Hacker News new | past | comments | ask | show | jobs | submit login

Never, ever trust another company for your business model.

Here's a previous comment: https://news.ycombinator.com/item?id=9262245

That was about FaceBook, but exactly the same holds here.


    Certainly I'd be incorporating an exit strategy,
    and my game plan would be to leverage off FB to
    start with, to grow as fast as possible, develop
    my own community and eco-system, and then be able
    to shed FB and replace it with something else.

>Never, ever trust another company for your business model.

This is just another thought-terminating cliche not unlike "if it is free, you are a product" and I can't help but respond.

Large technological companies create ecosystems where smaller players can coexist with "hosts" in a mutually beneficial relationship. Ecosystem companies are not able to predict the future of the ecosystem, its evolution for all intents and purposes is mostly unpredictable. Sometimes the relationship works out, sometimes it doesn't.

Startups have 70-90% failure* rate, the possibility of the host cutting of some parts of the ecosystem to integrate their food chains into the corporate borg is factored in the business and investments strategies of the smaller companies.

You should not trust another company for your business model, but building your tech business in the ecosystem of a large fish is an absolutely valid and quite possibly lucrative strategy.

*definitions vary

EDIT success/failure

The problem is that there is nothing preventing the "ecosystem host" from deciding that a niche an "ecosystem player" is exploring is lucrative and worth pursuing.

You can fly under the radar for a while but if Twitter/Facebook/Google/etc decide your area is lucrative, they will go after it, whether that will kill one of their "ecosystem players" or not.

There is no partnership involved, they provide a service as a huge "blob" that is making them money and they don't care about 1 out of 1 million of their regular users. Your small company cannot call itself a partner, you've as much leverage as any user when it comes down to deciding to sacrifice you in the name of something else more lucrative.

I'd say explore the benefits to bootstrap but have a strategy to depend less and less on them or your time will come. And if you really want to tie your business survival to one of these ecosystems, get yours and theirs responsibility down in writing (and even that won't help much because a breach of contract for them will mean nothing in terms of dollars, while for you it'll be your existence).

Definitely. Indeed, there's a very long history of companies creating ecosystems and then crushing the most successful entrants. One that stuck with me is Stac:


In the MS DOS days they were looking at building on-the-fly compression chips for hard drives. But it turned out that it could be done nearly as well in software, so they made a very popular product that basically doubled the space available on every hard drive.

Microsoft thought that was pretty swell, so they talked with Stac about licensing their technology, getting as far as inspecting their source code. But they decided to just build it themselves. Stac eventually won a lawsuit on patent claims, but it was neither easy nor cheap.

It makes sense from an MBA perspective: you can definitely boost your short-term numbers by fucking over your previous partners. You get a proven market and no significant competition. But what never shows up on the spreadsheets is the extent to which you've reduced long-term ecosystem innovation because savvier players learn you can't be trusted.

Ah. So you mean, "If you want to be perfectly secure for all future time you must be completely self sufficient."

Can't really argue with that. Maybe not useful advice, but it's self-evidently true.

It's more like:

    If you build a great business based on something over
    which you have no control, and which is such that the
    ones who do have control actually may have an interest
    in cloning what you do and terminating your facility
    without warning, be sure to have a "Plan B".
Yes, sounds obvious, but so many people seem to ignore it, and in some cases actively disagree with it. You seem to be disagreeing with it - what would you suggest as more useful advice and a more reasonable stance?

Honestly, would you build a business based on a Twitter or Facebook facility, relying entirely on them not arbitrarily terminating your access?

What is your advice?

Perhaps to realise these companies are walled garden abberations, and that in an open competitive world, the Eco system would exist but the monopoly would not.

As such encourage and build openness by default into your products, as for sure users want that (how many people for example wanted only to email other AOL users, or view only other AOL websites?

It's not a thought terminating cliche at all, it is very sound business advice and you will ignore it at your peril.

Not all that long ago I did tech dd on a company that had exactly one very large company as their main unwilling partner. They claimed that they were 'too big to fail', guess how that ended?

Yeah, Gnip founders ignored this "very sound business advice" and raked in ~$130M.

Startups are risky, startups that are growing in the ecosystem of one large megacorp are even riskier but they are prime acquisition targets. You decrease success chance for higher yields. This is like business 101.

The advice "do not base your business model on another company" evaluates to "do not increase your risks" which would be an absolutely absurd advice.

If one qualifies the original statement with several conditions for different contexts, it may become a solid advice. As it stands this model of understanding business strategies is useless.

