Spolsky: Does Slow Growth Equal Slow Death?

I work at Atlassian and I kind of assume it's us, partly because everyone else around the Internet seems to think he's talking about us, but mostly because the Fogbugz 7 manifesto talked in all caps about people threatening to "GO OVER TO THE AUSTRALIANS".

There's a good chance he isn't, though, because the company described in that article doesn't really bear a great resemblance to Atlassian.

The business models of Fog Creek and Atlassian are far more alike than they are different. Both companies are developer-founded, private, non-VC funded and growing "within their means". Both aim for bottom-up adoption rather than selling top-down through CIOs, and (although this is apparently changing at FogCreek?) neither have a traditional sales team. Both even started off writing an issue tracker (JIRA, FogBugz) then followed up with a content management app (Confluence, CityDesk).

Before this year we didn't have a growth strategy beyond "sell lots of software, and use the money we get from that to hire more people to make more software and sell that." We've recently got (temporarily) a little more aggressive about hiring because we figured the GFC was a great time for a profitable company to snaffle up any good people who were at loose ends, especially when some of our VC-funded competition were being forced by the same circumstances to lay developers off.

So yeah, we're hiring. http://www.atlassian.com/32/ :)

pseale• 17 years ago

I'm going to guess TFS/VSTS from Microsoft, which is both a relatively new product, and fits his description of aggressive expansion.

edw5194• 17 years ago

If increasing your growth rate is your objective, this looks like a very nice first step, Joel. You disguise a PR piece as an objective "how to" in a national business publication, coming across as an authority, the underdog, and an all around nice guy who really cares about his customers.

You're an engineer who claims to be "weak" in the sales department, but all evidence to the contrary: nice "sales hack". Any reasonable person who takes your advice would be a fool if he bought from your competitor. Not bad for a couple hours work. Kudos.

spolsky• 17 years ago

Jeez, everything is a conspiracy on Hacker News. How about I wrote it because I'm a columnist at Inc and I have to turn in one article a month? And how about, they hired me because they like columnists who are actively running businesses to write about the issues they face? (c.f. Norm Brodsky, the other columnist, who has some kind of a box storage business).

None of the plumbers and dog shampoo-vendors who read Inc. do software project management. The number of leads I get from Inc readers is laughable.

Also, you're confusing sales and marketing. They're different things. We're pretty good at marketing for a company our size. We're absolutely bad at sales.

tptacek3• 17 years ago

I don't think edw519's being sarcastic. I got the same vibe he did off that article, and I admired the article all the more for it.

edw5194• 17 years ago

Yikes, lighten up, Joel. I really meant what I said. I have no idea how many leads you get from inc, but I bet you get a few more with that piece.

What I didn't say in my original post was that this was an excellent example of generating interest in your business while contributing to the community at large, and that quite a few of us here can learn from it. But now, I'm afraid to give you another compliment because you may misinterpret it. (Oops, too late.)

There have been many times when my post was misunderstood because I accidently omitted the <sarcasm> tags. But this is the first time the inverse ever happened. Live and learn.

</nosarcasm>

oldgregg• 17 years ago

I for one think your writing is well suited for Inc.

wallflower2• 17 years ago

Atlassian Confluence (the corporate wiki product from Fog Creek's probable competitor) has pitiful search functionality, among other things, for all the revenue that product probably generates from site licensing.

That being said, Atlassian Confluence's "Wiki" is updated nearly every minute (according to the homepage dashboard) at my company. It's not just a Wiki; it's a Wiki that normal business users can use. Fog Creek seems to think they can take on their competitor by mainly focusing on features while IMHO Atlassian's strength is producing an adequate product that works and more importantly makes the person at a company who leads and drives its adoption a likely corporate hero. Corporate sales is tough but lucrative.

> "What set Oracle apart from Ingres," Moore writes, "was that [CEO] Larry Ellison drove for 100 percent growth while Ingres 'accepted' 50 percent growth."

