Ask people to name a ridesharing company and most would probably mention Uber. But there’s also Lyft, the key word being “also.” It’s a distant second in the U.S.

But Lyft has been making headway by capitalizing on Uber’s self-inflicted wounds. The number two in ride-sharing had made hay on social media and in PR. And now Lyft just made its biggest score off Uber to date.

Waymo, Google’s sibling self-driving-car technology company (both belong to parent Alphabet), has inked a deal with Lyft to develop and test autonomous-car technology. This is a major threat to Uber and a likely result of its legal battle with Waymo.

Ride-sharing as a business category has been hot, with most of the activity concentrated in Uber and its roughly $68 billion market valuation. In comparison, Lyft is valued at $5.5 billion. However, both companies have a basic ongoing economic problem. The operational realities of costs versus revenue are ugly.

Naked Capitalism ran a lengthy analysis on Uber last fall, although the criticisms can be generalized to the rideshare business model. Because companies use drivers who are treated as contractors, much of the gross revenue has to go to those individuals to cover vehicle costs, fuel, insurance, maintenance, and credit card and license fees before talking about pay for the drivers (and benefits for those who try to do this full-time).

As shown in Exhibit 2, for the year ending September 2015, Uber had GAAP losses of $2 billion on revenue of $1.4 billion, a negative 143% profit margin. Thus Uber’s current operations depend on $2 billion in subsidies, funded out of the $13 billion in cash its investors have provided.

Uber passengers were paying only 41% of the actual cost of their trips; Uber was using these massive subsidies to undercut the fares and provide more capacity than the competitors who had to cover 100% of their costs out of passenger fares.

The losses have only continued, at a reported $2.8 billion in fiscal year 2016. The company says revenue growth is significantly outpacing losses, but offered no expected break-even date. Also, the “loss statement doesn’t account for employee stock compensation, certain real-estate investments, automobile purchases and other expenses,” according to Bloomberg.

Rideshare businesses can’t easily scale because the major operational costs outside of pay — fuel, insurance, fees, and all automobile costs — are borne on a per-driver basis. The companies cannot get too involved by providing mass purchasing or they stand to lose the huge tax and regulatory advantages of working with contractors (like not having to pay part of FICA, unemployment insurance, and workers compensation). As Huber Horn, author of the Naked Capitalism articles, wrote, “Drivers, vehicles and fuel account for 85% of urban car service costs.” There are few economies to obtain.

Rideshare companies have two possibilities moving forward. One is massive dominance and the ability to then sharply raise prices and retain more of the gross. The billions in investment Uber has received is predicated on this strategy. But a sharp fare increase could be difficult to impose, sending customers running back to taxis.

The other possibility is to eliminate drivers to the greatest degree possible. Self-driving technology could, in theory, allow the companies to own and operate autonomous cars, gain the advantages of purchases in scale, and move toward profitability. Uber and Lyft have both been public about interest in developing a driverless fleet. In fact, Uber started its own division and hired a key employee out of Waymo, which then sued Uber in an action that looks as though it’s headed to trial.

Getting into a legal battle with a company whose parent has the massive resources of Google’s cash machine is like seeing a Braveheart square off against a battalion of special forces soldiers. The guy might win, but it’s going to be tough going and distracting, to say the least.

The distraction is only the small part of the benefit Lyft has received. The big one is the partnership with Waymo. The technical and financial resources there are deep, the expertise in self-driving vehicles is the longest established in the industry, and the desire to beat Uber, which once had a partnership with Waymo, seems significant. Given that Waymo apparently thinks its technology is near ready for commercial use, this could give Lyft the help it needs to rise over its rival. Lyft is also working with General Motors, one of its major investors, on self-driving vehicles. How many other companies have the manufacturing skills and ability to crank out large volumes of autonomous cars?

The lesson this offers is that it’s important to distinguish between strategy and tactics. Uber has often seemed to focus heavily on tactics — throwing elbows, pushing as hard as possible for what it wants immediately. But today’s smart tactics can be tomorrow’s foolish strategy. Pushing Google into the position of an adversary is risky, even if it is to make an initial gain. Particularly if you are losing more money a year than the vast majority of companies ever get as investment. Eventually the investors get tired and walk away. What happens if you don’t have the necessary cash flow at that point?