Commoditize your complements (2002)

https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/

by clomond • 5 years ago

62 25 5 years ago

wunderflix2• 5 years ago

I’ve read about this before. I can’t you tell you why, but somehow I think there’s something about that theory that doesn’t feel right. I can’t quite put my finger on it. Maybe someone else can?

bigbob2• 5 years ago

At the end he argues that desktop software is not commoditized but look at Sublime Text vs VSCode - granted it was probably released 15 years after this article came out, but I think that's a pretty good example of commoditized/interchangeable desktop software.

aidenn03• 5 years ago

1. He said software is hard to commoditize, not that it isn't. Google docs commotidizes MS Word for a reasonably large audience, but there are a million features of Word that google docs doesn't have, and if you use even one of them, it falls down.

2. I don't use VSCode nor Sublime Text, but are they really so similar that people switch between them freely?

As a comparison, I use vim. The best vim-emulation layer I've ever used is evil-mode on emacs, but even that has some hiccups that make it hard for me to switch between them (I hate that yanking to the default register on evil-mode also yanks to the system clipboard).

PeterWhittaker• 5 years ago

Well, it may be a case of what Jobs referred to in his quote about being able to connect the dots only when looking backwards, that in the present moment, one has to hope that the dots will connect.

What he refers to as IBM's RedHat strategy may have been their motivation, but it's hard to discount the value of a proven, growing revenue stream; they may simply have thought "hey, revenue, plus some possibilities!" Likewise commoditizing the PC peripheral market: I could buy it, if they had made big money in that market. The fact they went on to spend gazilladollars on OS/2 (which had its moments) and got M$ to write it (and gave Gates an opportunity to have his engineers learn the 386 in detail while not providing any transferable knowledge to IBM, since OS/2 was written in assembler...) makes me think Joel was cherry picking.

That said, it is an interesting argument: If you do a thing that makes something else desirable AND you can make decent margins on the other things, especially if they are consumables or have a repeating revenue stream, it may be worth quite a lot to do the thing.

I'm not wholly convinced, but I am intrigued.

blacktriangle2• 5 years ago

My guess is that this is a strategy for dueling giants. You look at his examples, it's MS, AOL/Time Warner, Sun, Oracle. This is not really a strategy for startups or even successfully midsize companies. Those companies are too busy working to make money off of their own products. To commoditize your compliment you have to have so much extra capacity you can develop a product at a quality level of a midsize company and then give it away for free.

So it's a very interesting theory, but its also pretty much irrelevant for most of us.

WJW• 5 years ago

I think this hooks into the "losers game" post from a few days back: most startups are still in the phase where they are overwhelmingly more likely to die because they fuck up than because some competitor manages to outmaneuver them. Big tech is profitable and won't die just by itself, so in that theater clever tricks like commoditizing your complement becomes much more important.

AnimalMuppet• 5 years ago

I don't see it that way. Take Bic, for example. They commoditized the razor handle in order to sell more blades. But for that to happen, they didn't need to be in a duel with some other giant who made their money selling handles. They just needed to decide that they'd make more money selling more blades than they would make selling more-expensive handles.

nostrademons5• 5 years ago

It's relevant to startups, though, because picking a vacant market where the complements are already commoditized is a winning strategy for a startup. It was like that for most of the tech giants of today. For Amazon, books were already a commodity, shipping was a commodity, and online checkout became a commodity shortly after they started. For Google, websites and hyperlinks were a commodity. For Apple, chips were a commodity, TVs were a commodity, floppy disks were a commodity, and programmers who knew BASIC were a commodity. For Uber, drivers were a commodity, cars were a commodity, and cell phones became a commodity.

The main difference between startups and big companies is the resources available to them. Startups do not have the ability to meddle in markets other than their own product, because they simply don't have enough people. They have to put all their efforts into their main product, because they don't have much effort to spare. But the most successful startups are the ones where the rest of the ecosystem already exists and they're just putting the keystone in place to change how society functions.

jbjbjbjb• 5 years ago

I think it makes sense in the right conditions.

Your product has a complement, that complement’s quantity demanded can be increased (maybe the price could be lowered by removing a monopoly or reducing costs) thereby increasing the demand for your product (your demand curve shifts right).

But I guess whether your business can capture any of that increased demand is another story.

gwern• 5 years ago

I have a page on this topic with many additional examples (most from after Joel's post): https://www.gwern.net/Complement

mrkramer• 5 years ago

Amazing article.

Commoditize your complements (2002)

Related Stories