My hunch is that savvy investors will make sure those tokens aren’t being spent on short-sighted window dressing, and are instead actually laying track for break-out performance and optimized operations.
Spending on AI is easy to announce and hard to audit. Any company can put a token bill in a press release. The question a serious investor will ask is what the tokens bought — fewer hours per task, lower cost per transaction — and whether that shows up in the operating results rather than in the narrative.
There’s already a live example of what that kind of reporting looks like. Klarna has taken to publishing its revenue per employee alongside its AI story. Last May it told investors the figure was approaching $1 million, up from $575,000 a year earlier, crediting AI for letting it curtail hiring and end its Salesforce CRM contract. In its fourth-quarter release this February it reported revenue per employee of $1.24 million, up 3.6 times since 2022, with headcount down 49% while revenue grew 104% and operating expenses fell 8%. Whatever you make of the company, that’s an efficiency ratio reported quarter after quarter, and the market can watch its slope.
Someone probably comes up with token/employee/EBITDA ratios and looks to see them improving quarter to quarter. The level of such a ratio will mean little on its own — every industry will have its own normal. The trend is what will matter. Investors will want to see whether the extra token spending improves results.
