What Colin Powell Knows About Winning

I was the first CTO at Hulu, and our obsessive early strategy was a lot like the famous general’s ‘main attack’ theory of war and politics

Eric Feng


Credit: Daniel Zuchnik/Getty Images

WWhen Apple reported Q3 earnings this past July, they once again set an all-time high in services revenue at $11.46 billion — their fourth consecutive quarter of all-time revenue highs from their services business. Apple’s services strategy has been years in the making, but also recently become more prominently discussed, with Apple disclosing services margins for the first time in January and then hosting its first ever services-only marketing event in March. As Tim Cook said at the end of that event: “From everything we’ve shared with you, you can see how important these services are for us.”

Will their upcoming Apple Card, Arcade, and TV+ services be a hit with consumers? Can Apple offset stagnating hardware sales with growing services revenue? Can they reinvent themselves from an innovative hardware company to be a leader in digital services? It’s still the early days, with lots of questions about Apple’s services future. But one thing is very clear: Apple has made services their main attack, and that in itself is a smart idea.

The Main Attack

Former U.S. Secretary of State Colin Powell first outlined his idea of “decisive force” in the 1992 National Military Strategy, in which he stated: “One of the essential elements of our national military strategy is the ability to rapidly assemble the forces needed to win — the concept of applying decisive force to overwhelm our adversaries and thereby terminate conflicts swiftly.”

Executing the main attack means both selecting the right target and then putting everything you can possibly muster into the resulting fight.

One afternoon a decade ago, I was at an offsite meeting at Kleiner Perkins, and former U.S. Secretary of State and four-star General Colin Powell was holding court, sharing incredible insights and wisdom from his distinguished 35 year military career. This is a man who was Commander of the U.S. Army Forces Command, led 28 military operations, received the Presidential Medal of Freedom, not once but twice, and was a New York Times best-selling author, just to name some of his accomplishments. As we sat in the room, transfixed and mesmerized, General Powell said something I’ve thought about constantly to this very day: My most important lesson from the battlefield is you must have a main attack.

The definition of a “main attack” in military terms is exactly as you would expect: “The part of an attack where the commander concentrates the greatest portion of offensive power.” There are actually two principles hidden within that single statement. The first — “the part of the attack” — is about identifying the right opportunity. The second — “greatest portion of offensive power” — is about fully committing to the chosen action with your full resources. Executing the main attack means both selecting the right target and then putting everything you can possibly muster into the resulting fight.

To take a real military example, in the Persian Gulf War (during which General Powell was Chairman of the Joint Chiefs of Staff), the main attack was a huge ground incursion on Iraqi forces in Kuwait beginning on February 24th, 1991. Crossing from the western border, General Norman Schwarzkopf led more than 150,000 troops and 1,500 tanks into Kuwait from Iraq. The main attack didn’t stop until there was a ceasefire and surrender by Iraq in just 4 days. It was a resounding victory that informed and reinforced General Powell’s theories around the main attack.

Powell would also apply the main attack theory to other aspects of his life, such as politics. In 2010, General Powell referenced a main attack when critiquing the first half of President Barack Obama’s inaugural term. While President Obama had passed legislation on economic stimulus, health care, and education, as well as made progress on energy reform, the Democratic party also lost 63 House seats during his first midterm election. The reason according to Powell? Because Obama didn’t have a main attack on which the Democratic Party and its candidates could hang their hats.

In an interview with Larry King, Powell said: “It’s not that other things aren’t important. Health care is important. Our energy policy is important, education policy. All of this is important. But when you’re starting out as the president, you have to figure out which is the most important. In military terms, we say what’s the main attack? Everything else is important, but what’s the main attack? And in my judgment, the main attack was to do something about the jobs situation, the unemployment situation.” Instead of focusing on a variety of important issues for the United States, Powell believed Obama should have focused on the single most important issue (which he considered to be unemployment) and thrown every available resource at his disposal until the jobs battle was decisively won. Yes, other critical problems for Americans would still need to be solved. But the President could have faced them with the momentum of having first solved the critical problem. The momentum of having won his main attack.

Colin Powell deftly applied a military strategy — the “main attack” theory — to political strategy. And it turns out it would be just as applicable, important, and useful in another non-military situation: business strategy.

The Main Attack in Business

Comparisons made about business are usually used to describe competition: Coke vs. Pepsi. Nintendo vs. Sony. McDonald’s versus Burger King. But some of the most important battles a company faces are internal struggles: when the problems are overwhelming, when priorities aren’t clear, teams aren’t focused, solutions too difficult to build, and more. These have nothing to do with outside competitors, yet can still threaten a business just the same. And in those situations, a main attack can be the best response to that threat.

I know this firsthand. I was the founding CTO at Hulu, and in those early days — back when we were still called either Newsite or Clown Co depending on whom you asked — we had no shortage of tough questions to answer about our business: How will we acquire enough content to keep users engaged? Will advertisers be willing to purchase our ad inventory? How can we differentiate from YouTube, MySpace, Yahoo, and other huge websites where people stream online videos? Can we recruit a world class team of builders? And unlike the endless problems, our time and resources were finite.

