Ranking tech companies by revenue per employee | Hacker News Reader

Ranking tech companies by revenue per employee

37signals.com

65 pointsspivey37 comments

These numbers are interesting, but using them as a ranking method is absurd. Revenue/Employee punishes vertical integration. Let's say that Facebook, instead of developing Casandra in-house, decided to use millions of dollars worth of Oracle RAC. Their revenue/employee would go up, presuming that they could have avoided hiring those good developers. But, their profits would likely be lower. What's the revenue per employee of a private equity firm? Isn't Chrysler an LLC wholly owned by Cerberus? Revenue per employee must be well over $10M!

Agreed. When I worked at a fabless semi company, we compared ourselves to other fabless semi companies using this metric, and given similar levels of vertical integration, it's not a bad metric. Using it to compare a hedge fund and a fast food chain is indeed absurd.

Haha - indeed. Centaurus Energy (TX based hedge fund) would amount to $88.2M per employee.

Every valuation metric requires context. Like sparky said, you can't compare two different types of businesses the same way. In retail, you want razor thin margins. In software you want very high margins. In capital intensive businesses like factories, you want lots of debt. In software businesses you want no debt.

Building tools in-house, like cassandra, would be considered a distraction at most public companies. Expensive efforts that are not within the business's core competency are signs that the organization may be making bad investment decisions or they are bloated and trying to keep employees busy.

In fb's case, their core competency is information, so building better information management tools is part of their core competency and they have the talent on staff to do it.

If GM or Walmart spent employee brain power on something like cassandra, they'd be blasted by the shareholders for re-inventing a wheel that is already good enough.

I think it's clearly implied that profit per employee would be far preferable, but in the absence of good profit data, they're saying that revenue per employee is vastly more useful in judging a business from a business viability standpoint than, say, users, or even users/employee.

If the startup's goal is solely to be acquired, then users might be appropriate, but 37Signals has always argued that that goal is generally foolish, and akin to playing the lottery. Their mantra has been that startups should be trying to build a business, not an acquisition target.

When I was investing in public companies, this metric was one I used to decide where to put my money. I think it is many times companies settle for hiring as many people as possible to get their profit margins slim. This MO produces companies with a lot of fat who need to lay off employees when the inevitable down turn arrives. They often times over promise and under deliver.

Most public companies I researched had revenue/employee between $100,000 and $200,000. This simply isn't high enough to provide any cushion.

The reason here, that Craigslist is at the top, is because they outsource information management to the community through flagging and karma assignments. Most of the work of the employees is handling the exceptions to the algorithms, which usually result in the form of an email to craig or a post to the feedback or help topics. These may result in a kind email from craig or banning of a spammy account.

They are so good, they already have phone based account validation. None of the other companies on this list have that. CL focuses their employee time and energy on what is important -- stuff only humans can do. The rest is done by computers and this is a brilliant sign that they are doing it right.

Most companies handle exceptions with bureaucracy, filling up revenue with salaries until the boat sinks.

I think it is many times companies settle for hiring as many people as possible to get their profit margins slim.

Do you have any insight about the motivation to settle for slim profit margins?

This seems counter-intuitive, unless by shrinking margins, they make a significant increase in the bottom-line. (i.e. they would rather make 1% of 40 billion than 20% of 100 million)

Is this sort of trade-off common? Does it work out favorably often enough that it can be backed by something other than managerial fashion?

In large organizations, the decisions made by the employees are often not related to the bottom line. They'd rather have more direct reports than increase profit margins. Middle managers feel powerful because they tell many people what to do, not because they do more with fewer people.

Of course the investors want the company to be more efficient, but not always. This is true in the retail industry where razor thin margins are the goal. Software is the opposite. Imagine if Walmart made 35% margins like many software vendors. They'd be seen as gouging their customers and their prices would be too high. Target would create slimmer margins and take walmart's customers.

Also, employee salaries are tax deductible as expenses, so companies that are intent on employee satisfaction -- something seen as beneficial for successful organizations -- tend to pay out much, if not all profits as bonuses. If a company is making lots of profits, they have to put that money somewhere, either investments in either infrastructure or labor and it's fairly easy to hire people. Otherwise, it goes to the government and no one likes that. Employees are expensive and hiring more is a quick way to legitimize a budget. Really, investors care about revenue growth more than anything. A company that is growing revenues and growing the share price is considered good. The profits could be paid out as dividends, but fewer companies are doing that and they are double taxed, so capital is more efficient if left in the organization. Exceptions are very large, old and stable companies, but that's a different kind of investor. Those are income investors, not growth investors like I was.

