Scott Rafer on Dalton's letter: "Learning the wrong lesson"
rafer.net
rafer.net
Can anyone explain more what he means by that? How can something that is a cash flow hit for a BigCo not be a cash flow hit for anyone else?
If @daltonc's app.net doesn't massively undercut FB's App Center revenue plan (or overwhelm it's distribution plan) somehow, what's the point?
I think your cannibalization concept is brilliant, but I don't think app.net is completely vulnerable to it. Caldwell's made it clear that he's not trying to find success of Twitter and FB's magnitude, that sustainable profitability is more important for this venture than unbounded growth and huge winnings, and he's not trying to steal Twitter's customers, whatever that would mean. But unless Twitter becomes wildly profitable I don't see them giving away ad-free unrestricted API access just to eliminate a small gadfly.
I think your concept has a lot of merit, but I'm not sure it applies in this case.
I mentioned elsewhere that, personal opinion only, of course, that this expectation is the sign of bad mindedness. Think of the logical implications of this point.
And, when I don't like the tone of a meeting, I politely leave. Facebook's offices are physically difficult to enter, not leave.
a) dev builds software on the platform that doesn't disrupt the platforms economics, distribution, or organizational limitations. b) dev builds software on the platform that does disrupt the platforms economics, distribution, or organizational limitations.
You seem to be saying that if he had done a 'proper job' he would have pursued strategy b. But surely that is precisely the case where you can expect to be bullied. It is strategy a. where you develop software that does none of those things, but adds value to the platform by adding value to the users of the platform where you are least likely to be bullied. In these circumstances you are no threat and you are complementing the platform.
Strategy a can only be pursued when you are not dependent on the platform and hence are much harder to bully.
This was great from a user perspective (puts a picture with an email so you recognize that this is 'Bill' who you met at the hackerspace even though his email is cypherdog26@crytomesh.org).
However it was NOT so great for Gmail:
1. Rapportive overwrote the section of gmail where they display ads.
2. It further entrenched Google's social competitors.
There is no way that Google was going to launch a set of Rapportive equivalent features in Gmail. However, Rapportive was beloved by users and exited to LinkedIn.
Facebook and Instagram was a perfect example. Facebook has money, time, and resources to build an Instagram competitor. But they didn't, because they still don't know how to monetize their mobile traffic (which should answer the question, "Why is Facebook's mobile app so awful?") They tried last minute to build Facebook camera, and when that had no traction because Instagram had already captured that market they had to cut their losses and buy it. My guess is Instagram realized that Facebook was basically forced to buy them and made them pay for it ($1B).
So I work on a startup in the news/journalism industry. If we're to follow the argument (which I believe is spot on), our goal is to build something that the New York Times or Washington Post wouldn't dare build, because it would destroy the revenue from their newspaper sales (yes, we have to compete against the likes of the Huffington Post too, but it's easier to illustrate it against the background of business models that haven't changed since the 1950's).
What Steve Jobs always preached was true - if you're not willing to cannibalize yourself, someone else will cannibalize you. As a startup, one strategy is to find those companies not willing to cannibalize themselves and help them out a little bit. But in doing so you run a risk; if you don't generate traction quickly enough and they see that it's inevitable, they'll just build it on their own and leave you hanging out to dry.
Southwest undercut united and American by avoiding business and first class, flying only one model of plane (737), which lowered maintenance costs and allowed easier substitution of planes among flights. They skipped the hub and spoke model and flew out of peripheral airports that were cheaper, and boarded like buses with no reserved seats to shorten the idle time a the gate, etc. no long flights meant no meal service. There was no possible way for United and American to match them on price. They were afraid to adopt any of these changes because it might tarnish their brand. They were too focused on competing with each other.
Dell offered lower costs than their competitors by charging up front for the computer before buying the parts (which were steadily falling in price). The competition were stuck with a channel model where they shipped finished computers to channel partners for stocking in inventory, so they not only had to buy parts far in advance but had to give a cut to channel partners and retailers, making their prices higher and profit lower.
Netscape had a dozen competitors to their browser. Netscspe offered free downloads to anyone, and with a wink asked for a licensing fee. No individuals paid it, but corporations did because they wanted to be in legal compliance. Best of both worlds (free for individuals and paid by corporations, but only after achieving critical mass within the corporation). They crushed their competitors with this model, who were charging license fees to all customers from the beginning, as competitors didnt get the benefit of rogue employee groundswell to achieve critical mass in corporations.