So how did we move forward? We launched our main attack.

Original Hulu.com homepage

What we realized mattered most at Hulu to all our customers, advertisers, and content partners was watching videos. Period. So video delivery was our top priority, and became our “part of the attack” (referring back to the definition of the main attack). We would make our video viewing truly world class, or as Hulu’s CEO Jason Kilar liked to say, an experience “worthy of remark.” That was the target we selected, and from there, we applied our decisive force.

All available resources were rallied around the goal of making video delivery the best we could make it; from the business team to the operations team to the engineering team to the product team. On the business side, we started requiring mezzanine files or full resolution master video files (i.e. the highest quality source material) from content owners when signing deals. Our ops team would go to any and all ends to bring these files in house, often traveling on location and hunting through storage rooms, closets, and basements for tapes and hard drives.

Original Hulu.com video page

Next, the engineering team worked for weeks tuning encoding profiles to create the best possible video files. We purchased multiple third party encoding services to try out. We hired away contributors to the FFMPEG open source project. We kept tuning codecs, resolution, video bitrates, audio bitrates, frame rates, keyframe rates, and dozens of other parameters, and ran focus group after focus group on the output like a chef tasting batch after different batch of a dish to perfect a recipe. We went all over the country doing deep packet inspection of video streams and partnered with content delivery networks to determine where to strategically host video files. And remember: this was still in the early days of video streaming before fiber, LTE, MP4, and adaptive bitrates made life far easier.

Finally, on the product side, we designed every pixel on the site with the goal of frictionless video delivery. We selected every video thumbnail by hand — more than 20,000 in the first year — cropped to a 16:9 aspect ratio (instead of the more common 4:3 aspect ratio) to make the thumbnails more visually informative about the content. We minimized non-video content, like reviews and descriptions, to maximize the number of videos we could present to users. We even embedded videos on static pages like the company’s About page and the 404 error page (when someone navigates to a URL that doesn’t exist) so that users were never more than 1 click away from watching a video.

Original Hulu.com 404 page

Did the main attack work? Did focusing on video delivery end up making the positive impact we were hoping for? Thankfully it did, as illustrated by the media’s reactions at the time:

GigaOm: “This is an awesome service, one that worked flawlessly… The quality of the video shows is good enough to enjoy without straining the eyes, and even in the full-screen mode, the Flash video looks pretty amazing.”

PCWorld: “Hulu may offer the best-looking, most watchable web video to date, rivaling the standard-definition content of regular TV.”

TechCrunch: “I was very impressed by the preview of Hulu’s interface and the bulk of its features… Hulu has done a good job keeping the user interface simple and highlighting the actual content of the site.”

Of course, having exclusive access to excellent TV shows made a huge difference. But I also assign a lot of credit for Hulu’s early success to having a main attack on video delivery. It gave us clarity and focus. The entire company had a singular mission “drilled into our hearts and into our heads” (to quote General Powell). And most importantly, it was a mission we never allowed ourselves to stop working on or investing in. The main attack has to be followed through to the end — the mission can’t stop until all resources are exhausted, and there’s nothing more to do besides declare success. Only once we were confident in our video delivery experience did we start thinking about other priorities, needs, and unanswered questions at Hulu.

And it turns out those unanswered questions were more straightforward to answer after winning the main attack. Content owners and advertisers were eager to work with a platform that treated their media assets with care and respect. Users chose to spend time watching videos that looked premium and high quality. We recruited builders who believed in obsessing over every pixel in the content experience. The main attack created a foundation for Hulu that we could build future success on.

And that’s the real key to the main attack. You’re not putting all your resources behind solving the hardest problem, but rather behind solving the most important problem that will make subsequent problems easier to deal with, and subsequent goals more likely to be achieved. Look what happened when new Microsoft CEO Satya Nadella made cloud computing their main attack. Since 2015, Microsoft has doubled its quarterly Intelligent Cloud revenue (from $6 billion to nearly $12 billion) and tripled its stock price. But more importantly their main attack has given the company a new identity that mirrors many of the fundamentals of the cloud. Microsoft is a platform company, focused on the enterprise, and embracing openness and flexibility for their customers — just like cloud computing is. Because of the cloud, partnerships are now easier to come by (like with once rival Sony), sales are more durable (because of recurring revenue from subscriptions), and Microsoft is more relevant with developers than its been in many years (hence the Github acquisition).

It’s the same with Netflix’s main attack on original programming, or Domino’s main attack on digital orders, or Target’s main attack on differentiated merchandise and private label. Or back to Apple’s main attack on services and Hulu’s main attack on video delivery. The main attack isn’t just another battle to be won. Instead, it can be the launching off point for winning a much larger war.



Eric Feng

Current: Co-founder of @cymbalxyz, Co-founder of @GoldHouseCo Ventures. Past: @Meta (via Packagd), GP at @KleinerPerkins, and CTO of @Hulu and @Flipboard.