The unicorn celebrates the wrong task

An agency team I know ran a rebrand in Asana, and the day they closed the last task in the plan, the tool fired a small celebration: a creature, a unicorn if I remember right, sailed across the screen while the account lead clapped at her desk and two designers stood up. It was a nice moment. It was also two weeks after the rebrand had missed its own launch date, and the reason it missed was a client approval that had sat unanswered through three meetings, none of which had ever been a task, because approvals that happen in rooms do not get rows. The unicorn did not know that. The unicorn celebrated a checkbox, and the checkbox was downstream of everything that had gone wrong, and I sat there thinking that the creature was not a frivolity at all. It was the product's thesis, rendered as a mascot.

The thesis is that completing things is the point, and the celebration is aimed at the exact moment work becomes visible. The trouble is that the invisible part is where projects live and die, and that part has no mascot.

The company diagnosed the disease, then billed per treatment

Start with the diagnosis, because Asana wrote it themselves. Since 2019 the company has published an annual Anatomy of Work survey, ten thousand knowledge workers, run and funded by Asana, and its headline finding, repeated every year and covered by Bloomberg among others, is that a majority of the workday goes to what they call work about work: status chasing, handoffs, updating the thing that tracks the thing. Treat the finding carefully, because it is vendor research with a built-in loop: a company that sells work management has defined most of your day as management of work, and its instrument has no category for the thought itself. Still, the direction is credible to anyone who has watched a campaign team for a week, and the honest reading is that coordination has metastasized and everyone knows it. The incentive is what makes it interesting. Asana is paid per seat to manage that coordination, monthly, per person, forever, with compliance, permissions, and timesheets sold as separate add-ons, and since late 2024 it has layered on AI Studio and AI Teammates that automate the very statuses the survey mourned. The new meter prices an AI request at fifty cents prepaid or sixty cents on demand, charged only when the work completes. The company found the disease, named it, and now sells the treatment by the dose.

That is not a scandal. It is the clearest business model in the category, and I want to give it its due: most vendors hide this loop inside a demo. Asana published the loop as a report.

The serenity has a balance sheet

The name is a yoga pose, a settled posture, and for most of its life the company performed the pose: founded in 2008 by Facebook's Dustin Moskovitz and Justin Rosenstein, launched commercially in 2012, taken public by direct listing in 2020 at a reported valuation of five and a half billion, praised by PCMag as an editor's pick across three separate years. Moskovitz, Facebook's cofounder, owns about half the company personally and has been able to fund patience out of that stake. The pose did not hold. The fiscal 2025 filings show revenue of $724 million against a net loss of $256 million, and in March 2025 Moskovitz announced he was stepping down as chief executive and the stock lost a quarter of its value in a day, which tells you the market's view of the brand without him. The board replaced him with Dan Rogers, a former ServiceNow and LaunchDarkly executive, an enterprise-growth hire by any honest reading, and in May 2026 the company paid seventy-five million dollars for StackAI, a no-code agent builder, folding agentic workflows into the roadmap. None of that is philosophy. All of it is the arithmetic of a per-seat business that has stopped compounding seats and needs a new numerator. Who benefits if you believe the demo? The demo is good. The meter is new, and the meter is the growth story the investors were promised.

I am not cynical about the pivot; I am suspicious of its timing, which is the same feeling with better manners.

Coordination is the work, so managing it is not the disease

The counterargument deserves a full paragraph, because I think it is mostly right. The work-about-work framing is lazy in one direction and true in another: in an organization spread across time zones, the handoff, the status, and the approval are not overhead on the work, they are the form the work takes, and pretending otherwise is how distributed teams rot. Asana is pleasant in a way that matters operationally. People actually update it, which is the entire value of a tracker; Jira loses to any tool its users abandon, and this one gets updated because the forms are clean and the approvals have a trail. The AI request pricing is more transparent than most of the industry: no charge for failed or timed-out tasks, a published rate, alerts at sixty-five, eighty-five, and one hundred percent of the allotment, and a spending cap an admin can set without a sales call. If an agent is going to chase statuses anyway, this is a defensible way to sell it. And the unicorn, I will admit, is not nothing. Morale is an input, teams that celebrate small completions ship more often, and a mascot is cheaper than a consultant.

Both things are true. The celebration is real, and it is aimed one step downstream of the failure.

It depends, then, on whether your coordination is a symptom or the substance. If your team's actual product is coordinated deliverables, Asana is among the best tools on earth for it and the meter is honest. If your team ships a thing, and the coordination grew because the org grew, the meter will happily institutionalize the growth.

The variable worth watching is what happens to the request meter as the new chief executive reports his first full years. A seat is a promise you renew. A request is a promise you interrogate. The unicorn will still fly when an agent closes the task. Nobody will be at their desk to see it.