Famous tech acquisitions’ cost per user

https://public.brightside.io/v1/chart/ece4947eb29375535376ca212578cf78

by RougeFemme • 12 years ago

269 95 12 years ago

public.brightside.io

selectout• 12 years ago

Didn't realize Aardvark only had 90,000 users. Huge win for the Aardvark team with that sale.

ghayes• 12 years ago

Uh, is it just me or is this chart having a hard time with comma versus period for "thousands separator"? Broadcast.com is listed as having $10,961 cost per user, but comes up as $10.961 on the chart. [Sale price of 5.7B for 520K active users]

SnowProblem• 12 years ago

It's not just you. I was about to post this.

On a related note, 11K per user is insane. What did Yahoo see in Broadcast.com?

icpmacdo• 12 years ago

I would guess for technology they developed not their user base.

genofon• 12 years ago

Mark Cuban

greendestiny• 12 years ago

What they saw was youtube.

blawson• 12 years ago

That's correct notation for a significant part of the world:

http://en.wikipedia.org/wiki/Decimal_mark#Examples_of_use

thrush• 12 years ago

Another interesting metric would be cost per user minute (would probably need to be normalized over a time period like day, week, month, or year, or quarter).

_sentient• 12 years ago

Interesting analysis, but that flickering tooltip issue is beyond annoying. A quick UI fix would be a big improvement.

abbott• 12 years ago

PayPal was a steal.

LandoCalrissian• 12 years ago

So was Youtube, pretty interesting to put it in perspective.

msoad• 12 years ago

We can add Viber for compression too: $3 per user

oskarth• 12 years ago

Wow. This graph is certainly illuminating: https://public.brightside.io/v1/chart/321ef2c603e8d67080afcd...

450000000 is a lot of users. Metcalfe's law says that the value of a communication network is proportional to the square of the connected users. Then add to this the amazing growth they are seeing. I don't think people quite realize what this means.

EDIT: In general, it baffles me how many armchair opinionators there are. Isn't it more interesting to try to understand why FB paid $16B, rather than saying why they shouldn't have?

TTPrograms• 12 years ago

This was exactly the chart I wanted to see after I saw the first one.

...but did they really put a LLS line through that dataset?

rkuykendall-com2• 12 years ago

It's an API that puts an LLS line through everything. If it isn't meaningful, ignore it.

Retric2• 12 years ago

Metcalfe's law is clearly wrong after a network get's overly large. I could in theory call a random person in China but if I have no way of communicating with them it's practically worthless to do so.

nl2• 12 years ago

Cat gifs are pretty universal.

But seriously: I've had people on conference calls when they couldn't understand the language, but someone wanted them to see slides. The fact they could join the call (in that case on Skype) definitely made the network more valuable.

You might argue that the fact you can't speak to some of them drops the value per user, and that is true to some extent. But the fact they are on there and attract other people (some of whom you can speak to) adds value for you, too.

dredmorbius4• 12 years ago

See "A Refutation of Metcalfe's Law", by Odlyzko and Tilly:

http://www.dtc.umn.edu/~odlyzko/doc/metcalfe.pdf

I'd extend the rule further and suggest that the value of a network scales as some Odlyzko and Tilly suggest, but with an additional negative function subtracting value (v) from the network:

    v = n(log(n) - f(n)

That is: as the network grows, the added value of each additional member is reduced (log(n)). Further, each additional member of the network exacts a cost to the network as a whole as well. Doing some simple modeling, I suspect that this isn't a strictly linear factor, but itself grows with n, quite possibly as the original Metcalfe's law suggestion. That is: any given member is increasingly less likely to be a positive contribution to the network, but might well present an equal opportunity to be a net negative to the group as a whole:

    v = n(log(n)) - kn^2

Where 0 < k < 1 and n > 1.

Moreover, let's look at some group sizes which might allow us to estimate for k in various contexts.

For software team size, it's very typical that a core engineering group has a size of 5-10 members, more or less. This suggests that k is about 0.2 for software development: every added team member exacts a cost, within a single group, of about 20%. This sees value grow for 1 < n < 8, then fall with larger n, hitting negative values at about n=14.

