Butter or sand in the gears? The question every founder must ask before choosing SF or Seattle

The City by the Bay may be considered the center of AI and technology, but that doesn’t mean every founder should flock there to set up shop, right? That’s the argument put forth by Yifan Zhang, AI2 Incubator’s co-managing director and creator of the AI House.

In a speech at last week’s Seattle AI Startup Summit, Zhang addressed the classic San Francisco-versus-Seattle debate. And while she heaped praise on San Francisco, she also highlighted the benefits of founders building right here in the Pacific Northwest. It was less about civic boosterism and more about founder diagnostics.

Zhang said the same qualities that make San Francisco great for some founders can work against others — particularly those building category-defining startups in the AI era.

Is your startup butter or does it have sand in the gears?

When weighing their decision to relocate to San Francisco or stay in the Pacific Northwest, Zhang proposed that entrepreneurs ask whether they’re building a “butter” or a “sand in the gears” startup.

As she explained it, “butter” startups are those that “succeed solely based on pure speed of execution, an extremely smooth customer experience, removing all friction for your users.” These are best suited for San Francisco. 

Alternatively, startups with “sand in the gears” have real-world complexity, hardware, human relationships, and regulatory edges, all of which Zhang believes may be considered flaws in the Northern California city. 

“The sand in the gears might give you some defensibility, a moat against the onslaught of competitors that are exactly the same,” she explained. “This is especially true in the AI era, when building has become so cheap. The people, the founders, willing to grind it out through these sand-in-the-gear startups, versus pivoting away from the things that are hard, will end up winning in these categories.”

Zhang used AI2 Incubator portfolio startup Friday Harbor as an example. Founded in 2023, the company faced a challenging technical problem: how to use AI to match borrower documentation against lender guidelines to determine who qualifies for a mortgage. Unfortunately, AI at the time wasn’t as great as it is today. There were shorter context windows and no agentic systems. Advanced reasoning models weren’t widely available.

“It’s also a hard problem when it comes to the customers you’re dealing with, a non-tech-savvy customer audience,” she said. Mortgage loan officers and originators tend to be skeptical of AI and distrustful of outsiders without industry experience.

But rather than throw in the towel and pivot, Friday Harbor chose to work through all the tough technical problems, staying focused on customer outcomes, and ultimately delivering mortgage underwriting that today benefits from the latest AI advancements. 

Zhang said that willingness to grind through hard technical problems is what sets Seattle apart from other tech-oriented U.S. cities like New York, Los Angeles, Austin, and Miami.

“We have a serious engineering culture here that’s heads and shoulders above every other city and goes toe-to-toe with San Francisco,” she said.

Building a startup in San Francisco vs. Seattle

Zhang knows what it’s like to build in both cities. She founded two startups — Gympack and Loftium — in San Francisco and Seattle, respectively. So how do the two cities compare for founders?

San Francisco’s biggest advantage is the sheer concentration of people singularly focused on founding and building great companies, Zhang said. Founders there absorb best practices faster than anywhere else, can draw from a talent pool that genuinely prefers startups over Big Tech, and get early access to cutting-edge technology.

That said, there are notable downsides, such as overwhelming pressure from investors. 

“You might be influenced to raise mega rounds when that’s actually pouring jet fuel in a plane when you’re learning how to fly,” Zhang said. “You may be influenced to pivot away from startup ideas that are good and are just a couple of tweaks away from being great.”

San Francisco can also be an echo chamber, Zhang said. When everyone shares the same startup-and-tech background, founders tend to hear the same feedback on repeat, whether or not it’s relevant to their company.

Seattle has its own advantages beyond engineering culture, Zhang said. Founders here are more willing to admit they have zero paying customers rather than ship a product that doesn’t work — a candor that San Francisco’s launch-at-all-costs culture tends to punish. She called Seattle’s humble approach a positive attribute, especially in an industry where the technology absolutely has to work.

Of course, Seattle isn’t perfect. With Amazon and Microsoft so prominent in the local tech scene, founders are more likely to get advice from people with Big Tech experience than from those who’ve actually built or funded early-stage companies. And the mentalities and best practices that worked at large corporations don’t necessarily translate to startups.

She also warned that Seattle can have a narrow view of who gets to call themselves a founder, often defaulting to people who’ve had senior or executive roles at Amazon or Microsoft. That mindset may work if you’re building a B2B SaaS company selling back to Big Tech, Zhang said, but otherwise it’s irrelevant.

Her advice: “Pay attention to the things that you uniquely bring to the table and build a company around that.”

In the end, Zhang offered this tip for founders: “Geography does impact your idea [and] your chances of success. So choose it as carefully and wisely as you choose your startup idea and industry.”