Hacker News new | comments | show | ask | jobs | submit login
How to Figure Out Your Competitors’ Revenues in About 70 Seconds (2013) (quora.com)
175 points by pud on Sept 21, 2014 | hide | past | web | favorite | 50 comments

Moz has 140 employees and is at a revenue run-rate of ~$33mm, putting us around $235K/employee. I know companies with higher revenue/employee than us (many of those are smaller) and some with lower revenue/employee (many of those have lots of funding and are going negative to grow faster). So long as you're applying additional metrics into the equation (funding, growth rate, cost/person by location/type, etc) I think this is a reasonable baseline.

That said, there's lots of very important metrics to understand if you're trying to do a comparison - rate of growth, margins, churn, etc. - all of these figure prominently into the health of a SaaS business.

The whole benefit of this method is that its simple. Its first order of magnitude estimation by taking the fact that headcount is usually close to linear with revenue and then applying maybe the next biggest twiddle. So you get maybe 80% accuracy recognizing the linear trend, and then another correction for the valuation.

As such, even 1 correction is kind of twiddly, since it does pretty good with average salary. But, its also in that range where you can do it off the top of your head in conversation, by simply referencing a well known labor number, and a +/- factor that most entrepreneur folks you're having drinks with will know.

Alternative method for high-profile startups: multiply number of employees by 0

Those are great numbers........

Too bad they can't be even close to used to predict the revenue of my SAS apps (so far....). Then we have the train of weekly "Sorry users, we are shutting down" posts here on HN (unlikely to happen if they are pulling in ~200K per employee no?).

There is a lot of variation. I view this method with more than a little skepticism.

Maybe when companies are successful they hire employees till the money's mostly gone. So it'd kinda work for profitable companies but not most startups.

Jason mentions in his article that this applies after the first few million of ARR.

Or you could just look it up, most people don't this info is available from a variety of places for established companies (such as Hoovers and Dun & Bradstreet).

Search them for free here: http://nuggety.com/u/nuggety/company-or-business-search

Hoovers and D&B numbers are incredibly inaccurate for new and privately held companies. I've seen them show $20m/year companies as $1m/year and vice-versa. They don't actually know, they estimate or allow companies to self-register, along with crawling public filings and the like.

Also, btw, I tried nuggety just now and got prompted to sign up for D&B for every search on their site.

The OP method is much simpler and at least as accurate.

Or you could use privco. They seem to be good. Someone in another thread mentioned MOZ. And their stated numbers are similar to what Privco quoted.

Walmart - $480 Billion revenue, 2.2 million employees, $218,000 per employee per year. So that is a company dealing in physical goods, and it kind of fits.

Which is kind of funny, because the average pay to an engineer in a tech startup is nothing close to the average cashier at Walmart.

The average marginal revenue is vastly different too.

Also the cost structure is nothing close to each other.

Or you could just get their annual accounts from the business registry, and see what they've reported there.

[won't be up-to-the-minute accurate, of course, and won't work for all jurisdictions. But it's a data source that far too few people use, IME]

No such requirement in the USA (at least not for private companies, some public companies may report a few accounts with high business concentration).

All companies in the UK must file annual accounts. They are available to anyone willing to pay the one pound fee per set of accounts.

Smaller companies don't need to publish P&L and cash flow. They can just publish a summary balance sheet. However, if you have the summary balance sheet for two consecutive years, then you can calculate the profit or loss.

About 60% of companies file accounts electronically using XBRL/iXBRL (rather than paper), and these accounts are available for free via daily/monthly zip files published on the Companies House web site.

> However, if you have the summary balance sheet for two consecutive years, then you can calculate the profit or loss.

But what if the profit is all paid out as dividends? Then there might be no significant change to the balance sheet, I suspect

You are correct :(

Misunderstood accounts to mean customer lists... Got it, that's not too bad if its limited to financials. I guess the intent is transparency for trade partners/employees/etc. Thanks for clarification.

In the EU, you have to file annual financial report and sometimes consolidated reports. You may be required (or want to) to file other documents, depending on the particular circumstances. All those are freely available for an online check. All you need to know is the company's EIN - employer ID number.

Applies to my company (GrantTree), huh.

However I know that one of our competitors had 7 employees in their first year and, because they mistakenly published their full accounts, I also know they made just £80k of revenue that year.

So it doesn't necessarily apply to all companies.

Note that writeup caveats that the rule of thumb generally works for SaaS companies with $1mm+ in annual recurring revenue (ARR).