To achieve 100% growth: Oracle salespeople were fired if they did not achieve a doubling of their sales quota on a yearly basis.

chrisbroadfoot2• 17 years ago

More specifically, the competing product he's referring to is probably Atlassian's JIRA (bug tracking software).

I for one thought that Joel sounded a bit sour and jealous. Is it that Fog Creek is losing its "market leadership position"? (was Fog Creek ever a market leader?)

pchristensen2• 17 years ago

I wouldn't say jealous, more like disoriented to realize that his business hypothesis might be invalid. He designed his company to be a high quality competitor among many, but now he's got to face a category killer which requires very different tactics. A core part of his business has to change to something he doesn't necessarily want.

jshen2• 17 years ago

It's a running joke at my company about how bad the search is in confluence. It's shockingly bad.

wallflower2• 17 years ago

One thing that helps (a little) is to try their "Boolean" search support. Use all-caps AND and OR. Example: svn AND instructions AND command-line

lawrence• 17 years ago

This is the best argument for considering venture capital that I've read in a while, and it barely mentions venture capital.

tptacek3• 17 years ago

Not so fast. First, Spolsky probably doesn't need VC based on their revenue; they can staff up a sales team without going out for more money. Second, Spolsky got to spend ~10 years turning the dials to figure out what works for his product. He's in a good position. He has the option of increasing growth, and that option is more attractive right now than maintaining current growth.

No VC funded company has a 10 year runway to figure out how to get their product right. Very few VC funded companies get 10 years, period: the investment needs to liquidate at some point.

Finally, Fog Creek all but stumbled into their current product. They had a "VC-ready" product when they started (CityDesk). If they had gotten funded, they'd be dead now.

SwellJoe2• 17 years ago

If they had gotten funded, they'd be dead now.

On what do you base this theory?

I don't know that you're wrong...I just don't see how it naturally follows from anything we know about Fog Creek.

zaidf• 17 years ago

I am all for not being secret. But there are exceptions and in this case, if I am really mapping out a specific strategy to go after a competitor, I wouldn't blog/write about it. I'd be curious to hear where he draws the line in sharing strategic plans.

The startup I am doing right now involves public data. I've been able to learn an incredible deal from my competitor just by googling around and finding the name of my competitor in the minutes of public meetings.

On a separate note, this makes me feel really good about my decision to take Sales Management this semester:)

albertsun• 17 years ago

Joel's blog is an excellent marketing tool and I've read and enjoyed a lot of posts there and have a greet impression of Fog Creek Software. I know all about the working environment, what their philosophy in hiring is, how they treat their programmers, etc. And I've learned a ton about software development processes and working with others.

One thing that I haven't figured out in all this time reading is, what products does Fog Creek actually make? I still really don't know what the products are, or what they do.

In his last post he wrote,

FogBugz, which is all about giving developers tools that gently guide them from good to great.

It's a nice description, but it still doesn't really tell me what FogBugz does. Maybe Joel could blog about how he identified the problem for the product to solve, what specific features he's really proud of that make software development easier, etc.

astine• 17 years ago

From what I recal, it's a bug tracker.

tedunangst• 17 years ago

Go to www.fogcreek.com, it describes all the products.

"FogBugz manages projects, tracks bugs, and even tells you when you’re going to ship..." It continues from there.

philfreo• 17 years ago

It's a bug tracker. Plenty of his articles describe it in more detail, if I recall.

elviejo2• 17 years ago

Plus they make Copilot a tool to give Tech support based on VNC.

elviejo2• 17 years ago

I'm I the only one that thinks this article is the stick part of the Hockey stick?

In other posts (http://www.joelonsoftware.com/articles/fog0000000017.html) Joel has said that: "Good Software Takes Ten Years. Get Used To it." And then makes reference to the hockey stick graph of sales figures for a couple of products

So ten years have passed for FogCreek and seems that now they are focusing on the stick.

I would say this is the next logical step in that strategy.

pchristensen2• 17 years ago

Joel knew you need developers to make the flat part, but it sounds like he's realizing you need salespeople to claim the steep part.