Super for gnip.

But basing your business model on being allowed access to other companies' data fails far more frequently than that it succeeds, and it fails far more frequently than the already pretty bad chances that you have as a start-up to begin with.

See the thread subject that you're commenting in for how Gnip's competitor fared.

Yeah sure, the countless resellers of MSFT, Oracle, IBM, VMWare, Redhat, EMC all do not understand business at all. If only some of the uptards knew how little they know, yet how much they run their mouth.

I think you're mis-understanding the issue, but that's fine.

Let me try to explain it a bit better: if you use documented interfaces and have an established supplier role with a larger company that's fine. But if you are taking data from some API and you're re-purposing that data in a way that the company supplying you with that data may not agree with then you may one day find your access terminated.

You probably should not include RedHat in a list like that.

But DataSift was actually buying the firehose from Twitter, and they got screwed anyway.

Which only re-inforces the statement, definitely does not detract from it. But normally speaking you write your contracts in such a way that one party can not unilaterally withdraw without serious penalties.

AFAIK the contract simply expired (I think it was 5-year old) and Twitter refused to renew it.

So, "Never, ever trust another company for your business model" is in fact a thought terminating cliche?

People will take any opportunity to stop thinking:

    Most people would sooner die than
    think; in fact, they do so.
        -- Bertrand Russell
In short, if it's a "thought-terminating cliche" then the problem may not be in the way it's expressed, but in the way it's interpreted.

If someone chooses to terminate their thoughts based on something someone says, without considering why they might be saying it, and why there may be value to be extracted from it, then maybe the problem is not with the expression, but with them.

Think of it this way. You may consider it a given that there are times and places where you can and should trust other businesses. This statement about never, ever trusting a another business therefore seems to be at odds with your experience. Do you ignore it, or do you use it as an opportunity to learn why someone might say it?

You're stuck in a loop.

1) Make an overgenaratilzation; 2) Get called out; 3) Make a defense, only to expose your overgeneralization yourself; 4) Get called out for it; 5) ???? 6) You're stuck in a loop.

Let's get back to an earlier question to help me understand your position. The original is:

    Never, ever trust another company
    for your business model.
In this comment[0] in response to a scenario wherein one company did trust another and got screwed you said:

    What's wrong is the lack
    of contract or agreement.
If a contract or agreement is required, then that would indicate that one should, indeed, not trust another company for one's business model. So it seems that you agree with the actual statement.

Is that the case?

[0] https://news.ycombinator.com/item?id=9363982

Those resellers are not trusting those other businesses, they have in place legally enforceable agreements. There's a difference.

I honestly don't understand the point you're trying to make, so let me parody your stance and you can explain where I'm wrong.

* I create and provide a service

* Said service has an interface

* You spot a service you can provide based on my interface

* You build a business using that interface

* You are successful

* I notice you're successful and build a clone

* I launch my clone and make the interface unavailable to you

* Lots of your customers switch to using my replacement

* You go out of business.

Is that all OK in your opinion? If not, what's wrong?

What's wrong is the lack of contract or agreement. If you jump on a business opportunity without formal contracts, it's your own fault. Generally speaking, if defense and risk mitigation is not your first thought when evaluating opportunities, you probably should stick to applying your technical skills, working for people who understand those things.

But of course, MBAs are for suckers who know nothing and should just cease to exist. At least that's what I keep hearing on HN.

So, what I hear you saying is this:

    Never, ever *trust* a company: always have
    formal contracts.  If you can't get formal
    contracts, don't work with them or rely on
That seems to match what I said, which was:

    Never, ever trust another company
    for your business model.
You seem to be saying exactly the same thing. You seem to be saying: Put formal contracts in place. If you trust someone then you don't need formal contracts. If you need formal contracts, you don't trust them.

Am I missing something?

It really does seem to me that we have pretty much exactly the same viewpoint - there is a risk analysis to be done, and risk mitigation is critical. My risk analysis says that if someone else is in a position to screw you over completely, don't proceed on blind faith. Instead, have contracts or agreements in place. In other words:

    Don't trust them.
We're back to what I originally said. Do you disagree with it? It seems like you're actually in complete agreement.

If you read carefully the discussion, you will see that my point is not that what you said is untrue. My point is that what you said is a "thought-terminating cliche". It's a over-simplified one-liner. Just look at how much had to be written in order for it to be more clear and valid. There is no trust in business relationships. How do you trust a legal entity? The people there change all the time, goals become misaligned, etc.