Hilariously, Microsoft in turn crushed Netscspe by goong a step further, making Internet Exploror completely free, no wink needed. Netscspe browser revenue dried up as corporations stopped paying for Netscape because they already had a free license to ie.
This was funny to me, as it led to all sorts of crybaby behavior by Netscape screaming "antitrust!", when all Microsoft did was tweak the strategy used by Netscape themselves to crush the others.
All of these cases are "strategy", which is defined by Michael Porter as the way that you are different from your competitors. Ideally, in a way they cannot copy. Notice how a strategy allows you to avoid competing, because you're playing a different game altogether.
Downloading Netscape back in the day wasn't the same as downloading Chrome or Firefox today. We're talking about 1997-8, when most people had 56k modems at best.
Apple's position with respect to Safari is nothing like Microsoft's position at that time. For starters, Apple doesn't even approach a monopoly in the desktop PC market. And Chrome and Firefox got big in a different era, years after Netscape was buried.
You seem greatly concerned to say that Netscape was "whining" and "screaming" and that it wouldn't "make sense" for Microsoft to have done otherwise. I am only concerned with the accuracy of the account of how IE got popular. If that impedes your Netscape bashing, that is only incidental to me and I don't even understand why you would still have any kind of interest in the subject given how long Netscape has been buried.
I'm sure that strategies exist on any platform. Zynga doesn't seem worried that Facebook will copy its games (even if they do hate paying the platform tax). Many products are pretty safe from Facebook. For example, build an app for dentists, or personal trainers, or pregnant women.
The risk is not very hard to assess. It's pretty binary per startup opportunity. Do the relevant incumbents' revenue models, distribution models, or inherent cultural limitations prevent them from competing with you. @paulsutter covered the first two well early in this discussion. For the last, Apple has great difficulty with social, Facebook hasn't figured out Touch, Google is excessively focused on those two competitors, etc.
Please also see these two comments: http://news.ycombinator.com/item?id=4336783 http://rafer.net/post/28638883246/mark-i-know-for-a-fact-tha...,
http://en.wikipedia.org/wiki/Precommitment
Note that the 'textbook' examples of precommitment involve the burning-of-bridges or burning-of-ships-at-the-shore.
If there's any hope of an alternative to Facebook/Twitter at this point, given the incumbents' long head-starts, it's going to be a 'zag' to the incumbents' advertising-supported 'zig'. Caldwell might not have gone into his Facebook meeting fully intending that sort of confrontation... but by the time he left, the choices were stark.
I thus view Caldwell's anecdote not as a complaint, "boo-hoo I was wronged", but as him sharing a difficult lesson: that Facebook's model makes it impossible for Facebook to be a magnanimous platform steward. Perhaps Caldwell should have seen the writing on the wall earlier, but many exterior-platform shops are missing the same point, as evidenced by the surprise and indignation after each new consolidating move by the platform proprietors.
Perhaps Caldwell could have intuited the reality beforehand, and written about it in abstract terms. The anecdote illustrates the problem, in a way that far more independent developers will understand.
I feel a responsibility to make it clear that such things are the norm and need to be accommodated in business planning and risk assessment from the beginning. Please note that I failed to do so the first time I dealt with Facebook, but that doesn't make it their fault -- rafer.net/post/168541483/lookeryupdate
Please also note, that I'd be damn excited for app.net's current iteration to take off and will go out of my way to use it if I can. I'd love a dev program that had ongoing stewardship built in, but no for-profit platform provider has ever made such a thing work over time.
I agree that it is helpful to point out that there are major risks when you develop for an alleged platform. But isn't that the point of his post?
I think the point of the app.net initiative is to create a platform that because of its different business model (subscription for infrastructure/service) will not suffer the conflicts that are inevitable when the business model is ads. And whilst this is indeed unusual lately, it is not long ago that such for-profit platforms were the norm. Older examples include, DOS, Windows, UNIX, Linux, X-Windows... More recently there is a profusion of companies providing commercial support for open source software or providing such software as a service.
And, I'm shocked at your statement about the older platforms. I don't know if your entire background is on your LinkedIn page, but what you say is simply not true of Windows, DOS, the Mac, and is very tough to credit to Sun or SGI in their heydays. Those vendors cherry picked their ISVs all the time and with no more rhetorical consistency, stewardship, or charity than Dalton describes. They did so because they had market share in their segments and could.
