Hacker News new | past | comments | ask | show | jobs | submit login
The only two ways to build a $100m business (versiononeventures.com)
124 points by bwertz on Sept 19, 2012 | hide | past | web | favorite | 70 comments

> Too often I have seen entrepreneurs believe that customers will automatically flock to their cool new service, completely underestimating how tough it is to cut through the noise and build an audience.

I usually think the opposite. Most of the time when I come up with an idea for a start-up I think "How the hell am I going to get users on this thing?"

My guess is that you've had enough failures to figure this out. It took me a couple.

I see this 'the product will see itself' thinking often with engineering peeps. They are usually very logical people that evaluate their decisions this way but the market can be very different in their purchasing triggers. Good distribution and brand recognition can sell even sub par products. There is much to be said for getting these parts of a business running well.

How do you go about to answer that question?

By making sure you're building something that is intuitive and that people are interested in.

I could be missing something but I find this article a little ridiculous and devoid of any useful information.

I work for Atlassian. It's a $100M business. It wouldn't fall into either category. We've gotten there by selling a lot of software over the internet. We spend hardly anything on customer acquisition. There is no viral effect to our products.

Another successful business I can think of off the top of my head is ZenDesk. They have to be approaching $100M. Their main sales channel is the internet (low customer acquisition). They have zero viral effect. You either have X support staff or you don't therefore you will need an X-user ZenDesk license.

There is no viral effect to our products.

You certainly know your market better than I do, but it seems to me that you are taking advantage of a viral effect. As people use Jira (for instance) and they move around from one company to another, they're likely to recommend using Jira at their new places when they get sick of Rational, Remedy, Bugzilla, etc.

People using Jira and realizing that there is a bug tracker that doesn't stink are very likely to bring that message to their next company. The word about a great product spreads... like a virus, albeit a slower moving one than pet rocks.

I think the most polite way to put it, is that the article is written with a certain audience in mind. You and I are not that audience.

Definitely true, the last three projects I've ran have utilised your products (Confluence & Jira specifically) and its through word of mouth only.

I should just mention that there’s yet another way to build a $100M company, but it involves starting with a $1B company.

the original line, from Warren Buffet (probably a misattribution), is:

"How do you become a millionaire? Make a billion dollars and then buy an airline"

I'm pretty sure that was Richard Branson.

Yes, it was specifically about Virgin Atlantic, which at one point stretched his finances to the point where he was forced to sell Virgin Records to free up the cash to keep it afloat.

Which is sad, because he won a lawsuit the next year for the exact amount he sold that company for.

I think to try to attribute this one at all would be risky. It's been applied to so many different industries (space, motor racing, wineries, farming, theater, video production business, trading, etc.), and said by countless famous people.

Does anyone know if OED does cliches?

I once heard “How does one leave Las Vegas with a small fortune? Go there with a large one.”

Argh! Beat me to it. I was gonna say "start with a $200M company and invest in Facebook IPO."

It's like the wine business: the key to making a small fortune is starting with a big one.

So $100M is the new $50M? I used to hear VCs talk about "We don't want to invest in something that isn't going to be at least a $50M business."

The other question is this, "Valuation or revenue?" If I can build a business with $30M/yr in revenue and 16% net income I shouldn't bother talking to a VC? I don't know, seems like you leave too much money on the table with this world view of $100M or nothing.

If so its a side effect of two things. One, is fund sizes have gonve from $100-200m to $500M+. So, the partners need to put more money to work. Second, early stage valuation inflation. $10-$20MM cap notes. And all that.

"Generally speaking, there are two ways (and only two ways) to scale a business to hit that $100 million threshold."

Actually, there's another way to hit >= $100M.

Be a strategic acquisition target. (Route two: The viral effect can overlap with this.)

I forget who blogged this (Dixon?), but the idea is that that to hit a >= $100M valuation, you should not be thinking of the standalone value of your business, but instead the value it provides to potential acquirers.

Sitting on crucial patents is one approach, but extraordinarily tricky.

More importantly, if your existence presents a strategic challenge to a larger player (e.g. Instagram), then your business is worth a lot of money.

Best rationale for why Twitter is @ $6-$7B.

-- They have the potenital to destroy N-times more value than they can create.

I'm more inclined to build a stable business that makes me enough passive income to continue to work on it. Building a $100M company sounds exhausting. 37signals has definitely influenced me in going in the direction of small and lean instead of big and greedy. However, I find the insight into the way VCs think fascinating.

