNewsWire’s position in the market is stronger than ever, even in the face of new and free competitors.
Daring Fireball’s feed traffic certainly isn’t representative of the RSS market as a whole, but NetNewsWire’s presence is simply remarkable nonetheless.
Here’s a breakdown by percentage of aggregators accessing Daring Fireball’s main feed, from 1 September – 10 October 2005:
|NetNewsWire Lite 1.0.x||4.0|
|NetNewsWire Lite 2.0.x||26.1|
(The 18.6 percent “Other” is mostly a long tail of many different user agents, although a few percent of them are apparently using Firefox’s “Live Bookmarks”, but I didn’t feel like re-running the script to count them separately.)
Arguably, looking at my main feed is statistically unfair given that it’s included in the default subscription list for NetNewsWire. There are quite possibly many people who are subscribed to DF’s feed not because they’re regular readers, but simply because they never bothered to unsubscribe from the default feeds in NetNewsWire.
More fair would be to look at the stats for the special RSS feed for my Tiger Details Report back in April. The top four user agents accessing that feed:
|NetNewsWire Lite 2.0.x||14.0|
Even here, with a brand new feed that is not listed in the default subscriptions for any aggregator, NetNewsWire’s share is significantly higher than Safari’s, and no one else is really even in the same ballpark.
And the numbers from Daring Fireball members who subscribe to the members-only RSS feeds (full-content for regular articles, and the Linked List feed) are just sick:
|NetNewsWire Lite 1.0.x||0.9|
|NetNewsWire Lite 2.0.x||12.5|
NetNewsWire’s market share among Daring Fireball members is simply titanic. These numbers are interesting in a few other ways, as well:
The number of DF members using the paid version is almost five times higher than the number using the free Lite version; for the Tiger Details Report feed it was only about two times higher, and for the default (free) DF feed, the numbers are close to even.
Safari’s market share here is curiously low, I suspect because it doesn’t make much sense to use Safari’s RSS feature to read the full-content Daring Fireball feed; Safari’s RSS-reading UI is clearly optimized for excerpts and blurb-length posts, and if that’s what you want, you might as well use the regular excerpts-only DF feed even if you are a paid member with access to the full-content feed.
I don’t track Bloglines in my free feeds, because I don’t think there’s any good way of accurately gauging how many people are actually using it. Bloglines reports that over 3,000 Bloglines users are subscribed to my default feed, but that counts everyone who has ever used Bloglines to read my feed, even people who tried Bloglines out for a day and decided they didn’t like it. Whereas with the members-only feeds, it’s possible to get an accurate count because each member uses their own username and password to access the feeds.
NewsFire is way more popular with Daring Fireball members than with non-members. Same goes for OmniWeb, PulpFiction, and Shrook for that matter. I think this is simply common sense: people who are willing to pay money to support Daring Fireball are more likely than average to be willing to pay money to purchase software.
And so the thinking of those who were surprised by this deal goes something like, I thought Brent was doing great on his own, why would he sell the app to a larger company and take a job working for them, when he was already living The Life, working only for himself?
The thing is, Ranchero was doing great, sales-wise. The problem is that they were doing so great that Simmons wasn’t living The Life.
One thing worth noting is that Brent was not alone at Ranchero; he was helped, tremendously, by the full-time help of his wife and partner Sheila. Sheila managed a bunch of the unglamorous aspects of running the business, most especially serving as the first responder for customer support. The fact that Brent had Sheila’s full-time help is probably the only reason he was able to fly solo for as long as he did. In fact, it’s not even fair to say that Brent was in the same one-man-show category that Michael Tsai and Gus Mueller are in. Sheila was the secret ingredient that kept Ranchero rolling as the NetNewsWire user base grew into the thousands.
But NewsWire’s continuing boom pushed things to the point where even with Sheila’s full-time help, Brent wasn’t getting to spend nearly as much time as he wanted writing code. Instead, he was spending time helping customers make use of the code he’d already written. Novelists aren’t expected to help people read their books; musicians don’t help fans play their CDs; but programmers are expected to help you with their software.