For an elementary school classroom where the ideal class size seems to be around 22-25 students, k would be around 0.08.

For Dunbar's Number, the number of relationships people can manage (100 - 300, typically set at 150), k is between 0.028 (100), 0.2 (150), and 0.113 (300).

For city sizes, it's likely that different cities offer different matches of positive and negative factors. k of 0.0005 gives value max at n ~ 10,000, k of 0.0006 is ~ 100,000, and 0.000007 is around 1 million.

To scale to 1 billion users with net positive value means you have to keep k to less than 0.00000001. That is: any one member can have only a 1 in 10 million chance of being annoying to other members.

logicallee• 12 years ago

isn't the obvious graph to ask for cost per user / revenue per user? (i.e. "acquisition price" per earnings per user.)

After all, that's the main way that we can consider a per-user cost to be "high".

tav• 12 years ago

A bunch of ZeroDivisionErrors doesn't look so pretty :)

Shivetya• 12 years ago

450 million is a number that immediately brings up the question, how many are real? This is almost as if the site is trying to make up an excuse for the purchase. However to me, having watched different entities out there I am always curious as to what is real and what isn't

chaz• 12 years ago

Interesting dataset, though I do have some quibbles. For example, registered users aren't the same as monthly unique visitors, which is how CNET is best measured. And ecommerce like Zappos have completely different metrics. One thing I would like to see is the user numbers normalized against web users at the time, which helps put late 90s acquisitions in better perspective.

pedalpete• 12 years ago

Would be interesting to see this go one step further to list annual revenue per user at the time of acquisition.

FigBug• 12 years ago

Most of those were a complete waste of money. Seems like the majority of the big money acquisitions did not live up to the hype.

arasmussen• 12 years ago

There's this thing called sorting which makes stuff a lot easier to read and more informative.....

Blahah• 12 years ago

Yup, I came here to say this.

Link-• 12 years ago

Me 3... Was actually digging into the source to see if I can find a way to sort this 'thing'.

bungabunga• 12 years ago

I am surprised Google's acquisition of the thermostat company is not included. $30,000 per user.

gregcrv• 12 years ago

I think you missed a zero... 3,200,000,000 / 1,100,000 = 2,909 dollars / user

dman• 12 years ago

I think the more fascinating metric here is cost per employee which gives us an idea of economic value created per employee. Whatsapp destroys the other acquisitions on this metric.

rtpg• 12 years ago

it's odd that we focus so much on just user numbers yet talk so much about aqui-hiring.

In most cases it's not just the users they're trying to get, but also the talent (with the idea that most of them aren't one hit wonders).

Acting like startup acquisition is a single-metric game seems like a gross oversimplification. This chart only proves it by showing there is little correlation between user count and buy price.

mmanfrin• 12 years ago

Broadcast.com's bar is incorrectly scaled down 1000x -- overlay shows cost per user is over $10k, but it lists at $10.

rkuykendall-com2• 12 years ago

I think it's the comma. The visualization code must just grab the first set of numbers.

gcb02• 12 years ago

indeed. it was the first thing i looked for... and when i saw $10 i was shocked that broadcast.com had any use let alone billions...

gcb02• 12 years ago

* user, not use. but that works too...

and does it include the youtube legal deals and fines? probably not

DivByZero• 12 years ago

Yep, that was the first thing I checked as well, actually broadcast.com is by far the most expensive acquisition.

There's probably a bug in the visualization library getting the thousands coma as a decimal separator, too bad :)

nl2• 12 years ago

Some of these acquisitions weren't done for the users.

Aardvark, Jaika, Dodgeball, Picasa, Broadcast.com, FriendFeed (missing), maybe Flickr as well as more recent acquisitions like DeepMind ($900M to Google, zero users) were bought for some combination of the team and the technology.

chavesn• 12 years ago

All this shows to me is how useless it is to compare by "cost per user". The note at the bottom says "Facebook acquired WhatsApp for 16 billion dollars, the largest startup acquisition to date. Cost per user was comparable to Google's YouTube acquisition.".

It's sort of close, but we're still talking about a swing of >30% there -- $5 billion dollars in WhatsApp's case.

