>I can tell you that nothing we have ever done at Fog Creek has increased our revenue more than releasing a new version with more features.

That may be true for Joe Spolsky. But he is one of the highest profile software people in the world. The 99.9999% of us who are less famous should probably spend more time on marketing.

For the rest of us, simply releasing update after update works to grow organically. Lots of things look for the "new update" signal.

Yes but at some point your feature updates will lead to bloat and maintenance issues. There is a problem with accumulating features before getting customers too. You will not know which ones are actually wanted and which you should have cut. But of course any feature will have some users so removing them afterwards will lead to complaints.

In my case publishing a relatively unknown app, any kind of press obliterates the increase in customers due to features.

I guess 80% of my revenue was made on days of and immediately following press releases.

You don't have to release new features. Customers (at least mine) appreciate tightening up current features almost as much.

I never tried the press release route, sounds like you had good success with it?

There is also quite a gap between what current users want (mostly no surprises and stable software) and what new users think they want (more features, I believe many have the mindset "it does all this but I really want X before I commit").

And of course journalists only want new releases and new large features. Which is why getting features is usually only feasible on launch.

Depends on definition of success :) My app is not very successful in that it brings in only a few k$ per year.

However it has been featured here and there and these features were always massive spikes.

Reaper goes this way. They have regular updates with fixes and the rare huge new feature like the notation editor.

Even the big guys. I moved from sublime to VSCode just because the relentless updates spelled the writing on the wall.

You are likely right. Yet I wonder if

a) Fog Creek had that high a profile when they put that advice into practice, and if not...

b) The people that Fog Creek had such a high profile among included its prospective/repeat customers (as opposed to just non-customer fellow software developers.)

I believe Fog Creek had (at the time of writing this) already been in both situations at various points. Joel wasn't famous at the start.

But the context of this statement isn't really related to fresh startups. it's aimed at software companies... presumably with existing customers and revenue.

And yes, it's true that there are lots of exceptions at any level... a big press score or whatnot. The underlying point is solid though, imo. Most software companies best strategy is to make more/better software for their customers to buy/use. Shiny, big, spiky "tactical wins" are neither necessary nor sufficient. They just loom large in our narration of events.

a) Yes. Joel had a very well written blog that was very popular with developers (I miss your blog Joel!).

b) Yes. Back then his customers were developers.

I recently asked someone who helped Fog Creek with sales-y stuff about this quote.

He said it had to be taken with the fact that Joel Spolsky's blog in its heyday had a tremendous amount of traffic from developers. He implied that it doesn't universally hold true.

Sounds like you asked a marketing/sales guy this question, and you got a marketing/sales guy answer.

I think the answer is pretty clear. If you already have a channel to customers you can focus on the product. If not, increasing reach can be much more beneficial.

That can work for others too. Maybe building an audience via a blog can be worth more than paid advertising, depending on the target audience.

You can interpret "new version with more features" as Product. Basically, he's saying that Product trumps Marcom.