Trying to get big.

The conventional wisdom is that it works better to get big with an unsustainable revenue model and convince people you can make it sustainable than to stay small with a sustainable revenue model and convince people you can make it big.

Spend whatever you can to capture the land, then charge rent.

Sometimes it turns out you can't hold the territory. Sometimes it turns out it's not worth much. But the times it works out are the times that make funds.

I hate this conventional wisdom idea and it really doesn't need to be that way in many cases.

Companies like Mailchimp and 37 Signals prove that the "logical" way to do it works pretty well.

(Not a criticism of your comment, just the get big fast idea)

I'm inclined to agree with you. It's not my field but I suspect VC pressure has something to do with it: they need their 10x returns, or their unicorns, to make up for all the losses elsewhere and "get big fast" is the most reliable mechanism for doing that. Not suggesting I'm a fan of it, btw.

It's a market issue. Databases are a very competitive business with lots of options and competitors. It is also extremely hard to sell.

Mailchimp is rather large, but 2017 I have the feeling 37Signals is rather insignificant, e.g. compared to Slack.

FWIW when I was still a student and not in a 9-5 I wondered what if would feel like to use basecamp. Now here I am happily going with flow using slack on the fre eplan. One of my coworkers sometimes asks if we shouldn't go with a subscription to have the whole logs but we rarely have to dive so far in the past at all.

> Spend whatever you can to capture the land, then charge rent.

Spot on. And this rent-seeking behavior is all about becoming quasi-monopolist in a niche and it's anticompetitive and anti-innovation.

Real R&D is done only to capture land, then it is replaced by PR and all efforts go into holding the trench.

What is the rent here, enterprise licensing? Quasi-monopolist? Does MongoDB have anywhere near a majority of market share for database vendor?

In technology, it's often the cost of re-integrating with something else. MongoDB makes this lock in ever better because it's API isn't SQL, so getting off is even harder than usual.

It's pretty easy to use the Mongo protocol to talk to Postgres, and Postgres JSONB beats Mongo BSON hands down

Cue gratuitous protocol changes from MongoDB.

If it's so obviously a bad idea, why do investors keep buying into it?

I don't really understand your question... this is the favored approach of many Silicon Valley VC firms, and they are widely considered some the smartest tech start-up investors in the world.

If it really is a bad idea, and those firms have been so successful for so long not because of it but in spite of it, then it seems like a huge opportunity for others to come along and displace them with a better model. There's a lot of money in many of these tech markets. But we don't see that happening.

I guess the answer is that it's unintuitive... it seems like an obviously bad idea to many people, but the people that have the most experience think it's a great idea and continue to succeed with it.

There are other approaches: in China, for example, investors generally considered profitability much more important and expect it very rapidly, like in the first 6-18 months (though this has eased quite a bit towards a more SV-style model in recent years).

The gripe then is that companies can't have any long term vision or go big because they have to be making profit immediately and constantly.

Something about green grass...

I remember reading that they are loosing money on the aggregate. While there are some successful investors, majority is not long term successful.

Yeah, I when I thought about it I realized when I say "SV VC" I mean probably only the top 10-20 firms or so - the "top tier" or whatever, Sequoia, Accel, KPCB, Benchmark, etc. (highly subjective but you get the idea) - and I think those guys tend to do OK, and it has been mostly the same ones my whole time in tech (~10 years), and I know many go back considerably before that.

I meant that a lot of those people are down with the land-grab now / rent later model.

It's a bad idea most of the time but sometimes it works. The smart investors have a portfolio of many bets, and it's priced in that most won't succeed. For the entrpreneur I think this is a bad idea but for the investors it's workable if they can do it a few dozen times over.

It's just injection of both money and risk, amplifying the result. It'll fail more often than usual. But if it succeeds, wow.

It's a bad idea for the small business owner/startup to pursue, because investors buy into it for the 1 in how-ever-many that get big. 1 facebook is worth many failures. And to be fair, they're not just throwing money at the wall and hoping it sticks - though there is some element of that. There's also the many companies that don't make headlines but are generally - if modestly - successful as well. So, the investors don't want to lose money of course, but they're okay weathering several losses if the occasional win is big enough to more than compensate.

On the flip side, if you're a small business trying to grow your company, this hole "carve out the market operating a loss in hopes that you'll make it big" idea doesn't work in your favor most of the time, statistically speaking.

Modest investment for modest gains for modest success for long term growth is the smarter bet for the business owner (usually), but that's not as flashy, headline worthy, or profitable for the investors as the go-big or go-home approach.

So, it depends on what perspective you're taking as to whether this "conventional wisdom" really is good or bad.

> The conventional wisdom is that it works better to get big with an unsustainable revenue model and convince people you can make it sustainable than to stay small with a sustainable revenue model and convince people you can make it big.

This used to be called fraud, but now calling it out as such is thought of as regressive and anti-entrepreneurial.

I'm sorry if my wording was confusing; the company in these both situations is convincing people (investors) that they can become those things in the future, not that they are at the current time when they're not (which would likely involve some sort of deception).

It's like:

Case 1: "We are spending like mad to capture what we believe is a valuable market as fast as possible. This means we are losing money like crazy right now, but it is working and we are growing fantastically. Once are in a large dominant position, we believe we will be able to retain that market, so we won't have to spend as much on growth, and we can focus our energy on maximizing revenue and being capital efficient and start making a lot of profit."

Case 2: "We have validated our product market fit and business model, and are profitable. Our market share, growth and revenue numbers are small but stable and positive. We believe we will be able to rapidly accelerate our growth and market share in the future while maintaining our margin and start making a lot of profit."

Both are good, neither are fraud, but (1) currently tends to be the favored approach for venture financing, and one way or another it tends to be venture financed companies that end up dominating the important tech markets (though that doesn't necessarily means it's casual).

No it's called in German risk capital (Risikokapital), because you know the risks.

Last year (through Jan '17) spent (very simplified):

    $29.9M on Cost of Revenue (hosting fees + the people who run things + support/consultants)
    $78.6M on Sales & Marketing (sales people)
    $51.7M on Research & Development (programmers)
    $27.1M on General & Administrative (managers/executives/HR)

This was spread across ~820 (as of July) employees. Hosting + Rent + Employees = $188M. Basically they spend $229k per employee which works out to very roughly: ~43% Tech, ~42% Sales, ~15% Management

By comparison, Cloudera is quite similar: 42% on Tech, 45% on Sales, 13% on Management. They made $261M last year and spent $448M, losing -$187M across 1,470 employees ($305k per employee)

A small correction, G&A is everything which cannot be directly tied to per unit of sale. So if they have say employee(s) who work on installing and maintaining an active directory software - used by everyone in the company, those salaries will also be part of G&A.

820 employees, plus some contractors probably. That's a pretty large headcount for that revenue - ~120k revenue / employee...

Brand awareness. Getting exposure at hackathons around the world, etc. costs a lot of money.

Been there, done that.

Sales&Marketing. They need to drive Revenue growth to sell a story to investors.

They launched Atlas, a cloud service. It sure ain't cheap to operate.

Sales and marketing?