As the wireless industry gathers in Los Angeles this week for MWC Los Angeles 2019 many attendees are going to have a sense of deja vu. 5G will once again be the primary topic of discussion, but most vendors and operators are still struggling to identify how they’re going to recoup the costs of deploying 5G networks.

“Most of the use cases we’ve heard don’t make sense, yet the capital requirements for densifying networks are very real,” analysts at MoffettNathanson wrote in a new report about the state of 5G in the U.S. market.

Data consumption and the revenue operators earn per wireless device began diverging a decade ago. Each wireless device consumed, on average, 8,863% more data in 2018 than it did in 2009, according to data from the Cellular Telecommunications Industry Association. During that same period, average revenue per device declined 12.6%.

Simply put, network operating costs and data consumption are going up while revenues continue to slide. MoffettNathanson concludes that the “if you build it they will come” strategy is part of every telecom network upgrade cycle, but 5G presents some unique challenges. Operators will have to identify better, more realistic use cases before the heavy lifting of network deployment and densification can get done, the firm wrote.

“Driverless cars are on every list of 5G use cases. But would anyone really make steering, accelerating, and braking, the most mission-critical functions of a driverless car, dependent on ubiquitous network connectivity,” MoffettNathanson analysts wrote in the report.

The firm went further with respect to industrial IoT, arguing that while factories of the future will be more broadly connected they’re also likely to operate on private networks, potentially cutting out the carriers in the process. Likewise for smart cities — the opportunity and benefits are clear, but “how many cities have the money?”

MoffettNathanson points out that 5G is about more than faster speeds and greater amounts of data. The optimistic view is that utilities, vehicles, and factories are ripe for a new era of connectivity and that low-latency applications will drive new opportunities in virtual reality and user new business models that don’t exist today.

However, according to the firm, there’s plenty of reason to be skeptical about each of those applications. Replacing wired broadband may be the only opportunity that “targets a tangible revenue opportunity that actually exists already” and moreover, according to the firm: “there are better, or cheaper, ways to deliver on each of the identified opportunities.”

The primary challenge that network operators and some vendors will be confronting again this week at MWC Los Angeles is that “some of the business opportunities identified for 5G wireless can, in fact, be met more cheaply or expeditiously by wired networks,” analysts at MoffettNathanson wrote. “Others can be served by new competitors with alternate networks.”

It’s a “classic chicken-and-egg scenario,” according to the firm. “Those applications (and the revenues they bring) won’t fully develop until the network is fully built. But fully building a network before there are revenue generating use cases is too risky.”

The analysts conclude that those use cases will eventually materialize but deploying a dense and ubiquitous 5G network with a robust 5G ecosystem of use cases will take longer than most people expect.