It seems counter-intuitive because investors seldom have control over this kind of spending and revenue/employee isn't a popular metric for investing like PE, PEG, OM, or PM. For example, look at the key statistics for MSFT at yahoo. http://finance.yahoo.com/q/ks?s=MSFT+Key+Statistics The number of employees appears nowhere and thus calculating revenue/employee is beyond the ability of most investors. Sometimes you can get # of employees on the profile page: http://finance.yahoo.com/q/pr?s=MSFT+Profile

Essentially, the rev/emp metric when making investment decisions is "outside the box." Most investors don't care or even think about using it as a filter, nor is it even really possible without better investing tools.

I don't follow the logic as it pertains to retail. It's quite easy to get revenue per employee to be as close to zero as you want. Simply hire more workers. Investors in retail stocks are not looking for revenue per employee to be as low as possible. They are more focused on things like sales per square foot of their stores.

Razor thin margins are the goal but only if it is achieved through the right process. Hence, people look at other metrics.

Razor thin margins are a sign that the retail business is operating as efficiently as possible.

Apple: 46710 million / 34300 employees = 1.36180758 million/employee

Dell: 52 900 / 94 300 = 0.56097561

Microsoft: 58690 / 93 000 = 0.631075269

Intuit: 7800 / 3260 = 0.417948718

Intel: 38280 / 79800 = 0.479699248

The stuffy large companies seem to make about 0.5 million /employee /year. The more aggressive ones(Amazon, Apple, FaceBook, Google etc) seem to make about a million/employee. ALmost certainly doesn't hold up statistically. Just something that struck me.

This stuff is interesting, and we should add more companies.

I took this table from wikipedia (http://en.wikipedia.org/wiki/List_of_the_largest_global_tech...), put it in a spreadsheet, and added a column for revenue/employees. Anybody can edit, so feel free to update with more recent figures and add other companies.

http://spreadsheets.google.com/ccc?key=0AqI3DInWs2nKdDdMcXlL...

What happened to that spreadsheet? It's been completely cut down (and not even to the top-N as there are many other tech companies with better numbers that were removed). Folks added many more companies than are currently showing and we had columns for Profit and Profit per employee (which I think is a much better metric).

The Apple number is very different if you don't count people working in their retail stores.

Well, sure, but why wouldn't you count their retail staff?

Some numbers for other companies, in a few different industries influenced by technology (taken from Wikipedia):

Dow Chemical 57514 / 46000 = 1.25 Million / employee

BASF 67787 / 104780 = 0.65

GE 157000 / 323000 = 0.49

Merck 27400 / 61500 = 0.45

du Pont 26100 / 60000 = 0.44

Alcoa 26900 / 87000 = 0.31

(These are revenue... profit margins will vary, of course.)

Some micro economics: An enterprise is supposed to add resources until the marginal revenue equals the marginal cost. In this case, it means the additional revenue produced by the employee equals the salary.

Some of these 'efficient' companies are likely leaving value on the table by not adding resources to the point of zero marginal profit.

It's kind of strange to call Craigslist an enterprise, in that case.

Realize that the chart you are reading is revenue divided employees. It is not a measure of the actual revenue generated by each employee divided by his individual salary. Calculating marginal revenue and marginal costs is not that simple for the sort of IT companies listed.

The point I was trying to make was that the proposed corporate 'metric' could easily oppose basic economic best practice.

That said, it should be no surprise that Craigslist is at the top - they might be leaving more value on the table than any other company in the world.

Do you have facts that backup your hypothesis?

Strategic choices don't exist in isolation however.

Craigslist owes it's success to a deliberate focus on pleasing it's audience (well that and luck of course). If the prevailing attitude had been on of extracting all available value, it's likely craigslist would't have survived to see it's current success. They would have wasted away at some local maximum.

While in theory it's true that you should always make moves that are +ev, in reality we cannot approach the complexity of operating a business that way. Change has a huge cost in organizations, and the larger and more varied the lines of business you're dealing in, the more expensive that change is.

Put a different way: organizations aren't markets. Organizations can't be ignorant of economic principles, but economic principles alone, particularly microeconomic ones, do not completely prescribe strategy.

Another way to look at that is that (eg Craigslist) with high revenue/employee money on the table are those with potential: Craigslist aren't going in top gear. Apple are having trouble keeping up. They have more profitable things to do with their time then take all possible money off the table.

This article implies that employing more people per dollar revenue is inefficient. However this is only true for comparisons in a particular industry segment.

Starbucks' revenue per employee is less than any of those but it doesn't make it less inefficient, just a different model.

