An archive of the best articles from Marc Andreessen’s 2007 blog. Get the ebook →

Maintained by your friends at Fictive Kin, the product studio of your dreams

Part 1 Guide to Big Companies: Turnaround!

So you’ve been hired/promoted/brought out of retirement to become CEO of and turn around your NASDAQ/NYSE/LSE-listed 5,000+ employee software/semiconductor/media company that’s recently been getting trounced by competitors, brutalized by the press, and savaged in the stock market.

Here’s your turnaround plan in 9 easy steps.

Step 1: Go dark and execute.

Your predecessor in the CEO job inevitably spent way too much time explaining to reporters, investors, analysts, and anyone else who would listen (that cute new Wall Street videoblogger?) why your company was actually doing just fine and how brighter times were just around the corner as your competitive position deteriorated and your financial results fell apart, and nobody believed it anyway.

Money talks, hype walks—when you’re hitting your numbers, everyone thinks you’re a genius and believes everything you say, no matter how silly. When you’re not hitting your numbers, everyone thinks you’re a moron and won’t believe anything you say, no matter how true.

So go dark, focus on the business, and don’t talk publicly for at least six months.

Mark “Who?" Hurd sets the gold standard here.

Step 2: But first, throw your predecessor completely under the bus.

Can’t forget this one! Tell Wall Street that your predecessor was such an incredibly dim bulb that in retrospect you can’t even understand how he got past security and into the building, much less was picked to be CEO. He completely fouled the financials and sabotaged the business and as a result, earnings for the next several quarters are going to come in way below expectations.

The fun part about this one is that your stock won’t even drop because everyone has already figured that out.

Step 3: Identify the 3-5 things that are working surprisingly well in your business, and double down on those.

Any big company, no matter how moribund and poorly run, has a number of products and projects that are going better than expected—and usually come as a complete surprise.

Drawing on Peter Drucker’s classic admonition to “focus on opportunities, not problems", figure out what these surprise successes are and double down on them.

Promote their general managers, elevate their business units in the organization, give them more funding, and get out of the way.

Step 4: Identify the 3-5 things that are consuming a lot of money and time and yet going nowhere, and kill those.

A good starting point is your predecessor’s pet projects—line ’em up and shoot ’em.

Frankly, they don’t even have to be consuming that much money. They’re almost certainly consuming time and management bandwidth, and they need to go.

You can also consider this a warmup exercise for Step 5.

Step 5: Lay off a third of the workforce.

Here’s why:

History shows that you’re going to have to ultimately do it anyway, either via death of a thousand cuts (or six to eight distinct rounds of layoffs), or all at once.

So do it all at once.

A company that requires a turnaround has, in all likelihood, hired too many people for the size of the business opportunity it actually has. This impairs profitability, driving away investors and submerging the stock price at precisely the time the company needs a healthy acquisition currency; this demotivates your great people by surrounding them by too many mediocre people and too much bureaucracy; and this slows everything in your company to a crawl because there are simply too many people running around who have to talk about everything before anything gets done.

Grit your teeth, offer the most generous severance and assistance packages you possibly can, and get it done.

Your ability to continue to employ the other two-thirds of your people is at stake.

Step 6: Reduce layers, then promote up and comers and put them clearly in charge.

A company that requires a turnaround has, in all likelihood, too many layers of management. Nuke as many of them as you can.

Then develop a list of your top 20 or 30 up and comers—strong, sharp, aggressive, ambitious director- or VP-level managers who want to succeed and want your company to succeed. And promote them, and put them in sole charge of clearly identified teams and missions. (And give them big ol’ fresh option packages.)

As CEO, you should only have at most one executive between you and these 20 or 30 up and comers once you are done promoting them and putting them in charge of their teams and missions.

If you don’t know who those top 20 or 30 up and comers are, if you don’t promote them, if you don’t put them clearly in charge of the things that matter, or if you have more than one layer of management between you and them when you’re done, you’re probably doomed.

Step 7: Figure out the single most important thing your company has to win at, and put your single best person in charge of winning at it.

’Nuff said.

Step 8: Look at the market, figure out 3-5 new areas in which your company is not currently playing or winning, but are clearly going to grow a lot—and acquire the best company in each of those areas.

Here you’re looking for growth—for products, trends, perhaps phenomena outside but adjacent to your current products and markets, that are going to grow a lot in the next few years.

You have to acquire, because if you’re in a turnaround situation, you aren’t going to have the time or bandwidth to build them in-house—unless you’re the very rare exception.

When you do acquire, you’re going to have to pay up, because new things that are growing really fast in growth markets are always expensive—whether private or public—especially compared to the PE multiple of a big company in turnaround.

So here’s hoping you did a great job at Step 5.

Step 9: In six months, relaunch the company with a single, crisp, coherent message and strategy.

Then go dark again and go right back to work.

Of course, there’s more to being the CEO of a turnaround than these 9 steps. There are a thousand other things you’re going to have to do. But these are the 9 most important.

To quote the great Tommy Lasorda: “This fucking job ain’t that fucking easy.”

Appendix for media companies only:

Step 10: For God’s sake, stop suing your customers.

This article was written by Marc Andreessen and originally published on his blog, These articles are probably some of the best writings on business and startups anywhere but they were taken down years ago. They live here now in this static archive.