That's why I specifically included this in my post: Playing naive to that reality went out the window in the mid-1980s with PC software and has been reproven by the platforms at least every 24 months since.
Everyone from RedHat to BeOS to Sega who behaved (somewhat) better did so because they needed their ecologies to grow. If they'd succeeded in building market share, I'm pretty sure I could predict their changes in policy.
I gather Zynga has that covered (and has an interesting relationship with Facebook)
I know where he stands about all of this. I don't want to drag him into this fight any more than he already is, but I don't think you understand pmarca very well.
I've got some great supporters and advisors who help me tremendously. It just pains me to watch someone great like that drift further from a startup.
But what does ‘require a new entrant’ mean here? Who requires it? There is surely only one meaningful answer - the consumer. Moving on... What sort of new entrant offering ‘can not be filled by a product line extension from an incumbent?’ I suggest in the context we should parse this to mean ‘... platform company cannot build a competitive offering.’ But of course, there is little that such well financed companies who control the platform can’t build. More to the point there is still less that can’t build if all they have to do is copy an existing product. So this is a very strong condition. Perhaps you mean - software the platform company doesn't want to build? One answer along these lines is to build software for a small segment the platform company cannot be bothered to address. But for an ambitious developer who wants to add significant value to millions of users I am not sure it is possible to develop such an offering - insulated from potential competition by the platform company - unless the platform company has a clear policy vis a vis developers and provides very clear signals with a very clear record to trustworthy behavior that gives a high degree of confidence it is safe to proceed. But of course developer is taking a risk that the platform company can bait and switch. I am unsure what ‘large companies with developer programs all consolidate their segments’ means. But if it means that such platform companies have an inbuilt drive to purchase their devs I can’t agree. The big platforms have so much software on them that even the wealthiest company couldn’t buy it all even if they wanted to. I don’t know what a cannabilistic cash flow is. This may be a term of art I am just unfamiliar with. But if it means a cash flow built on the back of the host platform, this surely refers to the cash flow of all apps built on a platform. So I am unclear this idea moves the ball forward. If on the other hand the idea is that it will cost the host money for no clear ROI then this is indeed always possible, but just as obviously if there is no clear monetization strategy the dev is taking a huge risk. It can work out but of course most times, not surprisingly, it doesn't. (Clay Christen makes much of this dynamic and the challenges of marginal rather than absolute costing when he discusses such cases as Blockbuster competing with Netflix. But the undiscussed part of his argument is the hindsight effect. Blockbuster were right to ignore most of the market entrants. Most never got any traction. With a thousand flowers trying to bloom, which one will see the sunlight?) So building software that doesn't seem to make financial sense will doubtless reduce the risk of the platform company competing, but you'd better have something up your sleeve. The Instagram example is indeed important. But it is important precisely because of the very different behavior of Apple and Facebook vis a vis their devs. Apple has had long experience of working with devs and a deeply wired understanding of how to foster and manage relations with their developers. iOS is a platform. Apple provides some core apps but other than that it is a free range and the overwhelming majority of devs can be confident Apple isn’t going to compete with them. It seems to me that the whole point of DCs argument is that Facebook declared itself a platform, provided APIs, provided assurances that his product was seen as valuable and would be welcomed and then, having given him the confidence to build it, turned on him threatened to cut off his access to the platform on which his app relied and then offered to buy him out. This is very different behavior.
So to be honest, I don’t find the argument very clear. I’m not sure we agree on what the responsibilities of a company that declares itself to be offering a platform are. I can’t extract a convincing strategy here for how you feel devs can best to develop on a platform. And whilst you seem to feel you have some sort of lesson for DC about how to work on a platform even if the company baits and switches, I can't fathom what that lesson is.
If you don't have a clear strategy that protects you from competitors big ans small, don't start a company.
But if Godzilla steps on you, dont act surprised and dont expect a lot of sympathy.
In Zynga's case for instance, FB was pretty darn dependent on the advertising $$ Zynga paid them (among other factors). In Instagram's case, I believe that Facebook wished to avoid pumping up mobile app traffic that quickly. Etc.
Obviously, YMMV. The strength of the answers varies wildly, and startups often lose the bet -- even when the bet was calculated well. C'est la vie, et c'est la guerre.