The 37signals Effect
lbrandy.com
lbrandy.com
Add new features. Improve the product. Make it do more. Make it kick ass. Put more work in. Make it scale better. Make sure your product is far better than the competition. Integrate it with other useful tools... etc etc. Maybe tackle a few "hard" problems.
Bad developers whine and complain that their product has been stolen.
Many people seem obsessed with the fact that Campfire is so much easier to write than, say, Mathematica. And so it is, particularly after 37signals gave away the toolkit to do so. But if you're trying to make money with a chat app, building it is the least of your problems: There's probably thirty or forty chat apps sitting on Sourceforge right now. Your biggest problem is Metcalfe's Law: nobody wants to use a chat app which their friends aren't using. What you need is publicity. Ideally, you want to convince half the readers of Techcrunch to try out your chat app in the course of a single week.
Google handed 37signals that publicity on a plate, and they ran with it. Complaining about your product's imminent extinction at the hands of a mean, nasty corporate behemoth might be a good idea even if your competition is Microsoft -- obviously, your odds of getting those guys to back down are infinitesimal, but at least people might hear of your existence, it might be better to burn out than to fade away, and you never know: Maybe there will be room for more than one company in your product space, and if you play your PR cards right you might emerge with a reputation as the "classic", "alternative" brand and be able to ride Microsoft's coattails to fabulous wealth. Or modest wealth, anyway. This is how Apple Computer managed to stay alive long enough to be rescued by Steve Jobs.
But 37signals got extra lucky because their opponent was Google, which treasures its reputation as a kinder, gentler company that is certainly not mean like Microsoft, and is unwilling to tarnish that reputation over what (to them) is a simple demo app. So, when challenged, Google folded immediately and left 37signals in possession of the field. More importantly, Campfire is now a household word, at least in households where people read Techcrunch or Digg.
Meanwhile, you're welcome to try to compete with Campfire yourself. Bolt on features, make it do more, add plenty of nifty-looking toggle switches and a skinnable interface, make it scale so well that it runs on a netbook, etc. Go nuts. Don't skimp on the work if you don't want to: Work day and night! Just don't come crying when you can't make any headway in the market because it turns out you've been working on the wrong things.
http://www.quantcast.com/campfirenow.com
In the world of chat, Campfire is pretty small. It's pretty funny for you to say they are in possession of the field - they've only done 10m chat messages since they started!
Some random math: Been going almost 3 years - 1000 days odd 10m messages sent! (From website)
That's an average of ONE message every 10 seconds.
MSN User: Okay, I use msn, but I can't get to it at work.
Non-Chat User: Um, I'm not sure how that works.
Yahoo User: I use yahoo, I'm busty_hot_chick_16@yahoo.com
Business User: Guys, just go to campfire.1234.com and sign in.
It's just a very solid, inexpensive private web chat that we don't have to host ourselves. It works everywhere, unlike IRC, SILC, and AIM, which almost everyone filters. It is completely brilliant for conducting meetings. We're all smart people; could we have built a substitute? Sure, if we wanted to be in the chat business.
So, I think along with "time" and "expertise", you should also add "tolerance for needless pain".
You know, last time I checked, the BSD devs were still using icb. Maybe Campfire is also for the people without the time or expertise to set up icb. =)
* IRC runs its own custom protocol, which firewalls go out of their way to block.
* IRC has archaic limits like 7-bit clean text and short lines.
* IRC requires custom client software.
* IRC can't natively include binary data, images, or rich content.
* IRC can't provide transcripts without more archaic software to log in and watch channels.
* IRC's authentication support is significantly worse than that of the web.
* IRC's authorization features are limited to chankeys and +i.
It's in use on millions of websites. It's the most used groupchat protocol on the internet.
So I'm not quite sure where you get your data from.
Maybe you personally hate IRC for some reason, but to say "I can't imagine why anyone would waste time using it in 2009." Is just idiotic.
By "in use on millions of websites", I presume you're cheating by saying that any website that mentions IRC is "using" IRC.