RyanMcGreal• 17 years ago

> If you're growing at 50 percent a year, and your competitor is growing at 100 percent a year, it takes only eight years before your competitor is 10 times bigger than you.

Assuming you both start out the same size.

MartinCron• 17 years ago

Even if you don't agree with what he has to say, his writing, especially for Inc. Is really good.

I'm having a hard time seeing the comparison of FogBugz vs. Mystery competitor (consensus says Atlassian?) and Oracle vs. Ingress. Picking a db vendor has a lot more lock-in and switching costs than using a different bug database. Also, there are so many more competitors in this space, with plenty of good free options. Lastly, the barrier to entry is so much lower, I mean, I could write a usable software project management tool in a weekend, it's trivial. It would take at least a week to make a fully functional relational database system.

ghshephard9• 17 years ago

Joel is definitely talking about Atlassian in this article. In particular, one of their products, jira, goes head-head with FogBugz. Joel has actually referred to them in previous articles, in terms of customers threatening to go to the "Australians" if he didn't implement such-such a feature. (http://www.joelonsoftware.com/items/2009/07/20.html)

Atlassian is really the Major Grower in this space - good sized developer community, and "App Store" equivalent (Plugin Exchange ) - They have about 50,000 plugins a month downloaded from a library of 400 - 130 of which are for Jira.

The latest rev of their product, Jira 4.0, has one major feature - a SQL like query engine, they call it "JQL" - Jira Query Language - which allows you to do everything you might expect/imagine. Their searching was _already_ better than FogBugz, but JQL makes you not even not want to compare the two.

Interesting Notes:

Google for "Bug Tracking Software" doesn't bring up FogCreek, but does bring up Atlassian.

Atlassian is like the "Anti-Fog Creek" in terms of it's thoughts around office. I've been to their San Francisco Offices (which are supposedly like their Sydney) - One Big Wide Open Space. No Cubicles. No Offices. Nothing - just pure Agile openness.

Go to Atlassian.com - you are 4 Clicks away from having the software on your computer, and it takes about as many seconds. And it figures out whether you want OS X stuff, linux files, etc...

Go to FogCreek.com - 30 seconds later I _still_ couldn't figure out where to click to try their stuff - and when I do, I finally land on a page where I need to start providing Email Addresses, Phone Numbers, accepting terms of services. It's almost like they are _daring_ you to try the competition.

What's cool, is that _both_ Fog Creek and Atlassian are the ultimate results of Startups going head-head. Someone will have to write a book about these two companies.

http://www.atlassian.com/about/history.jsp Atlassian: "In 2002, fresh out of university, Scott Farquhar and Mike Cannon-Brookes, With total start up costs of $10,000 charged to a Visa card, they built JIRA, a professional bug and issue tracker, and Atlassian was born. By 2004 the company had grown to six developers."

http://fogcreek.com/About.html Michael Pryor founded Fog Creek Software with Joel in September 2000.

Fog Creek's Major Problem, and the reason why I suspect Atlassian is going to wipe them out if they don't get their Act Together, is that Fog Creek has, to my knowledge, three products:

  o Bug/Project Management/Software Project Tracker (fogbuz)
  o VNC Hosting Service for Tech Support (Copilot)
  o Forums for Software Developers/SysAdmins StackOverflow/ServerFault
  o 25 Employees

Atlassian, on the other hand, has:

  o Issue Tracking Software (jira)
  o Wiki (confluence)
  o Source Code Reviewer (fisheye)
  o Code Review (crucible)
  o Continuous Integration (bamboo)
  o Code Coverage (clover)
  o Identity Management (Crowd)
  o 200 Employees
  o 15,000 Customers
  http://www.atlassian.com/about/customers.jsp

I don't want to come off sounding like an Atlassian fan-boy, particularly as I love reading Joel's articles, and I don't think I've ever read anything, ever, out of Atlassian that caught my attention - but he has his work cut out for him. Of course, maybe Joel needs someone like Atlassian to shake him up and get him going...

ghshephard9• 17 years ago

Interesting to look at their pricing for jira vs fogbugz - Clearly each of their pricing (for anything but the smallest companies) is almost the same - basically they discount the software and profit off the maintenance contracts):

