Big is bad, small is beautiful. Urban planning learned this lesson the hard way more than a half-century ago. The tech industry is relearning it today as it ventures into city-building.

Consider the most exciting new places for urban innovation. Station F in Paris, a massive tech hub, aims to gather 1,000 startups under one roof. New Lab in Brooklyn turned an old shipbuilding complex into an idea factory for small firms and professional networks. These places achieve the scale of the old corporate towers and technology parks. But instead of clear-cutting obsolete buildings to make way for monotonous walls of glass, they re-purpose history’s shells to provide an artificial reef for lots of little efforts to take root.

Google, Facebook, and Amazon want to be here. But they’re kept at a safe distance for good reason. They’d ruin the party. Silicon Valley works on a myth that originated with Xerox PARC in the 1960s - that the R&D largesse of big corporate anchors produces knowledge spillovers enough for the whole region to feed from. But today’s tech giants don’t work this way. They don’t disperse ideas and talent - they suck them in like black holes and lock them down with non-competes and predatory patents. In fact, it is their skill at stopping the leakage of ideas and talent that has powered their success.

What’s worse, when tech giants come to cities, they often destroy the naturally occurring habitat of the creative types they so dearly seek. They snap up and snuff out promising homegrown startups, take talent off the market, and drive up the cost of office space and housing.

Economic development officials throughout North America haven’t adjusted to these tactics, or even recognized just how much more these companies take than they give. They still compete to offer up steep discounts, tax breaks, subsidies, and dossiers of data about local talent. All in the cause of landing a deal with a big tech company. Amazon’s HQ2 search didn’t just fuel these insecurities, it preyed on them.

But the HQ2 fiasco asked the wrong question—do you have what it takes to be a tech headquarter? Instead, cities should ask themselves the same question that early stage tech investors have honed in on over the last decade--how do we create the conditions to increase the number and quality of startups?

It’s a challenging shift to make. Moving away from big deals and mega-projects requires economic development policy teams to think less like investment bankers and more like angel investors: increase deal flow while radically lowering transaction costs. And they’ll have to find ways to communicate the collective impact of lots of little wins with the same vigour of today’s big-deal press conferences. Cities that figure out how to do this effectively will turn the tables on big tech companies. They’ll hitch their future to a more resilient ecosystem of small and medium-sized firms, instead of one big footloose corporation that doesn’t share their interests or destiny. And they can start setting the terms for vetting companies, instead of the other way around.