It is a fairly well known fact in the business community that the majority of mergers and acquisitions are a failure when it comes to increasing shareholder value, benefiting customers or any of the other metrics that are used to judge the "success" of an acquisition. Whenever you read a news story about some startup being acquired or two large companies merging, there is a greater than 50% chance that the resulting product or company will be of less value to customers and shareholders than if the deal had never happened.

When it comes to software company acquisitions, there are additional factors working against success that go beyond the typical laundry list of reasons that are given for why M&As usually result in failure. With technology company acquisitions not only are there a minefield of people and financial issues that have to be dealt with, there is also the real problem of what to do about technology mismatch that often exists across different companies.

Whenever a large software company acquires a startup, the first order of business is often an attempt to move the startup's application onto the larger company's technology infrastructure so that it can get benefits of "economies of scale" or some other buzzword that is typically a euphemism for "we bought you so now you're our bitches" that is not grounded in business realities. This often requires application rewrites that have the unfortunate consequence of causing the shipped application to stagnate as all efforts are poured into recreating the same application using a different technology. In addition, the founders of the startup typically get frustrated with what they [rightfully] deem as a pointless exercise and eventually move on to greener pastures. There are a number of examples of this that have occurred in the "Web 2.0" space as shown below

From Fred Wilson's post We Need A New Path To Liquidity

So if you can't take a company public, how do you get out? M&A has been the primary answer in the web/tech sector for the past eight years. And it's been a great period to sell companies. We've sold three in the past couple years out of our Union Square Ventures portfolio, delicious, FeedBurner, and TACODA, to Yahoo!, Google, and AOL, respectively. Were we happy to take their money? Yes. Were we happy with the outcome? Yes. Were they good buys for their new owners? On the face of it, yes.

But if you look deeper, I wonder. Delicious grew nicely for a while under Yahoo!'s ownership but recently the user base has fallen off pretty dramatically. I double checked this chart in compete and alexa and they all show the dropoff.


Well, what about FeedBurner? Clearly Google has done a good job with that acquisition. Well I am not sure. I don't see any integration between Adwords and FeedBurner yet. I can't buy FeedBurner inventory through Google's text ad interface. I honestly don't see any additional money flowing to me, the publisher of the feed, since the Google acquisition. There's no way to know what the rate of signup by publishers has been since the acquisition, but I wonder if it's increased much.

And TACODA? I know that TACODA had an incredible fourth quarter post the acquisition by AOL, blowing way past the numbers we were projecting in our annual budget. But in the first quarter, AOL fired Curt Viebranz, TACODA's CEO, and many of the top members of the TACODA team are now gone from AOL. Another acquisition messed up.

But who am I to complain? We got paid right? So sit down and shut up.

From Joshua Schachter's comment on "How Yahoo dropped the ball with a pointless 3 year rewrite (from mod_perl to PHP)"

The writer is accidentally correct - we were told that it had to be in PHP to get ops support. Curiously the PHP part didn't take that much time - the majority of the "business logic" is in C++ which took forever and ever to write. I think the open question now is whether the remaining team will be able to innovate or be stuck in complicated codebase hell. 

From Dennis Howlett's article Google Sites - spoiled by usability issues

After 16 months at Google developer’s hands, the outcome is substandard. This is such a pity. In its JotSpot incarnation, it was far from perfect but that didn’t matter because JotSpot was shedding light on a new way of collaborating. Since passing into Google’s hands, the guts have been ripped out and then re-assembled with as much Google ’stuff’ as they could cram in but rushed to completion.

At the very least, Google should get rid of the gadgets addition facility and rework it. Otherwise, I sense the SMBs at which it is aimed will find the service a turn off.

Google has a real chance to differentiate itself from Microsoft - which is clearly what it wants to do, while adding significant numbers of users to its Google Apps offering. It won’t do it this way because despite all the gripes around Microsoft products, the fact is Microsoft offers a more polished experience. Until Google truly understands this, it will find it difficult to adequately compete. In the meantime, offerings like Wetpaint and Ning have little to fear.

From Ryan Paul's article Jaiku users flee to Twitter as a result of Google's neglect

Unfortunately, Google has allowed Jaiku to languish and is now suffering a backlash from frustrated users who are beginning to mass-migrate to Twitter, a competing microblogging service. Jaiku's external feed servers, which are used by third-party Jaiku client applications, have been down frequently during the past week, often returning 504 gateway errors or nothing at all. During the brief stints when the feed servers are operational, they have been extremely slow and often out of sync with the actual content—typically lagging by between four and 13 hours. These problems have been noted by many users and several third-party Jaiku client application developers who discussed the problem with Ars. Users also complain that Jaiku's IM bots and the third-party Jaiku Facebook interface are exhibiting problems as well.