And if you look at the comment wherein I used that single line, I then went on to add this:

    Certainly I'd be incorporating an exit strategy,
    and my game plan would be to leverage off FB to
    start with, to grow as fast as possible, develop
    my own community and eco-system, and then be able
    to shed FB and replace it with something else.
Is that not enough of an expansion of the thought, pointing out that using FB (or whoever) is valid to start with provided you have some sort of exit strategy?

I disagree that it's a "thought-terminating cliche" - it certainly shouldn't be. There's a difference between trusting someone, versus using them to your advantage but having an exit strategy.

I agree entirely that there is value, sometimes significant, in exploiting the ecosystem created by a larger fish. You'll see that in my comment, wherein it says:

    ... leverage off XXXX to
    start with ... be able
    to shed [it] and replace
    it with something else.
You are saying the same thing I've said, but emphasizing that starting in the jaws of a larger fish is sometimes worth doing, and sometimes they'll deign to let you survive there.

Consider, though. If you become successful - which is surely your aim - then you will come to their attention, and they will do whatever they like.

So I repeat: Don't trust them. Use them, yes, but don't trust them.

I think this needs more detail as to what stage in the tech lifecycle the businesses are in.

Obviously, I'm going to trust an electricity company to provide my business with electricity. I'm even going to trust laptop manufacturers to continue to provide laptops I can code on.

In both those cases, the thing I'm depending on is a commodity, where there is also either strong competition or regulation. There isn't much risk in depending on them.

The stage to be wary of is when a whole industry is still at product stage and people are trying to move them into being platforms, or using them to Innovate/Leverage/Commoditise (ILC, see Simon Wardley) to play ecosystem games.

i.e. Where the business models and how the industry will be vertically and horizontally sliced are not yet settled.

Those are the ones you should be wary of dependence on.

Perhaps the pitfall is in using the term "platform" to describe something that is actually just a very popular (and network-effected) product.

Sometimes trust is essential and diversifying would be a loosing strategy as the larger partner will react specifically to your diversification but will help develop your business until you are not looking for something on the side. Generic business strategy rules don't work anymore, the conditions are changing too fast.

I agree, however, that blind trust is always a bad path.

Allowing your business model be beholden to the whims of another company is setting yourself up for failure. Sure using an ecosystem built by another company can be a great way to gain traction but by limiting yourself to this ecosystem you are also severely limiting your growth potential. Not only that but you will be easily and quickly dispersed with when you become successful enough. Having validated your idea the larger eco-system company will simply replace you with a clone that they own. Unless you have some serious in-house advantage (patents and people) that is not easily replicated then you will fail.

There's always risk but it varies greatly by platform. Twitter is by now notorious for doing this behavior. On the other hand Apple has a flourishing ecosystem of app developers. And FB has improved greatly moving from being Twitter like to closer to Apple.

It's a function of maturity (young startups have growing pains on what they want to be and how they monetize) which can definitely create collateral damage. Mature companies like Apple don't experience the same degree of angsnt.

It's no accident that Twitter, which has generally underperformed and disappointed more than its peers, is still screwing up and over its ecosystem far more than others.

Or be acquired by the larger ecosystem company what would be the desired endgame for many startups. This alone may be worth all the risk of being squeezed out of the market.

The example is obvious: Gnip vs Datasift.

As I mentioned unless you have some compelling reason to be bought (people or patents) you don't stand a chance.

> Startups have 70-90% success

Don't you mean failure rate?

> You should not trust another company for your business model, but building your tech business in the ecosystem of a large fish is an absolutely valid and quite possibly lucrative strategy

Yes, as long as if you have a Plan B if/when your relationship is cut (a valid Plan B is to close shop and try to do something else with the expertise gained)

>Don't you mean failure rate?

Sure, thx

>Plan B

Certainly. Datasift has 25 various information sources. Twitter is really important, but doesn't seem to be critical for their platform.

It is if it's 90% of their business by revenue.

I've worked for small companies in the past that had dozens of partners, but most partnerships don't really do anything except give you brand logo to mutually share on each other's web pages.

It's like anything else, partners provide revenue to you following a power-law distribution, with a long-tail that provides nothing and a very select few that make up the lion's share.

Why not? Remora survive by not getting too close to the shark's mouth. Same applies to business.

Businesses are all mouth.

Well 50% mouth and 50% anus, sometimes its hard to tell which half you're talking to though.

I worry media companies are effectively in this position with regard to social.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact