AT&T’s technical team likes to talk about the company’s 5G fixed wireless trials and what the company has learned about millimeter wave (mmWave) deployments using the 28 GHz and 39 GHz spectrum bands.

At the recent 5G New Horizons Summit in Austin, Texas, Dave Wolter, assistant vice president of radio technology and strategy at AT&T Labs, told the audience that the company experienced speeds in excess of 1 GB/s with latency less than 10 milliseconds. And at the Brooklyn 5G Summit in April, Melissa Arnoldi, president of technology and operations at AT&T, talked about how the mmWave signal had very little impact from elements like snow or rain. “We learned that mmWave signals can penetrate foliage and glass and walls better than anticipated,” she said, adding that the company could deliver up to 1 gigabit of speed at 900 feet from the antenna.

But AT&T CFO John Stephens cast doubt on the viability of a commercial deployment of 5G fixed wireless earlier today by telling financial analysts at the Cowen and Company Technology, Media & Telecom Conference that it isn’t cost efficient to deploy the technology and offer service to residential customers. “There are challenges for us but it’s not the network. It’s the cost efficiency,” he said, noting that it will take a lot of fiber or small cells to backhaul the traffic from the mmWave antenna to the cell site. And fiber and small cells are expensive to deploy. “For a residential broadband solution, the economics don’t make sense,” he added.

Stephens, however, did hint that there might be industrial use cases for 5G fixed wireless, such as in manufacturing plants where businesses need to connect a lot of machines back to a central point.

Nevertheless, AT&T has said it will launch 5G in 12 markets by year-end, but that will be a mobile 5G service based on the 3GPP non-standalone 5G New Radio (NR) standard. And it has said it will be the first to have 5G, a claim that Verizon has also made. “We are leading the U.S. in 5G,” Stephens said. “Our country is in good shape with our ability to lead 5G and have it up by end of year.”

There seems to be a lot of doubletalk in this message, something ACG Research analyst Chris Nicoll questions. “It seems like they are talking out of both sides of their mouths,” Nicoll said when asked about AT&T’s stance on 5G fixed wireless. “I think it shows just how fractured 5G is.”

Fractured indeed. Just compare Stephen’s comments with that of Verizon, which says that it will make 5G fixed wireless available in three to five markets by year-end. The company has announced two of those markets — Los Angeles and Sacramento, California — but so far has kept the others under wraps.

And unlike AT&T, Verizon has said its 5G fixed wireless deployment won’t add a lot to the company’s CapEx budget. At an investor call last month, Verizon CFO Matt Ellis told investors that the company does not expect to see a big acceleration in CapEx to fund the 5G rollout.  Instead, he touted the expansion of the company’s software-defined networking (SDN) and its edge networking (which it calls “Intelligent Edge”) as well as its densification of the existing 4G LTE network.

The stark difference between Verizon and AT&T makes one wonder if AT&T is just downplaying the 5G fixed wireless opportunity because it wants to protect its DirecTV investment. In other words, why spend money to deploy a 5G fixed wireless service to residences in hopes of offering a residential video and broadband service when you already have a satellite pay-TV service that is struggling to keep customers.

The company purchased DirecTV in 2015 for $48.5 billion and was committed to turning it into the country’s largest pay-TV provider. But instead AT&T been fighting to keep its customers. It lost 188,000 DirecTV customers in the first quarter of the year and now has 20.3 million DirecTV subscribers. The company has been adding customers to its streaming service, called DirecTV Now, but that service is not as profitable.

With six months left in the year to claim that 5G first-mover advantage, AT&T has not yet said how it plans to market its mobile 5G service nor has it revealed any of its launch cities. Does that first-mover advantage really matter? Not when the 5G ecosystem is still so fragmented.

“What really constitutes a deployment?” Nicoll asked. “Is four cities using a pre-commercial standard a deployment? That’s almost just dipping the toe in the water.”