Fog Creek: $10,000 + $5000/year Atlassian: $8,000 + $4000/year

http://fogcreek.com/FogBugz/PriceList.html http://www.atlassian.com/software/jira/licensing.jsp

But, Atlassian is smart, and gets you when you are small:

  50 Person Software Shop:
  Atlassian: $2200 + $1100/year
  FogCreek: $8000 + $1825/year

And, what if you are a small startup - Where FogCreek/Atlassian were 2 Years after they were founded,

  10 Person Software Shop:
  Atlassian: $10
  FogCreek: $1899 + $365/year.

Given that the products are _roughly_ equivalent, and the customer has decided for some reason not to just use Bugzilla or Trac , which product do you think your typical 10 Person Small Software Shop will install.

Or, if you are using Bugzilla, but you are a cranky release engineer, and you want to replace it with something with a little more horsepower (not to mention benefiting from the QA/Regression of all those customers of Atlassian/FogCreek) - which product do you think you might be able to get away with expensing.

Now, finally - You have grown from 7 engineers to 20 Engineers - there is a CLOSE TO ZERO chance that you will EVER switch from FogBugz to Jira (Or Vice Versa) - But, Jira is starting out with a lot more customers, so they experience a profit wave.

Basically by the time the companies grow to 100 Users (where FogCreek/Atlassian are making profits off of the Maintenance, as it's been 3+ years since anyone has called for support) - Atlassian has a much bigger slice of the pie. This does NOT bode well for fogcreek in the coming years, as all of that Atlassian market share finally starts to pay dividends, further accelerating their growth, allowing them to invest more in their product, and further gaining market share...

And that, is one major reason why Atlassian is growing faster, and eating Joel's lunch.

Ironic, given the essays that Joel has written on pricing. I realize that he probably doesn't want to price his product at $10 and make it sound "cheap" - Atlassian was brilliant, and they did the following:

"Get started for $10 All proceeds go to charity" - "www.roomtoread.org" - it's like they READ Joel's article and said to themselves "Well, we can't cheapen our product to gain market share, but we _can_ raise money for charity"

Carlfish3• 17 years ago

While your analysis of why we're doing the $10 deal is spot on, we only introduced it a couple of months ago so any benefit we might see is still in the future. Previously the cheapest edition of JIRA was (if I recall correctly) $1200.

johns• 17 years ago

A couple additional points that aren't readily available:

1) Fogbugz has an integrated wiki, but I don't know how it compares to Confluence and you can't get it separate from Fobgugz.

2) Fog Creek has a code review and hosted Mercurial source control product that's in beta called Kiln.

3) You can get FogBugz on Demand for free with the student and startup edition, but that's only for 2 users.

gommm• 17 years ago

Wow the 10 person software shop promotion is interesting... I'm currently starting a company and was looking for a good issue tracker.

Up until now I've been using fogbugz for other projects but the license is a bit steep when starting up...

Poiesis• 17 years ago

Hilarious. Joel's said previously that site licenses and volume discounts are a bad idea ("Bad Idea #1: Site Licenses. "):

http://www.joelonsoftware.com/articles/CamelsandRubberDuckie...

ghshephard9• 17 years ago

Another Data Point that it is Atlassian he is talking about:

2004, 6 Employees -> 2009, 200 Employees = 100% Annual Growth, as quoted in the article against FogCreek's noted 50% Growth.

holdenk• 17 years ago

Maybe you and I are using different local Google's but for me googling for "Bug Tracking Software" both with and without quotes brings up Fog Creek in the first 5 (as of Nov/3/09). On .com atlassian is 2 bellow Fog Creek. On .ca it brings up Fog Creek right after atlassian

ghshephard9• 17 years ago

Apologies - I was unclear and misleading. I meant to say "The Advertisements" for "Bug Tracking Software" don't show Fog Creek, but do show Atlassian.

nl• 17 years ago

"I don't think I've ever read anything, ever, out of Atlassian that caught my attention"

In the old days Mike Cannon Brookes used to blog a lot. I think there is a good argument that Atlassian did more to save the Java platform (back in the dark 2001-2003 period) that any other company (including Sun).

