Could ridesharing shift to an ad-supported business model?
In a recent blog post, tech entrepreneur Andrew Chen discussed what a free, ad-supported service tier for Uber might look like. He argues that “free, ad-supported Uber rides are inevitable,” adding that, if Uber doesn’t do it, one of its competitors will.
It’s easy to see how Chen’s observations might be evidence that the Silicon Valley set is prone to overzealously apply the Internet company business model to everything. Namely, to offer a service for free that gathers commercially valuable data, and then make money from serving targeted advertising to a massive user base.
Unlike taxis, which currently have lots of ads and yet persist in their outsized expense, Uber’s ads would be not be delivered in a scatter-shot manner. Instead, they would be tailored to the passenger and based on their history, preferences and location. This makes them much more valuable to advertisers.
With more than a million rides per day, Uber has the requisite massive user base. In San Francisco, one of Uber’s most mature markets, its revenues are three times larger than the city’s taxi market ever was. And that’s not counting the city’s other ridesharing services, like Lyft and Sidecar.
Uber also collects massive amounts of information about its users. It’s not far-fetched to think that tech investors see the real value of Uber in the data that it collects, and see moving people as an ancillary function. What’s more, Uber characterizes itself as a “technology company” and not a transportation service. Certainly, its $41 billion valuation doesn’t seem to be supported by the raw numbers of the vehicle-for-hire market, even if one expect those markets to grow significantly.
While a totally ad-supported business model may work out great for Gmail, Facebook, Twitter and YouTube, these businesses are all highly scalable with relatively low labor costs and capital expenditures.
It’s not immediately clear that a similar model would work well for Uber, which has significant costs associated with scaling its service. Drivers have to earn enough to pay for gas and vehicle maintenance, while also earning a decent return on top of those expenses.
Let’s do some off-the-cuff math. Here’s a receipt
from Uber for a ride in DC. It was a little under 18 minutes, and cost $9.90.
If Uber used a simple model showing 30 second video ads, it could deliver about 36 ads during an 18 minute trip like mine. This is about $0.28 per view, which is a pretty high number.
The average CPM (cost per 1,000 views) for video ads on the Internet can vary widely based on a number of factors. Analysis from Credit Suisse puts the average at $24.60. No doubt, that cost can go higher for a quality, targeted audience with a high conversion rate (percentage of users that complete a desired action). But even if we put this at $50 CPM for 30 second ads, about the top of the current market, this is just $0.05 per view. Assuming a driver has zero down time (which is highly unlikely), this only adds up to about $6 per hour before expenses.
Chen suggests that Uber could demand CPMs as high as $100 (although he doesn’t get into the details of such a scenario, like how long each ad would be), based on the high cost of Snapchat’s advertising debut. There’s a good case to be made for why an Uber passenger would be a particularly valuable ad target, especially with all of the information Uber has about them — including their location, credit card and destination.
But even if you get $12 per hour with a $100 CPM, you still have to pay costs out of that amount, like Uber’s cut, and the drivers’ gas and vehicle maintenance. When Vox’s Tim Lee recently drove for Lyft, he came out with about $14 per hour, after accounting for gas and Lyft’s cut, and a significant amount of downtime. This is more or less consistent with Uber’s hourly wages in D.C., which it says are about $17 for UberX before expenses.
So, in spite of Chen’s optimistic assertion of the inevitability of free rides for all, it seems like the numbers fall just short of working, even with pretty optimistic assumptions, which do not even consider issues like mobile data costs and putting screens in every car. This is, of course, just looking at the United States. It may be viable in countries where labor, gas, and capital costs are lower.
Chen also suggests a few other variations on how ad-supported ridesharing could work, including app installs and business-to-business lead generation. Neither of those is particularly compelling as the sole basis for a broad, free service tier, but either certainly could help supplement revenue from targeted video ads.
Even if ads can’t fully support a free ridesharing service, they do have the potential to take a buck or two off each ride. That buck or two could be crucial, as transportation network companies grow more competitive and face new questions about the employment relationships with drivers, as well as new, more substantial, insurance requirements.
Indeed, one of the primary challenges Uber faces is finding ways to expand its market, which would both allow it to pass on positive network effects to users through lower prices and simultaneously increase its advantage over smaller competitors. Having a totally free or deeply discounted “loss-leading” service could potentially attract a huge number of new active users to its service. Uber has shown in the past it isn’t afraid of taking a loss to do this, and has plenty of cash with its latest funding round.
One thing that could make entirely free ridesharing work, however, is Google’s rumored plan to enter the market with autonomous vehicles. An autonomous vehicle without a driver, NHTSA Level Four, would seem to make the math work by removing the labor cost. Another major expense, the cost of insurance, also could decline in view of the meaningful increase in safety associated with the elimination of human error (so long as outdated insurance rate-making rules, like California’s Prop 103, can be overhauled).
Offering a free service at launch would also be a great way for Google to introduce this new technology and to overcome Uber’s dominant market position. For their part, Uber may already be anticipating this threat, having recently announced a partnership with Carnegie Mellon University to develop advanced technologies for ridesharing, including autonomous vehicles.
You may be asking yourself: ”This all sounds great! When can I expect the future of free driverless cars?” Well, unfortunately, government doesn’t move at the speed of innovation. Before autonomous vehicles can be embraced, they will have to clear regulatory hurdles at both the state and federal levels.
Where the government isn’t behind the curve of innovation, it is actively seeking to stymie its advance. In Sacramento, a bill has been introduced that would prevent ridesharing companies from sharing their data with third parties in the name of protecting consumer privacy.
While it’s well-intentioned, and there are some good ideas in the bill, overly broad privacy laws can also pose an existential threat to data-driven innovation. For instance, sharing some of this data would be essential for TNCs to work with the insurance industry or advertising companies to lower costs in the manner described above.
Whether free or otherwise reduced in price, ridesharing services and consumers stand to gain ground if they are willing to embrace the use of targeted advertising. Government regulators should get out of the way and embrace the future.