There are many things missing from "cost per user" but I think the first thing to ask is "user doing what"?

davidiach• 12 years ago

I think you are right. I have accounts on many websites, but I am not worth the same amount as a user on each and every one of them.

chucknelson2• 12 years ago

Agreed - how do we know the value difference for a "user" of Zappos and a user of whatsapp (among other dimensions)? This graph is not very insightful, it's just visualizing the easy way to look at and quantify the purchase.

alexeisadeski3• 12 years ago

Oh my god...

I'd forgotten completely about YouTube. At under $2B, that must have been one of the best acquisitions of the whole bunch.

YouTube for 1/10 the cost of WhatsApp.

aragot• 12 years ago

How many were they at YouTube when they got bought?

I tend to think more about acquisitions as capital per employee, as in, that's how efficient they are at creating value. It's more interesting that comparing users because of the "users doing what" question. I'm disappointed i can't find graphs about those. Last companies I checked (but I can't remember which ones), the capital was around $1m per employee. Except for WhatsApp.

billspreston2• 12 years ago

Best acquisition? When was the last time you gave Youtube money (or Google money to watch videos)? The fact that their videos are no unwatchable because of quality stutters and the inability to rewind/jump, it's only a matter of time for a new video site to take over.

jonnathanson• 12 years ago

Yup. "Cost per user" strikes me as a very '90s metric, for lack of a better way to put it. It is an unfocused average. It presumes that the singular purpose of any acquisition is user count. This makes especially little sense in certain cases, i.e., the Facebook/WhatsApp acquisition, wherein a large proportion of the WA users are likely to be duplicative with FB users.

"User doing what?" is a decent start. But of course, we have a host of other considerations to ponder: technology, platform, integrations, competitive blocking and tackling, etc.

FWIW, I still can't wholly grok the WA purchase price. The price seems largely to have been driven by competitive bidding between FB and Google. Also a healthy sprinkling of defensiveness and fear, mixed with a dollop of "because we can." Factor in some "we believe our stock is overvalued, and as such, we will realize a discount on price correction," and I suppose you arrive at a figure that will end up south of FB's nominal purchase price. The one thing I can say is that Instagram looks likes a ridiculous bargain in retrospect, and especially in comparison.

That said, to whatever extent we do want to talk about user counts, it's worth noting some history. Around the time Facebook had an estimated $15 billion valuation, on the heels of Microsoft's advertising partnership/investment in 2007, Facebook had about 50 million active users. WhatsApp has 465M+. Apples and oranges, to some extent. But still. Worth considering.

yogo• 12 years ago

Yep, I was hoping to see some sort of estimated lifetime value in the popup (horrible btw, Chromium 33.0/Linux).

bicknergseng• 12 years ago

I think the cost per employee is even more interesting. $320 mil for every WhatsApp employee, compared with ~$76 mil for everyone at Instagram.

Now is a good time to be in the yacht business.

rayiner• 12 years ago

Kind of dumb to compare U.S./Western Europe users to primarily non-U.S. users as is the case for Whatsapp. I'd bet a 19 year old in India is an order of magnitude less valuable from an advertising standpoint than a 19 year old in America. Not only because people have less disposable income in those countries, but because kids have less control over their parents' disposable income.

rubyn00bie• 12 years ago

Uhhh... neat, but can we see profit/revenue per user in that list? I think this is pretty much useless without that context.

TorKlingberg• 12 years ago

For many of these each active user drives many passive visitors. Examples: Geocities, Flickr, YouTube.

jiggy2011• 12 years ago

How are we defining a user here? For example in the case of youtube, do we count people who watch videos? People who sign up and comment? People who upload videos?

dustcoin• 12 years ago

Relevant to yesterday's news: Acquiring MtGox for "free" would cost ~$336 per user if the leaked crisis document is accurate.

talon88• 12 years ago

No Tumblr on this chart?

allochthon• 12 years ago

One remarkable thing about this list -- with the benefit of hindsight, I only like five or six of the acquisitions all that much.

simonsarris• 12 years ago

I do not like charts like this because their existence implies that cost per user somehow matters, or is a deciding factor even, in the price of an acquisition.