They started javablogs.com, which back then was the real hub of open source java development. That encouraged Sun to start their blogging push, and that engagement was what protected Java against what seemed back then to be the inevitable rise of .NET. (Obviously the JBoss open source project and Apache/Jakarta had a lot of influence then too, but Atlassian was the most visible company during that period)

michaelneale2• 17 years ago

And its all about bug trackers. Gosh this is a boring business fight to watch - well I guess its slightly less trivial then facebook apps ;)

ctrager4• 17 years ago

As a developer, I say not trivial. This isn't just about bug trackers, because bug trackers integrate with version control, maybe our IDE, and soon code review tools.

A bug tracker can be like email, on the screen all the time, with a lot of our communication with co-workers going thru it ( and in fact the good ones incorporate email into them very well. )

So, it's the software we are living in. Next to our text editor, it's maybe the MOST important software.

Not counting Pandora, of course.

lucifer• 17 years ago

From the history article:

"In addition to eschewing traditional enterprise software, the company forged a different sales and marketing model, relying on word of mouth and selling their products online without the use of an outbound sales force. The company remains private and has no institutional or venture capital investment."

swilliams• 17 years ago

I'm helping a new non-profit startup get going and we ultimately went with Atlassian stuff since we just plain and simple cannot afford FogBugz.

I've used FogBugz and like it quite a bit, but the email response I received from them basically reiterated what was on their website: no discount for groups over 2 people.

If FogCreek came up with something competitively priced as Atlassian's "starter pack" I'd consider switching, but there is also a significant cost of moving all the data over the more we grow.

ctrager4• 17 years ago

Disclaimer: I'm the author of an open source bug tracker, BugTracker.NET, a sorta hobby project that got better over time and sorta, kinda, "competes" with FogBugz. Also, I receive money from Atlassian to display their ads on my website, but other than that, I know nothing about their business. So, I'm not any sort of insider. But I do pay attention to the space, and here's my two cents.

Some of the income for these companies comes from the hosted versions of their apps. I think free and somewhat viral services like GitHub put a lot of pressure on Atlassian and FogCreek. 1) Downward pressure on what they can charge. Consider Atlassian's recent price drop. and 2) Classic Joel "Fire and Motion" pressure to match the competition's features. Kiln is FogCreek returning defensive fire.

Not to mention less viral players but really nice playes in the hosted space like Unfuddle, Assembla. Christ, even Google.

So, along with hosted apps, some of the income comes from installed software. There, both companies compete against open source, like Trac, Redmine, and even my own BugTracker.NET. Traditionally some stodgier companies have tended to be fearful about running open source alternatives, but there might be a generational change going on. There could be a tipping point where open source developer tools start being perceived even by the stodgier companies as being the more comfortable, safe, mainstream choice. Safer than commercial. That might already be happening with version control.

ghshephard9• 17 years ago

Note: Atlassian just increased their prices. Our legal department was going to use their professional 3.x license to manage issues for 300 employees (You'd be amazed at how far people push Jira - a nominal software defect tracker is used by my company as a ticketing system, time tracker, project management system - we've got north of 20,000 issues entered and tracked through it) - The previous price was $2400 on a quote expiring on Oct 24th. The New quote (for 4.0 "Enterprise" - no more professional) is $8,000. Both Cheap, but the new prices certainly aren't a decrease.

I doubt that many enterprise size companies (100+ employee companies) host their issues at either FogBuz or Atlassian. Security. I'm betting the vast (90%+?) of their revenue comes from license sales.

BTW - Don't forget Bugzilla as a bug tracker - we were quite happy with it until we switched to Jira (ironic if you know where the name comes from).

And you are right - Generational Change is happening. Right now. All of the people who where in their 20s and 30s In the late 90s, are now Managers, and Directors - and have some clue about open source, so it's coming into companies from the top, and bottom now.

ctrager4• 17 years ago

You are contradicting my unsupported opinions using actual facts? That's not fair.