When a solo developer reaches the point where an app becomes so popular he needs help, there are only several options. One is to hire help, to build a company around the products. Two long-standing examples are Bare Bones Software and Panic; Rogue Amoeba is an example of a newcomer. But there’s nothing easy about this. For one thing, even if you hire just one person, you’ve greatly increased the amount of revenue you need to generate. And it needs to be steadier revenue — if you’re only supporting yourself, you don’t have to write paychecks every two weeks. Running a company with even a single employee involves a ton of overhead; the whole point of The Life is the idea that you can just sit around all day hacking on code and watch the money roll in through Kagi or eSellerate. That’s not what it’s like if you own a software company with employees.
Another option is just not to worry about keeping up with the flow of support issues. Tempting, for sure, but certainly not a good way to build up a reserve of goodwill from your users. There’s a lot of overlap in the Venn diagram of the intersection between successful indie Mac developers and indie Mac developers with a reputation for being responsive to customers. From what I know of Brent Simmons, this was never an option for him — his personality is such that he’s incapable of being unresponsive to his users.
And then there’s the buyout solution. Instead of building a new company around the product, sell the product to an existing company. A decade ago, this was a common pattern for indie Mac developers — build a product, get some revenue, and then sell to a larger company.
NetNewsWire users are worried about this NewsGator deal. For one thing, NewsGator is largely unknown, or, rather, is largely unknown in the Mac world. Another basis for concern is the memory of other acquisitions that have gone sour in the past. (The old joke about Symantec is that their slogan should have been: “Where good software goes to die”.)
But mostly I think NetNewsWire users are worried because they’re confused, because they don’t understand why Brent would sell Ranchero if he was already doing well financially on his own. And so I suspect there’s some lingering concern that Brent has simply cashed out and will be walking away from NetNewsWire.
Not that there’d be anything wrong with cashing out, if that’s what Brent wanted to do, but I’m extremely bullish on the future of NetNewsWire, and reason is that I don’t think the motivating factor for Brent was the money, but rather the time. I got to talk to Brent about this extensively last week at the Web 2.0 Conference. I don’t know and didn’t ask the financial terms of the acquisition, but I think it’s safe to assume Brent did well by the deal. And but I do know and did ask what Brent’s role is going to be moving forward: he’ll be devoting nearly all of his time to improving NetNewsWire, which he was no longer able to do at Ranchero.
It ought to work out well for everyone.
Brent and Sheila get rewarded financially for their success and work over the past two years, and Brent gets to devote his time and energy to programming. By selling his company and working on NetNewsWire as a NewsGator employee, Brent has actually gotten closer to The Life than he was as an indie.
NetNewsWire users get a generous transition deal: two years of subscription service to NewsGator online, including synching storage, access to other NewsGator software, and all updates to NetNewsWire released in that time. Considering that NetNewsWire 2.0 was a free update, this makes NetNewsWire 1.0 look like one of the best deals in Mac history — if you bought NetNewsWire 1.0 in February 2003, you not only still haven’t paid a dime in upgrade fees, but you won’t for another two years, either. And not only will the free NetNewsWire Lite remain, but they plan to add synching features to it.
As for NewsGator, in the course of one week they’ve gone from having almost no presence in the Mac RSS market to being the most important third-party RSS developer on the platform. They now publish the best desktop news reader for the Mac as well as what’s generally regarded as the best news reader for Windows, FeedDemon. And they’re doing it right — they’ve kept the original developers on board, and they have no plans to merge the two into a Frankensteinian “FeedWire” or “NetNewsDemon”.
This is a story involving a lot of money, a lot of users, an ambitious company, and a popular new technology that’s heading into the mainstream. But what it really boils down to is a crackerjack programmer carving out a space where he can continue focusing on one of the best apps the Mac has ever seen.
Bradbury, like Simmons, was hired to continue developing the app he sold to NewsGator, and likewise has the title “product architect”. Simmons and Bradbury aren’t doing Mac and Windows versions of the same app, but are building two apps that more or less do the same thing, one for Mac and one for Windows. ↩