When Google announced the acquisition, the company promised new features within a few months, but we have seen no evidence of any development at all. Registration is still closed and new users can only join the site by receiving an invitation from Google. The Jaiku developers have been completely and totally silent since the announcement of the takeover, and the official Jaiku blog—which used to have several messages a month—has had no new posts at all. The Jaiku Team feed has also not received any posts from Jaiku developers since the acquisition.

The stories are the same except that some of the names are different. A startup gets bought and immediately stops innovating because all their development time is being spent porting the code to a new platform. During that time newer, more agile competitors show up and eat their lunch. Why I find this to be such a conundrum is that when you buy a technology startup, you are primarily buying three things

  • customers
  • employees
  • technology

However the standard operating procedure during Web software acquisitions is to discard the technology and consequentially tick off the employees who made the product a success in the first place thus creating an exodus. The application rewrite plus employee exodus leads to product stagnation which eventually leads to lots of pissed off customers. Thus the entire value from the acquisition is effectively thrown away.

This is the default situation when it comes to acquisitions in the software industry. For every successful acquisition like Google + YouTube there are two or three that are more like Google + Dodgeball. So if there is a startup whose product you love that you hear is being acquired by a one of the large Web companies, be happy for the founders and be sad for yourself because the product you love is likely going to become a neglected bride.

Disclaimer: I'm an employee of a large software company that has displayed similarly counterproductive tactics when acquiring startups. Although no examples are provided in the post above, I'm sure some can be found from it's history of acquisitions.

Now Playing: Dire Straits - Money For Nothing


Tuesday, August 12, 2008 5:01:04 PM (GMT Daylight Time, UTC+01:00)
This is called hype cycle, you are drawing the wrong conclusions.
not relevant
Tuesday, August 12, 2008 5:36:36 PM (GMT Daylight Time, UTC+01:00)
Having left a large company that did a fair number of acquisitions, i can say that often both parties are equally guilty. The big company wants one cohesive whole because it's easier to maintain when folks leave (or when there's some new security threat or some new technology that can be used). The little company often sees a chance to ditch something crappy they built for something far better. As you note, the best solution often lies somewhere in between, but it's very rare that anyone associated with both ever sees that magic balance point, and even rarer that someone manages to fight both internal and external forces that continually try to shift that point.

In an ideal world, Big Companies would be structured for acquisitions and have tools that work as services that can be plugged into foreign architectures easily. Likewise small companies would structure their code as modular as they can in order to accept those kinds of services.

In reality, startups hack together stuff that often barely works and push it out as fast as they can while big companies have to support legacy stuff they can't afford to take down and re-write.

All the while, the original customers grow angry because that "underground" service that they loved sold out, and they start paying attention to the next "hot" service.

Thus the circle of life.
Tuesday, August 12, 2008 6:12:20 PM (GMT Daylight Time, UTC+01:00)
I'm interested in seeing what happens with PowerSet. As I recall, they are a Java house built on top of Hadoop. Considering that Dryad is a pale comparison to Hadoop (at this time), I wonder if they'll be encouraged to move to it?
Tuesday, August 12, 2008 6:19:15 PM (GMT Daylight Time, UTC+01:00)
While an entertaining article, many of the ideas you present show to me your lack of understanding of the world of mergers and acquisitions and business as a whole. I won't beat you up about them but only point out a few.

1. The Fred Wilson post you point to is full of wild opinion. For example, the graph that is prominently displayed shows Number of Unique Visitors. What does this tell us, other than there was an initial shake off after the, as the first poster said, "Hype Cycle" ran its course. I see an uptick at the end, does that mean the user base is again increasing? Ah! Growth! And because he "does not see any integration between Adwords and Feedburner yet" does not for Google a failure make. How long was it between Apple's acquisition of Next and the introduction of OS X? 4 years?

2. The Google-Answers post you point to that shows that "the majority of mergers and acquisitions are a failure" shows me a different picture. One third of them, according to the post, "achieve the intended goals which were the stated reasons for the business deal". Does this mean that the other two thirds fail? What happens when the stated business goals change after acquisition? In the business world we tolerate "flip flopping". Also, on the same topic, how many internal R&D projects fail within a corporation? It depends on what your definition of failure is. If you gain a technology or process, but lose the product as a whole, is it still a failure?

And the final piece I will state my opinion on:

3. You state that "when you buy a technology startup, you are primarily buying three things

* customers
* employees
* technology"

I will have to correct you here. When you buy a technology startup you are buying technology, technology, technology. You typically have a customer base. You probably have employees. What you are lacking is the technology of the company you are acquiring. Unless of course they are a competitor and you are just slimming the field. Purchasing the startup is probably a fraction of the R&D needed to gain the technology. There are so many other reasons to acquire a company that are not customers or employees.