I know this is HN and we can assume certain things about the businesses being profile but I wish that the headlines reflect the narrow niche the articles talk about. There are certainly other ways to grow a business outside of the software/online space.

yeah, but we are talking tech / software on HN

Not all tech ventures rely on bits alone. There are a fair number of tech products that are physical, too.

In consumer space sure, but not in B2B or enterprise.

B2B/enterprise would likely be covered in his first point about having a high LTV so you can invest in user acquisition.

I'm surprised he doesn't mention TAM (total addressable market) at all. A high LTV doesn't matter if you're just selling things to male cat breeders in the Northwest between the ages of 27 and 33.

For the high LTV option, can one put a specific number on this (i.e. $1,000)? Or is it simply having the ratio of LTV to Acquisition Cost is greater than X?

a large addressable market is always a pre-requisite for building a big business

Are these two ways really that different?

Essentially all the article states is that a profitable business needs to have profits per user that scale according the cost per user acquisition. Viral is cheap acquisition, so profits can be less. Non-viral costs more per user and typically has a smaller market, so profits per user need to scale accordingly.

Completely different.

To put the article another way, there are two ways to make big money:

* Make lots of money per user

* Have lots of users and have low expenses.

No-man's land is having a £1 app that you will never realistically get more than 5,000 users on.

Does method 1 even exist for consumers? Small businesses, sure. But apart from a few gold mines like TurboTax its hard to get consumers to spend any money on SaaS.

Dropbox, Flickr, Apple, etc-- all sell to consumers with a high ARPU. Also, cell providers, insurance companies, car companies, etc.

Good rule of thumb-- anyone who advertises heavily to consumers has a high ARPU.

Dropbox is probably the best example for an app that scaled charging consumers

Sure it exists. Have a services that is free to use but includes charges for one particular highly desirable function. One example is a game with in-app purchasing of new levels, character wardrobe, etc.

evernote, dropbox, rescuetime, pandora, spotify all beg to differ

yes, you find this quite a bit in e-commerce but agree on SaaS

Wow, this doesn't cover models that have worked for companies like Microsoft or Apple. Were they viral, or did they depend on big LTV? Not completely. LTV helps (people buying iPhone again and again), but they are not subscription-based companies. I think this model falls short to describe business.

MSFT and AAPL both benefit from immense network effects, which translate into high LTVs (Windows/Office platform, iOS ecosystem). The presence of a strong network effect can often be a strong driver of virality, if not a prerequisite for viral acquisition for a lot of businesses because the incentive to share is directly correlated with the value delivered to the user/customer.

...and even more impressive, they did it before the terms 'LTV', 'network effect' and 'virality' had even been coined.

As he mentions in passing, marketplaces seem to be a good solution to both of these options. Once you reach a critical mass you potentially have a large viral effect, user lock-in, and a high lifetime value as users get used to purchasing on a regular basis.

The only trick is conquering that chicken-and-egg issue.

agreed - marketplaces are probably the most attractive business models once you get through the initial traction challenge

Or number 3, leave the boring B2C world and go enterprise

That just means the customer acquisition cost is even higher. You're not paying for Google ads, you're you paying salespeople to go out, wine and dine, and get new customers. As a result, the margins on products needs to be higher to support the sales process.

I'm biased, but I don't think this is as big a problem as you suggest. Yes, customer acquisition is more expensive and takes much longer... but each customer is spending orders of magnitude more than in a typical B2C sale.

And I think there could be considerable cost savings in reduced overhead from having to support fewer customers at a higher price point.

This used to be a lot more true, but it's hard to apply it as a blanket statement anymore. Some counterexamples come to mind:

- 37 Signals

(because you're going to object that they're the only ones, here are some more)

- Fog Creek

- Atlassian

- Rackspace

- SoftLayer

- Amazon EC2

- Balsamiq

- Heroku

- ZenDesk

- Visual Website Optimizer

To my knowledge, none of them have travelling salespeople. Acquisition cost is likely higher than with consumers, but likely nothing like the costs associated with an outside sales force (or their prices would have to be higher).

Everyone says 37 Signals is an exception, but I think that's just lazy thinking. There are lots of enterprise businesses that don't do expensive outside sales.

There are different risks selling to businesses, but B2B businesses don't need massive scale to be able to pay their rent. So if you're bootstrapping, B2B is likely to have a higher probability of non-failure.

Vast majority of the companies above make products for developers/designers, and that is not a coincidence. Developers/designers live on the web, seek and find solutions themselves. Rest of the enterprise market is quite different and much harder to reach, hence drastically higher customer acquisition costs.