Insofar as this works, can anyone explain why it might or might not work? Or perhaps why it must be coincidental.

If I squint, I can imagine it being related to a few things. Average salary at a SASS startup is roughly the same across startups. Marginal revenue is very high. And then there's something about keeping revenue and costs pretty close even though the entity is in 'start-up mode'. Or perhaps that's more a function of average round size and average targets for making the round last.

Still seems weird that this would come close. Or maybe it's bogus...

I think the insight is that if a company has lots of revenue per employee (headcount correlates pretty well with overall operating costs) they'll take that as a signal that they're doing well and can afford to expand. And then they'll do this until they no longer feel like they're in such a secure financial position.

This, by the way, is why the multiplier is lower for well-funded companies - if they have more runway in the bank, they'll feel better about a lower level of revenue per employee, since they don't have to operate at a profit in the short term.

Let's say to were making $50,000 per employee in revenue and paying each employee $80,000. You'd be losing money on each hire (and therefore would begin firing people). If you were able to make an additional $100,000 per new hire, you'd continue to hire until you cannot make marginal gains. It's no surprise that the number ends up being 2-3x salary / employee. That's what would be most profitable.

I want to see Valve on that chart, they should be way up on top, maybe around 5 million per employee or so?

I think probably closer to $10m, but... yeah. A lot.

How could a SAAS company survive long term if they are making less than 200k per employee per year? The fully loaded cost of an engineer which I assume are the majority of their employees has to be close to 200k or more after you include desk space, taxes etc...

I think the assumption is that the company is in growth mode, spending all their revenues on staff to build the company, generating zero profit. The ~$200K sum is approximately the loaded cost per employee.

Well-funded companies have extra money to spend so they'll have more people than their revenues would indicate, which is simulated by dividing their revenues by a lower cost per employee. (Conversely for poorly funded companies...)

Not all employees are software engineers. Even inside the engineering department QA people aren't earning that much. Plus there are roles like marketing, account managers, analysts, IT, HR, etc that don't pay that much (at least compared to software engineers).

People have made various arguments for why this is arbitrary and may not apply.

A reason why it may: VC advise start-ups on what their run-rates, burn-rates, and staffing levels should be. VC operate on standard formulae.

One result is that you'll see a fairly strong pattern.

Or at least that's my hypothesis.

Phenomena and relationships such as this are why there may be relationships between data patterns, though causality may not be on the basis that's first apparent. And why there may be strong autocorrelation between variables.

I am not so good in English and business language: I am wondering, that companies which are "well funded" have lower revenue per employee as companies that are modestly funded. Seems to me contradictory at first sight .... (unless "well funded does mean funds from investors and well funded companies are still startups ... but still does not seem to fit in, because startups are more likely burning money and the examples wont fit).

Can somebody explain?

You basically have the right idea, well funded in this context means that the company has drawn a lot of investment capital. Companies without investor capital have to generate enough revenue per employee to actually pay for the cost of doing business.

Start ups which are growing quickly can operate at a loss, supported by venture capital, and thus may have lower revenue per employee.

If you're well funded you're not worrying so much about revenues at the moment, while if you're not so well funded you depend on them

Hence the "value per employee" has to be higher.

That does not work with companies dealing in physical Goods ... At all

This is for SaaS companies as described in the article, title maybe needs this detail

What would you use for a physical goods company?

Like others have said this is for SaaS startups with more than 1mm ARR. It won't work for your local Kwik-E-Mart and it won't work for non-startups.

Hey Chris, its Kris!

This is skewed heavily by employee wages. In SV they can be 2x what they are in the midwest so adjust for location.

Simply ridiculous. Sure it works out for some companies, so does my "secret" stock market strategy if you conveniently ignore the 90% of other cases where it does not work.

Just absurd.. what about 90% of all those startups that fail that have venture capital and have 10+ employees?

Exactly. Link-bait title.

What the article shows you is how to compute the burn rate of your competitor. It proves nothing about revenue. Therefore, most VC-funded companies can afford to have a high burn rate with no revenue [until they fail].

The math works if you assume your competitor is established and stable. But if you are entering the market, you'd like to know how you are impacting your competitor, and that method won't help at all.

doesn't necessarily work for Uber no?


70 seconds for a multiplication?

Looking up a head count number followed by multiplication

I figured out how you got the italic to escape your content.

End a message like this:

    Like this: *.**
Like this: .

Actually I just ended the last word with an asterisk (as in to imply I was correcting my parent comment), completely forgot about its use for formatting. Have edited it out now.

Those are billable seconds

Carry the 1.

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