Here's a few more.
http://www.quantcast.com/p-1as4Vqt4eFyrI/traffic/syndicators
You could probably change Mibbit to use the MibbitChat protocol and most of your users would never care.
I already layer things ontop of IRC such as typing notify, recent chat on channel join etc...
IRC is a simple, solid chat protocol. It's about as ugly and broken as HTML IMHO.
You don't get users by forcing them to use some swanky new protocol/system off the bat. That's a recipe for quick failure.
You know, I made the same mistake in 1998; we started (and got funded) a company that basically took IRC and:
* Added arbitrary TLV encoded data to messages
* Used link state routing to solve the netsplit problem
* Used FEC to provide reliability over the resulting mesh network
But even though you've kept current on IRC and I wrote it off a long time ago, I'm pretty sure I can still win an argument that IRC is archaic and ready to go.
Those problems you 'solved' are not problems any user cares about though. They want IRC to be easy, and usable as a webchat.
But the world certainly wants to be able to upload files and images into group chats. No user cares about IRC; they want group chat to be easy as using a website.
I assume your startup failed. Perhaps that's tainted your view a tad.
You are the opposite of right; IRC is dying, and will be replaced with web chat systems with backend protocols nobody thinks about.
Why would I count page views? Counting profit is much more fun.
I think you'll see an explosion in online webchat this year and next, and IRC will be one of the clear winners. I don't like the protocol, it's a complete mess to be honest, but that's what'll happen.
Ah, I should have gone ahead and written "the field of battle", which is what I meant. Which, in this case, was their own small piece of turf, not the entire world of online chat. Obviously they don't dominate the world. Nor do they seem to want to. They seem perfectly content to own their particular chat niche, and they probably would be happy even if that niche were little more than "users of other 37signals products talking to each other".
Umm I'll bear that in mind thanks ;)
Good Internet entrepreneurs might do this. Good developers might be good Internet entrepreneurs but then again , they might not be. I know I'm a pretty good developer and I have some ideas about cool web applications which might make some money and some ideas about the most effective ways to compete. But, you know something? I won't kid myself that my development skills or my native intelligence would automatically tell me how to respond to competition. I respect marketers, managers and investors who literally make it their business to know how to compete.
"Bad" as in "being a bad sport", not incompetent.
After a while adding features is more about keeping customers and killing off the competition than it is growing the market. Most users probably use 25% of Photoshop, but they don't all use the same 25%. Which makes it much harder to compete with Photoshop.
They made a sample application to show off their new product. Is the problem that it's too 'similar'?
I'm guessing the exposure would have been far less if 37signals just kept quiet and ignored the google app. I sure would have not known about the google sample app, and if I had I would have assumed that it was just a barebones concept, not something I could depend on or get support from if it didn't work.
My general rule of thumb is to always act as though any competition does not exist, because otherwise your simply advertising them. This means having features that 'no one else has' so you don't even have to name the competition.
Why risk PR problems for a silly sample application?
It's that their marketing advice (like Joel's) can be a little off-base. Both created a HUGE cult following with great writing. This cult was a perfect match with their products (webdev geeks for 37s and all dev geeks for Fog Creek). It's a beautiful sales engine, but it isn't particularly repeatable unless you are charismatic, a great writer, and are dedicated to creating a community like this.
In absence of this particular bottled lightning, you have to build a sales/marketing engine with some combination of sales guys, adwords, print ads, SEO, viral, whatever... AND (of course) a great product that keeps getting better.
(this is all assuming you sell stuff to businesses).
If you choose to build an app that is a "simple solution to an obvious problem" then, yes, obviously it will be easy for competitors to emerge.
The author starts on the right track, but ended with an extremely weak conclusion. Of course developing RoR has helped them. That doesn't mean another developer can't reap similar benefits from releasing open source software. Heck, I bet there are developers who have benefited from releasing rails plugins.
IMO, 37s hit home run by (a) building and releasing rails and (b) initially creating products that are targeted at many of the people who use rails (basecamp & campfire are for developers/freelancers).