I posted products I know, and I happen to be a developer. If you google (space) SaaS, you'll find bootstrapped companies serving many other parts of the enterprise.

Agreed. I am working on one :) I just meant that if you are addressing developer market, sales costs are lower as developers live on the web and there are many mediums to share information such as hacker news.

I find that other parts of the enterprise market are surprisingly harder to reach. In enterprise companies information sharing is often discouraged, many folks working in the enterprise are hesitant to mention what products they use let alone promoting (or criticizing) them. Successful SaaS companies addressing the enterprise seem to either gain traction elsewhere and then try to penetrate enterprise or spend tons of money for marketing and sales.

Plenty of those do have sales people.

In the end why leave massive wads of money on the table, do both.

Yes. I've had a few conversations with SoftLayer's sales people. I am not sure if they travel, but they do pick up the phone.

Not to mention the very long and unpredictable sales cycles. The money is good, but it's a different set of problems as you said.

Exactly. The B2B markets are typically harder to penetrate for startups, unless the startup has plenty of cash to drag businesses to their product.

Go enterprise and don't get lured to the "enterprise for 1.99" business model, propagated by a few exceptions that succeeded (e.g. 37signals). It's a lot "easier" to scale charging proper money for B2B products than betting on scale...

I wouldn't say 37S were "enterprise" , they started at the low end of the business market. I'm guessing a lot of their initial customers were freelance rails developers.

They do have enterprise level customers but that's more of a "trickle up" thing.

You have to hire out execs from previous enterprise companies, or work at one and get to know people in the company (as well as their problems) that you solve with your technology.

not necessarily. Cold calling works rather well, in most products. The sales process is definitely long (several months), but it does work... (my first company was a B2B for companies with 50+ employees and we didn't know anyone in the industry)

That said, knowing someone definitely helps. A LOT.

Wouldn't that just be #1?

Anything that claims there are only X ways of doing something will always have somebody come along and +1 it with a new approach.

I'd also say that anybody setting out to make a 100m company is somebody who has already made a sucessful company already. But software is the new music thesedays as how whats the most a single application has sold for compared to what the best single song has made and I'm thinking Instagram even compared to tbe beatles on a app compared to one of there great songs will still win.

I will also correct the title for programmers and say: "There are only 10 ways to build a $100m business, FF if you include exceptions"

"Error establishing a database connection"

Refresh worked for me.

Let me summarize. The 2 methods.

1. Have customers who come back frequently and spend money. 2. Have a lot of new customers who spend money. They tell there friends to come, and spend money. Rinse. Repeat.

Another brief takeaway for me was that businesses that grow quickly either

1. see an extremely reliable 4x return on advertising expenditure


2. do not have advertising expenditures at all.

One caveat to the article is that it is not in every business's long term interest to grow quickly.

Interesting article and insights. I'm wondering if there's a hybrid model where you can enjoy both - high LTV AND high viral co-efficient. Or if that's unrealistic. Maybe Ebay?

Amazon probably qualifies. eBay, sure. iTunes (sort of; arguably, the hardware drove everyone to iTunes originally, and the hardware was not cheap to produce or bring to market; but once the install base became stable, it grew VERY quickly through word of mouth).

Also: a high viral coefficient does not necessary = social, even though that's what people assume these days. I think the world has seen way too many businesses in recent years that are "social" for no other reason than that social is supposed to have magical network effects. The reality is that not every product, experience, or piece of content is truly social by nature. Trying to tack social onto a use case that doesn't support it is not going to yield a viral coefficient.

what about having a 5 million/yr revenue run rate. With "valuation" multiples of 20s, you can get to a value of $100 million.

In my experience of being part of the acquiree, you can get acquired for a gross earnings multiple of between 30 and 40. So as long you're only spending 2.5m per year to get your 5 mil run rate then yep, find the right acquirer and you can get that 100m valuation but there's a lot luck involved.

Valuation multiples are not against revenue, they're against earnings.

more broadly, i've often said that there are two ways to build a massively successful business (e.g. $100m business, or one worth a billion or billions):

1. don't fail.

2. don't fail to massively succeed.

bootstrapping is all about completing mission number one. not going bankrupt. not falling apart. not having a bubble pop from under you when you need it for your business plan.

but people all too often forget about #2, and think they will just magically be a $100m company. It doesn't work like that. If you want to be a $100m company, do two things: don't fail; don't fail to be a $100m company.

this is genuine advice and I can give more details if anyone is unclear about anything.

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