All in all, it was the kind of quarter investors were looking for at Roku (ROKU -10.29%). The streaming service and device maker beat analysts' top- and bottom-line estimates, sending the stock higher following the release of its second-quarter results last Thursday. March's launch of the company's own branded television sets appears to have been well-received, too. 

When you take a deeper dive into the numbers, though, old concerns are rekindled. Namely, the company is still losing money despite its expanding size. It's not clear when -- or even if -- sales growth will eclipse Roku's costs.

The rest of Q2 numbers aren't quite as thrilling

Roku lost $107.6 million ($0.76 per share) on sales of $847.2 during the three-month stretch ended in June. All three figures are measurable improvements on year-ago results. Most of the stock's gain since releasing last quarter's numbers, however, is the result of topping revenue estimates of $774.5 million and a projected per-share loss of $1.26. Guidance for the quarter now underway was better than expected as well.

The company showed other signs of forward progress, too. There are now 73.5 million actively used Roku accounts, up nearly 2 million from Q1's final count. These people collectively watched a record-tying 25.1 billion hours of video on their Roku devices during the second quarter.

There's just something not quite right about this growth, though -- namely, it's increasingly expensive. It's also slowing down before Roku has even gotten out of the red and permanently in the black.

The graphic below puts things in perspective. While last quarter's top line was healthy, the figure only extends the quarterly revenue range seen since late 2021. One might have expected more in light of the organization's own television debut.

Chart comparing Roku's stagnant revenue to continues operating and EBITDA losses.

Data source: Roku. Chart by author. Figures are in millions of dollars.

Yes, earnings before interest, taxes, depreciation, and amortization (EBITDA) and operating losses are narrowing. They're still losses though, and it should be noted that profit progress is purely the result of cost-cutting. Roku dialed back its research and development, sales and marketing, and administrative expenses from the fourth quarter's peak levels. 

Chart comparing Roku's revenue to shrinking operating expenses.

Data source: Roku. Chart by author. Figures are in millions of dollars.

Cost-culling is beneficial as long as it doesn't crimp a company's ability to do (or grow) business. It remains to be seen if that's going to be the eventual outcome here.

Perhaps the biggest concern surrounding Roku's Q2 metrics, however, is the deteriorating value of Roku devices' regular users. The company added another 1.9 million accounts to its total customer headcount. But the total number of hours these people watched didn't budge. Neither did the average amount of revenue they produced following the clear contraction of the average revenue per user (ARPU) figure during the fourth quarter of last year and first quarter of this year.

Data source: Roku. Chart by author. Account data is in millions. Streaming hours data is in billions.

Chart comparing Roku's active accounts to ARPU and streamed hours.

On a per-user basis, Roku users are actually watching less TV.

Then there's the mostly overlooked detail about Roku's published ARPU calculation. That is, it includes sales of devices like TVs or streaming players. If you take those sales out of the equation and only consider "platform" (or advertising and promotion) revenue in the calculation, the ARPU figure actually peaked back in late 2021. That's not encouraging given the company's business model, which is monetizing users as an advertising medium rather than turning a profit on TVs and streaming players themselves.

Still plenty of risk on the table

Doomed? No, that's overstating the risk here. Roku could conceivably find a way to extract enough per-user revenue in the future to work its way back to net profitability. Indeed, the recent spending cuts don't appear to be adversely impacting the business -- at least, not yet.

Too many of the numbers from the second-quarter report raise red flags, though. Why is the number of streamed hours stagnating when the number of active accounts is still growing? Why isn't the ARPU figure recovering from its recent setback? How long can growth be maintained when marketing and selling expenditures are contracting?

External headwinds don't help either. These include the fact that 83% of U.S. households already own a smart TV, or a streaming media player, or both (according to Hub Research). Most homes own more than one television set, and as such may need more than one Roku-powered device. Yet the market's already pretty well saturated -- not to mention crowded with competition -- limiting Roku's potential for further growth.

Similar headwinds are blowing overseas where Roku has even less of a foothold.

The point is, Roku isn't out of the start-up woods yet. It's seemingly moving in the right direction given last quarter's numbers. Those results are far from assuring, however, that Roku will swing to a respectable profit soon enough to justify the stock's lofty price now. Analysts aren't even looking for an actual profit until 2027. Much can happen in the meantime to derail this still-fragile company.