Even if that was the major metric in determining a price, there's often a very large difference between the number of users currently and the expected number 1 year or 5 years out.

You don't buy a company because of the number of users it has right now. You buy it because you think you can do something better with it, or to be defensive in a field, or to expand your own customer lists.

What's more, not all users are created equal! When Flickr for instance was aquired, what percentage of users were paying users?

I think its hard to make any real generalizations here, but better than a chart of cost per user, I think this is probably closer to a chart of how badly Company X feels it needs some audience. It's some formulation of growth, panic, defensiveness, etc.

chucknelson2• 12 years ago

Overall I think there are a lot of factors, many that we the public are not privy to, that form the basis of these deals. All of this analysis and speculation is a form of entertainment. And that's fine I guess...

dublinben• 12 years ago

When Flickr for instance was aquired, what percentage of users were paying users?

Probably a lot more than there are now. As a former paying user of Flickr, Yahoo really stopped all development for about five years. Now, I have no reason to upgrade to a pro account.

craigyk• 12 years ago

Also, the metric amuses me since it implies most of the value of a tech acquisition is in the users, not the actual tech or talent being acquired. Or the timing and context of the acquisition. Is this trying to say Whatsapp was a good deal? lol.

GBKS• 12 years ago

The Critical Path discusses really well in the latest episode. http://5by5.tv/criticalpath/112

asanwal• 12 years ago

Fair point, but like it or not, cost per user matters.

- When investing in or acquiring companies, people may use this metric in addition to price/sales or in rare cases for tech companies price/EBITDA valuation multiples. It forms the basis for comps analysis which bankers, M&A types and VCs do all the time.

- Companies use the ratio in similar ways when fundraising or talking to acquirers. "Based on how many users we have and comparable transactions, we're worth $X million"

- While you may buy users looking forward, you also may look at how you can "monetize" existing users, i.e., if we can get $X per existing user, that would be worth Y.

It is of course not the only metric that an acquirer evaluates (at least the good ones), but they do look at it as do investors, bankers and startups. So good or bad, it is here to stay.

darkarmani• 12 years ago

What, showing lines of code is out these days?

EGreg• 12 years ago

One way this chart could be useful to an entrepreneur is to estimate a ballpark their startup in a certain market would fetch if they sold it. The number of active users or leads is monetizable in a sale, and can be one of the assets.

dredmorbius4• 12 years ago

their existence implies that cost per user somehow matters

Well, it does matter, after a fashion, and it's a one (of many) metrics that an acquisition can be considered on.

I found it interesting to look through the list for obvious winners and loses, and actually, that would be a much more useful plot: showing the cost-per-user on one axis against, say, the 3-5 year investment performance on the other.

Actually, better than that even would be to look at a number of dimensions and plot these against one another looking for patterns in the data. R is particularly good at this with its scatterplot matrices:

https://personality-project.org/r/r.graphics.html

https://personality-project.org/r/figures/splom.jpg

balladeer• 12 years ago

Don't you think,

number of users it has right now.

and

or to be defensive in a field

often correlate?

ChuckMcM• 12 years ago

Customer acquisition is actually a very big cost. If you're able to compute the lifetime value of having a customer you can easily make an economic argument if the acquisition makes "sense" or not based on that.

I am amazed at how well that particular calculation works some times although I understand the statistics.

rm999• 12 years ago

It would be interesting to compare this to the equivalent metric for companies that went public: the market cap per user at IPO (or maybe X days after the IPO if there's a lot of volatility?). I'm curious how the valuations differ between IPOs and acquisitions. On one hand IPOs are determined by the open market instead of opaque decisions by a small handful of people. On the other hand, private acquisitions are probably less regulated and allow for more insider information in the pricing (?).

Some other dimensions I'd be interested in are the number of employees, year, current number of users, and (for public companies) current market cap. For acquisitions I'd be curious if there are any decent objective metrics of the "success" of the acquisition.

kamilszybalski• 12 years ago

You forgot about how much Microsoft paid for Hotmail.. i think it was $40/user.. $400 million for 10 million users

michaelochurch• 12 years ago

The cost per employee numbers are what have me raging. Given how much value engineers produce (even mediocre ones, much less us closer to the 10x persuasion) how the fuck did we end up being so lowly valued?