If you flip the logic round and apply it to customers, 37Signals looks very inefficient. Oracle has few high paying customers whilst 37Signals has many cheap ones.

Comparisons of rev/customer or rev/employee are always interesting but only meaningful in a single industry segment.

This feels a little like asking the revenue per pound of a person. If two people earn the same amount, and one if more slender than the other, the slimmer person makes more per pound than his friend. If he were to increase his weight would his earnings increase? I suspect they won't. This is like the view that Craigslist could increase their earnings by hiring employees. Companies like people are organic complicated things. Reducing them to ratios is essentially meaningless.

This is silly in the sense that it really doesn't consider companies who have attempted to scale horizontally in attempts to capture different markets than their core interest.

I'm going to go out on a limb here and say Microsoft, absolutely could fire everyone not related to developing/selling/supporting Windows, still make 5 billion dollars a quarter from software sales, and easily be at the top of this list.

Likewise, Google's pure search and adwords team is probably much more 'efficient' at generating revenue than say the Android team, but you get diminishing returns at a certain point with your core product line. The smart thing for a company to do is to grow the company's revenue (and hopefully profit) when the time is right into other, less high margin sectors.

CraigsList is at the top because they do one thing very well. In five years, if they were a public company, their shareholders wouldn't be too happy if they kept their #1 efficiency spot but didn't continue their growth.

This metric puts a high advantage on companies that outsource people-heavy parts to other companies. Buying your customer support from an outside company greatly increases your revenue/employee even if the cost for support is the same or higher than in-house.

Interesting comparison, yet there are so many wrong implications that can be made from it that I don't Even know where to start.

Compare to: short guys in glasses earn 100x as much as tall athletic guys. "Proof": compare Warren Buffett and Bill Gates with the local college basketball team.

They're not really drawing implications in that article. Profit(or revenue)/employee does provide a fairly decent, albeit rough, measure of efficiency. I think it's an interesting metric that more companies should look at. If that metric starts declining while your hiring is increasing, then it might be a good sign that you're becoming too large and bloated.

"Dividing revenue by employee gives us a better look at which companies are the most efficient. And that’s what you see in the chart at the top of this post."

craigslist's numbers are misleading here. Sure, they're extremely efficient by the above metric, because they were lucky enough to become a de-facto standard marketplace. I challenge anybody to find another example of a software company with 30 employees that is nearly as successful.

MSFT revenue per employee is $631k, between eBay and Facebook.

Source: http://www.wolframalpha.com/input/?i=microsoft+revenue+/+num...

(PS. Do I deserve extra karma for posting this cool WolframAlpha link ? :)

How is CL misleading? They make a shitload of money with very few employees. Saying they were lucky isn't even relevant to the argument. Based on revenue per employee, they are fantastically efficient.

MSFT was "lucky" enough to become the de-facto standard in their market, yet they aren't as efficient as CL.

Efficiency? More like profit margin. Labor intensive industries can be very profitable, provided the scaling is efficient.

No.. efficiency. Your margin is relative to your total costs, of which employees are generally not even a large fraction. For instance take a site like Photobucket, they have a relatively small headcount (~150'ish I believe, most of which are not developers). Their cost drivers are infrastructure and bandwidth which I imagine is in the millions/month.

This is a (very) rough measure of how efficient you are in terms of generating revenue. Craigslist is a big site, if they only have 30 employees (as cited above) than I have to imagine they have huge parts of their day-to-day operations automated. That's efficiency.

"Efficient in terms of generating revenue" (per employee), sure, but that isn't the usual definition of economic efficiency. In particular, there's no real reason, when looking at a company's efficiency, to look at only a raw count of employees, and not take into account things like how much those employees cost, and what your non-employee costs are. A company with half the employees but orders of magnitude more infrastructure costs and lower profits, on the same revenue, isn't more efficient, even though it would rank better on this metric.

Your analogy is poor, not just because neither Buffet nor Gates are short. I agree that the author's list of companies was selective, and a larger list would have provided more insight. But revenue per employee is an important metric for understanding the scalability of a company.

I agree with you that my analogy is poor, and that revenue per employee is an important metric.

My point was that showing craigslist there is misleading, for the reasons I listed in my other comment.

It still amazing how well Google scales, and comparing it to craigslist shadows this fact.

More than anything, they just don't include that many companies. They need more in each size category, and some sort of standard deviation would be interesting... actually there are all kinds of interesting statistics you could run.

very interesting and gives a good idea!

still a very high-level view of things and rather imprecise (as probably most comparisons) ...

cheers

Reply on news.ycombinator.com