Microsoft + Seinfeld + Gates: The $300 Million Campaign Nobody Understood
Microsoft paid Jerry Seinfeld $10 million and Bill Gates an unknown sum to star in ads that had no discernible message, pulled them after two spots, and replaced them with a campaign that actually worked.
The first spot opened in a shoe store. Bill Gates, shopping for shoes. Jerry Seinfeld, shopping for shoes, recognizes Gates and follows him around. They talk about churros. Gates wiggles his toes to indicate yes or no to questions. Seinfeld mimes eating a churro and asks Gates whether Microsoft will one day make computers that are “moist and chewy like cake.” Gates wiggles his toes. The screen fades. A Microsoft logo appears.
This aired in September 2008, during NFL football, in the slot that cost the most money available to a media buyer. Tens of millions of people watched it and had no idea what they had just seen.
The Context
By 2008, Apple’s “Get a Mac” campaign had been running for two years and had done genuine damage. The ads featured a cool, young Mac (Justin Long) and a frumpy, put-upon PC (John Hodgman), and they made a clear argument: Macs are for creative people who don’t want to deal with crashes and viruses; PCs are for people who’ve given up on joy. The campaign was clever, mean, and effective. Microsoft’s market share wasn’t actually threatened (Windows ran on the vast majority of computers in the world) but the cultural narrative had shifted, particularly with younger consumers, and Microsoft’s brand had come to feel defensive and stodgy.
The brief that went to Crispin Porter + Bogusky, the agency behind Burger King’s Subservient Chicken and other genuinely unconventional work, was reportedly about humanizing Microsoft. The company wanted to feel less corporate, more approachable, more human. The insight Microsoft was reaching for wasn’t wrong: Windows is used by billions of real people to do real things, and that diversity of real use is a story Apple’s premium positioning can’t tell.
Crispin Porter, riding a run of creative success, pitched the Seinfeld idea. Jerry Seinfeld’s Seinfeld reruns were in heavy rotation; the comedian had become comfortable as a cultural institution rather than an active performer. And Bill Gates, having recently handed day-to-day operations to Steve Ballmer, was newly available for the kind of self-deprecating public appearance that would have been off-brand for a sitting CEO.
On paper, you could see the logic.
The Campaign
Two spots were filmed and aired. The first was the shoe store, with its churros and toe-wiggling. The second showed Seinfeld and Gates having moved into a suburban family home to “reconnect with regular people,” sharing a bathroom, helping with chores, and engaging in gentle comedy with the family. At one point Gates wore a Conquistador costume to cheer up a child.
Neither spot mentioned Windows in any meaningful way. Neither made an argument. Neither had what Seinfeld’s own show had in every single episode: a setup, a development, a payoff. Seinfeld’s comedy is observational and precise; it works by identifying a specific shared human experience and exaggerating it until the recognition becomes funny. The Microsoft spots had no such specificity. They were just two famous people being mildly odd near each other.
The creative theory, apparently, was that unexpected and absurdist content would generate buzz, and that the humanizing effect would come from the sheer unexpectedness of Gates submitting to weirdness. But buzz in the service of nothing is just noise. Viewers weren’t confused in an intriguing way. They were confused in a frustrated way, which is a different experience entirely.
After two spots, Microsoft pulled the campaign. The “I’m a PC” campaign replaced it, featuring real Windows users from around the world: teachers, scientists, artists, children, all identifying themselves as PC users. That campaign worked. It was simple, inclusive, and made an actual counter-argument to Apple’s “Macs are for cool people” positioning by demonstrating that PCs are for everyone.
Why It Failed
There’s a version of the Seinfeld/Gates idea that could have worked. If the spots had a point, if the celebrity weirdness had been in service of a specific argument about Microsoft, or Windows, or what computers actually do for people’s lives, the unexpected pairing might have carried the message further than a conventional ad could. Celebrity casting is a delivery mechanism. It accelerates reach and attention. But it can’t generate meaning where none exists.
Seinfeld’s comedy requires precision. His stand-up is famously specific: it’s not “relationships are confusing,” it’s “why does the person leaving the voicemail tell you to call them back if you already have their number?” The Microsoft spots gave him a vague assignment and a famous co-star and then trusted that something funny would happen. That’s not how Seinfeld works. That’s not how comedy works.
There’s also a comparison worth making directly. “Get a Mac” was effective not because it was funny, though it often was, but because every single spot made one clear argument. Sometimes the argument was about viruses. Sometimes it was about ease of use. Sometimes it was about compatibility with creative software. The humor was a wrapper for a message. Microsoft’s Seinfeld spots had the wrapper and no message.
Pulling the campaign two spots in generated its own news cycle, one that made the failure official and permanent. If you’re going to fail, failing quietly is better than failing with a press release.
The Results
The Seinfeld ads are now a standard business school example of celebrity-driven campaigns that forgot to have a point. The $300 million figure attached to the campaign refers to the entire marketing period, not just the two Seinfeld spots, but the number stuck to the story and made the failure seem more spectacular than it already was.
The “I’m a PC” campaign that replaced it performed well. It was picked up by marketing press as a genuine recovery and generated positive coverage. The lesson within Microsoft seemed to be absorbed: the subsequent campaign, “I’m a PC,” was everything the Seinfeld spots weren’t. Clear, warm, inclusive, and rooted in a specific counter-argument against Apple’s elitism positioning.
Windows’ market share didn’t suffer permanently. Microsoft is a technology company with distribution advantages that advertising can’t create or destroy. But the brand perception damage from Vista persisted longer than it should have, partly because the Seinfeld campaign consumed the time and budget that could have been spent on a coherent recovery narrative.
The Lesson for Today’s Marketers
The error here isn’t “don’t use celebrities” or “don’t be unconventional.” Crispin Porter built its reputation on unconventional thinking that worked. The error is using unconventional execution as a substitute for a clear creative idea rather than as an amplifier of one.
When a campaign brief says “we want to feel human” or “we want to start a conversation,” those are directions, not ideas. A creative team still has to find the specific, surprising, true thing about the brand that the execution will demonstrate. “Surprising and human” is what you want the audience to feel after watching the ad. It’s not the ad itself.
There’s a reason “Get a Mac” worked and this didn’t, and it has nothing to do with production values or celebrity caliber. Apple’s campaign had an argument. The argument was unfair in some ways, oversimplified in others, but it was an argument. Microsoft’s campaign had a vibe. Vibes don’t sell anything.
The other lesson is about creative courage. Unconventional work requires nerve at the agency level, but it also requires clarity at the client level about what a successful unconventional campaign looks like. When the client isn’t sure what success looks like, the agency can’t tell whether a bold idea is right or just bold. Someone in the approval chain needed to ask: “What is someone who watches this supposed to think or feel or do differently?” If that question doesn’t have an answer, the spots shouldn’t air.
Key Results
- Campaign duration: Two spots aired before the campaign was pulled
- Seinfeld fee: Reported $10 million
- Total campaign budget: Reported $300 million (entire campaign period, including replacement)
SWOT Analysis
| Strengths | Weaknesses | Opportunities | Threats |
|---|---|---|---|
|
|
|
|
Key Takeaway
Celebrity casting is a delivery mechanism, not a message. A $10 million star cannot compensate for the absence of something worth saying.