Regarding "You'd be amazed": No,I wouldn't. Whatever the humble origins of a bug tracker, user pressure does push them to accrete more features and complexity, so even my own little BugTracker.NET also has evolevd to be also a ticketing system, time tracker, and a tiny bit as a project management system.

(Please let's not talk about permissions and customizable, enforcable, workflow...)

Here's my favorite way I've seen BugTracker.NET stretched: http://frap.cdf.ca.gov/projects/hazard/btnet/bugs.aspx, by the "California Department of Forestry and Fire Protection".

You're probably right about the 90% for now, but I think the Generational Change thing also affects attitudes towards hosted versus installed.

And, you have more and more big companies whose OWN business depends on THEIR customers trusting THEM to host their customers data, so I think that is a force for attitudes about hosted solutions changing too.

I work for a company that makes software for futures traders. Our customers include the trading depts of the biggest banks you can name. Presumably security conscious and with the expertise to manage the software/hardware themselves if they chose to. We offer both installable and hosted solutions. Plenty of these big banks have opted for the hosted solution.

BerislavLopac• 17 years ago

So the main point of this article is that the biggest risk a company can take is to not take any risks, right?

stcredzero• 17 years ago

I think the reasoning in this article also applies to programming languages. By that reasoning, it's by far better to have a crappy programming language that panders to mainstream expectations than to have something innovative that gives you a more powerful paradigm.

I also think that sometimes the more powerful paradigm still trumps sheer numbers.

rfreytag• 17 years ago

Perhaps Joel has discovered why VCs are so dismissive of 'lifestyle' companies.

If 37Signals starts to see market erosion then we'll know the VCs are correct and its grow fast or wither till irrelevance.

Personally, I hope the VCs are wrong and that in fact what Joel has to do is undercut his competition in a way they cannot answer without destroying their business model. This is what Clayton Christenson talks about in "The Innovator's Dilemma."

geebee2• 17 years ago

I'm a little sad to read this article. Are you always in a position where slow growth eventually means slow death? Or is it possible to find a niche where you can emphasize craftsmanship, quality, and a decent pace?

The reason I ask is that a lot of software developers actually like the idea of working for a small, highly profitable firm that stays, well, small and highly profitable.

If it is possible, what are some of the strategies you can employ to identify and/or hold on to this niche?

BTW, when I say "small", I don't mean small in the sense of total profits or number of users. I mean "small" as in craigslist small - a company that can keep the headcount low, avoid employing lots of non-technical workers (like sales people), and hang onto what makes a "small" company such a great place to work.

hello_moto• 17 years ago

Not many owners have the mind of Craig. Most owners of small businesses I worked for didn't settle for a nice beach house, they want a yacht too! and they want it fast! like ... tomorrow.

I blame humans for being greedy.

geebee2• 17 years ago

Eh, I wouldn't really "blame" someone for wanting those things (some would say greed is good, others (like me) would say that the honest pursuit of wealth isn't "greedy", but a lot of that just comes down to how we define the word).

My question is for those people who actually prefer to stay small - how you can find a strategy and niche to make this possible.

I suspect that a big part of it is how and where you sell. Vendors of for-pay developer tools need to convince a corporate procurement department to hand over the cash, whereas craigslist merely needs to convince people to post want ads, either for free or for very small payments. As a result, I suspect that the "sales people" approach wouldn't do much for them.

Joel went with the micropayment thing for a while, but maybe his space isn't a really great one for a small niche-company? Hard to say, since developers are a pretty unique group of people who might work well as a niche market?

petenixey• 17 years ago

Awesome discussion. It's been what 10 years since people really started blogging and we started seeing the progress of companies live and as it happens?

We've seen what happened to Google/YouTube etc. and how they moved into hypergrowth and uber-size but this is now the beginning of the next phase for the non-VC (bona-fide?) companies.

Really interesting article Joel, thanks for sharing.

Spolsky: Does Slow Growth Equal Slow Death?

Related Stories