Some times the goal of the shareholders, management and board is much different than the goals of the employees, the acquired company and its founders, and the customers that are using the technology. Management's job is to make business decisions and hire others to make technology decisions. It doesn't make their goals wrong, only different. As a person who has worked and consulted in the tech sector (hardware side, and software side) I have seen, and still see, those who pummel "management", which is always bandied as if it is a dirty word, for their decisions without any first hand knowledge of the situation involved.

I'll leave it there.

Again, thanks for an entertaining read.
Tuesday, August 12, 2008 8:06:03 PM (GMT Daylight Time, UTC+01:00)
not relevant,
There are six examples in my post along with several links to reasons why mergers and acquisitions fail. It seems you got distracted by the chart and decided to focus on the shape of the curve instead of the content of my post. This reminds me why I try not to add pictures to my posts since it tends to distract people from the content.

If your thesis is correct and technology acquisitions are only about technology AND there are numerous examples of startups being forced to reimplement their technology on acquisition...doesn't that mean we are in VIOLENT AGREEMENT? :)
Tuesday, August 12, 2008 8:19:07 PM (GMT Daylight Time, UTC+01:00)
you can tell you work in software...i'm sure you don't want to re-invent the wheel, but how about some MBA type research?

Remember Netscape? Their rewrite is the stuff of MBA case studies...

Wednesday, August 13, 2008 2:26:50 PM (GMT Daylight Time, UTC+01:00)
Having had the distinct unfortunate luck to have been a part of the gobbled up enterprise in such a merger, I can report that the primary motivation of the executives in the absorbing organization is to assert control, nothing else matters, in fact, quality and customers be dammed is the rule.

The larger companies executives, out of fear or greed or whatever, have only one purpose and goal, and that is to establish complete authority and ownership of the new aquisition. The more intelligent, skilled, and valuable the subordinate new employees are, the more likely they are to get the axe, as the new overlords eliminate potential competition, and after asserting control, seek to place thier own flunkys / brown noseers into positions of power beneath them.
Barnell Fitzcolmb
Wednesday, August 13, 2008 8:06:51 PM (GMT Daylight Time, UTC+01:00)
Do large companies really destroy software value? It takes two to tango, no?

By committing to being acquired, aren't startup founders also to blame for any destruction in the value of the software? It would seem that their commitment to the value of the company precedes the value of the software. If the founders aren't truly invested in the value of the software, how can anyone else be?
Wednesday, August 13, 2008 9:09:41 PM (GMT Daylight Time, UTC+01:00)
This is an interesting analysis. I've wondered for a while why so many acquisitions go bad -- but I've never thought of just why it always happens that way. Other comments are mentioning that it takes two to tango -- which I agree with, but we have to remember that management (owners) typically hold all the cards in the end.

We've all been there -- working for people who are horrible, and don't understand how to make connected and coherent business decisions. And the 'people magic' gets destroyed.

I'm sure post-acquisition employees could do a better job of trying to be team players, but if the culture and strategic plan off the acquiring company is strong enough, it should all work .. right?
Thursday, August 14, 2008 7:38:49 AM (GMT Daylight Time, UTC+01:00)
Feedburner is a good example. I don't remember how long ago Google bought them, but it's certainly been over a year. So far I've seen pretty much 0 improvement to their service, other than making previously paid features free to all users. There's still no simple way to add Feedburner to Blogger, which Google also bought. Why isn't there a one-click widget? Or have the two integrated? It would only make sense... but so far both services seem to have languished. Blogger has been getting steady improvements, but for some reason a lot of it is still a mess..
Friday, August 15, 2008 3:31:38 AM (GMT Daylight Time, UTC+01:00)
Delicious's traffic is now higher than the *peak* of the chart you're featuring... it's old!
Wednesday, August 20, 2008 4:51:52 AM (GMT Daylight Time, UTC+01:00)
When Microsoft acquired PREfix, they very wisely *didn't* pressure us to rewrite it. It was actually a difficult decision because we were using a non-Microsoft C/C++ front end, from Edison Design Group, and the lawyers were concerned that this could potentially expose the Visual C++ people to non-Microsoft intellectual property. We actually weren't able to build the code for first month after the acquisition until things got resolved. In the long run, though, it was clearly worth it. A year or so later, we embarked on replacing the front end with the AST toolkit, which by then had been used successfully by PREfast and had much better VC++ compatibility -- the right thing to do at that point. But if that's where we had started, we probably would have been added to the long list of casualties from an unncessary immediate rewrite.

Which is a long way of saying: I agree. Well said.
Comments are closed.