"Trust us and we'll make you a lot of money." It's the simplest pitch there is, and a surprisingly effective one, under the right circumstances. It works equally well with investors, employees, and clients. 

It has to, in fact, because the " Trust us" approach involves a sort of confidence flywheel. If any one of the individual components stops spinning, or even slows down a little, the gears quickly threaten to fly off.

At this time last year, the flywheel was functioning perfectly for just about all of the billion-dollar startups known (thanks to venture capitalist Aileen Lee) as unicorns. But as VCs and other investors came to recognize the outlines of a private tech bubble, they began questioning the math behind those 10- and 11-figure valuations. Cue flying gears. 

A deep dive by BuzzFeed's William Alden into the inner workings of Palantir shows what can happen when the flywheel starts to wobble. A seller of big data analytics, Palantir is the third largest private U.S. startup by valuation, after Uber and Airbnb. A key difference: Where those two derive their revenue from zillions of smallish transactions with consumers, Palantir relies on a relatively small number of big contracts with government and corporate clients, that--in theory, at least--pay millions of dollars a year for help with marketing, cybersecurity, and supply-chain efficiency. 

Here's how the flywheel had been working at Palantir: Its ever-growing valuation, pushed to $20 billion in the last round of funding, allowed the company to hire top-tier engineering talent for considerably below-market rates by making up the difference in stock. Meanwhile, a $2.5 billion war chest enabled Palantir to offer its services to clients like Coca-Cola and American Express on generous terms, "bookings" whose massive notional value was largely conditional on performance or deferred into the future. Steady growth in those bookings then allowed Palantir to raise yet more capital at a still higher valuation. 

As a closed system operating in a vacuum, it functioned perfectly--until outside forces intruded, in the form of investors spooked by the bubble talk, demanding to know "how long until bookings translate to cash," in the words of a Palantir financial analyst quoted by Alden. Pretty long, is the answer; while Palantir claimed $1.7 billion in bookings for 2015, only $420 million of that was actual revenue. 

The obvious solution is to squeeze clients to extract more of that cash they're ostensibly committed to spending. The problem is, BuzzFeed found, when Palantir has tried to do this, Coke, AmEx and Nasdaq all declined to extend or expand their deals, saying they weren't getting enough value to justify the high price tag.

Just as investors started to fret about ever realizing the value of their shares, employees were feeling the same way. Like other late-stage startups whose valuation has outpaced its revenue growth, Palantir is ill-positioned to go public anytime soon and has shown little inclination to do so. With the prospect of a big windfall looking more remote, employees who had accepted below-market salaries reevaluated the tradeoff, leading to a wave of departures in early 2016. To stem the tide, CEO Alex Karp increased salaries across the board by 20 percent. (In a Quora post, Palantir co-founder Joe Lonsdale says the spike in departures was the result of a cohort of employees hitting a vesting "cliff" that allowed them to leave without sacrificing the equity they'd accrued.)

In response to similar forces, other companies are tightening their belts, hoping a new focus on bottom-line economics will persuade investors, private or public, to give them the benefit of the doubt. Dropbox, for instance, has turned a gimlet eye on the $38 million it spends annually on employee perks, ending its local shuttle service and paring back meal hours. 

Palantir executives have suggested they could do the same thing if needed. At an internal presentation in February, another financial analyst said the company could "be profitable right now and turn off the growth engine." Indeed, the company told Fortune in March that it was projecting profitability in 2017.

To climb out from deep in the red while giving everyone a 20 percent raise and losing big clients--it doesn't take big data analytics to see that's a tall order. In the current atmosphere, it's going to take more than "Trust us" to convince anyone--and suddenly that flywheel is wobbling everywhere you look.

Published on: May 10, 2016