If they built rails and then came out with an app for insurance agents, well ... there obviously wouldn't be the same amount of overlap
Instead of articles like this, I'd like to see someone say "here's how to translate Getting Real methodologies" into X industry.
There are a lot of different ways to do things. We only have the answers that work for us. We don't pretend to have your answers. That said, we don't preempt everything we say with "it depends." Reasonable people understand that everything depends.
We simply share our way based on our experience. Take whatever value you find and leave the rest behind.
From the front pages of our book: http://gettingreal.37signals.com/ch01_Caveats_disclaimers_an...
"We acknowledge a large part of our own success is due to our popular blog. Unless you have a popular blog, some of the advice may not work for you"
From "Founders at work", Joel Spolsky when asked for advice for startups:
"I would recommend that you create a weblog and have millions of readers every month from around the world that read it. That's not really necessarily followable."
I thought that was a pretty cool honest answer.
Companies that write a blog with 5 subscribers are simply not playing in the same league. Yes, of course, you have to start with 5, and I'm sure you did at some point, but to suggest that you can apply X ideas to achieve Y without discussing these enormous 'external' conditions that are directly attributable to this achievement is not telling the whole story.
Which is longer than you've worked on any of your software, if I'm not mistaken. That would, in some ways, make it the biggest barrier to entry. The software is good, but certainly not impossible for good people to replicate in a relatively brief time, so what really seems to differentiate you is the blog/rails/following - that is something that cannot easily be replicated.
In all fairness, some of this may be a by-product of the very close association between 37signals and Rails; for all its strengths, I feel the Rails community often exhibits some of the same characteristics.
e.g. saying "My problem with 37signals is..." instead of "I think many people's problem with 37signals is..."
The former gives you possession over your subject, the latter is both redundant in saying "I think" (you wrote it, of course it's what you think!) and uses "many people" as a crutch that requires the ability to read minds to verify. It isn't that it's an false thing to say, it's just empty of meaning, and should be omitted.
Here's the archives for the magazine "The Smalltalk Report" which featured an article series "Getting Real" http://www.macqueen.us/stIndex.html
Here's a great article to get you started: http://www.macqueen.us/smalltalkReport/ST/ST01/15al.pdf
[EDIT] this is not bashing _anyone_. Just recalled the article series while reading this blog post and thought there may be some people here that have never read The Smalltalk Report. Good stuff!!!
For a product company, 37Signals does a lot of blogging, conferences, a book and so on. I think the guys are good at this, and they really enjoy it. It makes me think that, in a way, 37SIgnals is really an educational/PR company, that happens to make money from their software. If you want something like what they make, you'll need to pay someone to host it - why not them, since they are nice guys?
It might therefore be a good start to read Seth Godin's Tribes as well as Getting Real rather than just Getting Real on it's own.
the RoR advantage is not really copyable, and provided them with a huge advantage in terms of fans and press. But it did not win them most of their customers and going beyond a niche of small web dev shops its meaningless. My guess is most of their customers dont know what RoR is.
on a side note it would be nice to do a 37signals moratorium for a month or so, seems like all I see lately on HN lately.
What this leaves out completely is the approach of being different. Why do it the way everybody else is doing it? IMHO there is a huge potential in a lot of areas which could be used by a different approach to solving problems that have already been solved.
Even if what you are saying is true, you've missed an all-important middle step. They won the hearts of the nerds. The nerds wrote blog posts, subscribed to their blog. They got picked up by the nerd media outlets. This turned into actual media attention. Their page rank is high. They have numerous "real media" write ups. They have their own wikipedia page. To your average "boss", they are "legit".
"Add features!" vs "Don't add features!" "More is more!" and "Less is more!"
All are wrong/overly simplified.
Just because "Less is more!" is not right, doesn't mean "More is more!" is right.
"If you don’t make the full effort to understand what is going on, and you just believe imitating some subset is going to reproduce the results, you’ve entered cargo cult territory."
Uh, that is exactly what he's saying, no?