I think the answer is that we suck at fighting for our own interests. A traditional union might not be the answer, but something like the SAG might be the way to go. It's absolutely inexcusable that computer programmers don't, for example, have the right to have a representative when negotiating with management or HR.

joshconstine• 12 years ago

Who made this? I want to cite it properly on TechCrunch?

Grue32• 12 years ago

Please don't, Techcrunch quality is already low enough even without charts based on bad statistics.

bedhead• 12 years ago

Cost per user makes EV/Sales look like an ultra-precise and dependable measure of value.

brianbreslin• 12 years ago

Can someone remind me what aardvark did?

abvdasker• 12 years ago

My thoughts exactly.

Apparently it was a Q&A site that asked questions to friends and friends-of-friends.

This chart is the sweetest of schadenfreude — so many terrible acquisitions for huge sums of money.

watermarkcamera• 12 years ago

No wechat or weixin on this chart ?

malcolmmcc• 12 years ago

Ironically, I only make Google money through YouTube when I'm not a registered user, because when I am I have AdBlock on and it's only when I'm in an incognito mode or different browser/profile that the ads get shown.

So yeah, users not necessarily the most relevant metric there. Which, upon reflection, makes YouTube even more of a steal than it looks in the chart.

carlosrt• 12 years ago

Revenue Multiple would make a more useful chart to me. Broadcast.com had $60M in revenue when they sold. Knowing they sold at ~98x revenue seems more useful to me when pricing a stock.

vishrad• 12 years ago

I really want to see Txn Size / Employee. Equity never gets meted out equally across employees. However as a worker bee I am the 'dog that eat the scraps from the table'. When the $/employee is high, even non-founder, non-dev#1 employees make out well. IMHO, the last such exit was YouTube, before WhatsApp ofcourse.

sek• 12 years ago

A FeedBurner "User" is not equal, this was a publisher service.

The real target were the feed subscribers, or maybe they are where can I check the data?

The same goes for Youtube, how many visitors did they have? They bring in the ad revenue.

Compared to Facebook/Whatsapp where you can usually only participate by being a user.

joelthelion• 12 years ago

I think a more useful coloring would be by period: "first internet bubble", etc.

ig1• 12 years ago

So this story has 210 points, 80 comments and has been up for 13 hours and yet no-one has mentioned that the data is completely wrong.

There's no accounting for inflation. Comparing a nominal 1999 dollar amount with a nominal 2014 dollar amount is like comparing a euro amount with a dollar amount without doing an conversion (which would actually still be more accurate).

Grue32• 12 years ago

Absolutely useless metric, both mathematically and economically.

When a service has a lot of users, it has no room for growth and cost per user will be small.

When a service has a few users, but has potential to grow exponentially, it will have high cost per user.

Economically there's a large difference between services with respect to their monetization. A service where each user pays $1000 will have higher cost per user than a free service.

hankScorpi0• 12 years ago

Cost per user is moot when compared against companies which inherently have different users, across different types of devices. Take our latest spot of head-scratchery ala WhatsApp, it's easy to do math and say: "oh right they paid about 40$ a user." Did they really?

WhatsApp has a huge set of "new" users on new and old devices. So, how again is a WhatsApp user on some random (non-smart) phone in India, as "valuable" to Facebook as one sitting in the Valley with their iWhatever.

drakaal• 12 years ago

Several sources estimated that 1/3 of whatsapp users were spam accounts. So the numbers are a bit off. I would also like to see this adjusted for "Today Money" which would put things like Geocities WAY up there.

I quoted a source at $80 per actual users base on the Monthly Uniques.

http://www.forbes.com/sites/shelisrael/2014/02/25/2329/

MaysonL• 12 years ago

And how much is Time Warner Cable going for? Somewhere around $2K/subscriber? WhatsApp, with 30X the userbase, costs ~1/2 as much.

Famous tech acquisitions’ cost per user

Related Stories