Meta's Muse targets the inertia that Stanford economists say roughly doubles subscription revenue

A seller whose revenue is doubled by inertia and friction collects about half of it from customers who never actively chose to stay ^[8][1]. The average US consumer spent $1,887 on subscriptions in 2025, according to Mastercard and FT Strategies ^[4], and half of that is about $940 a year ^[2]. The figure is illustrative. The doubling comes from the sellers in the Stanford paper, and the Mastercard number averages every kind of subscription.

Mahoney's four-to-one result ^[7] shows how an agent gets at that half. An assistant that lists each recurring charge forces the decision that inertia puts off, and Mahoney said AI agents could weaken both inertia and friction ^[9]. Subscription managers have offered the listing for years; Muse puts it inside a much broader personal assistant ^[1][3].

Those years of tools did not stop the category growing. Subscription spend rose 7.7% year on year in July, faster than overall card spending, according to Bank of America payments data, with entertainment and retail about 43% of the total ^[5][6]. ScribeUp, which builds subscription tools into banking apps ^[17], says its median user has more than 12 recurring payments and one in four has 20 or more ^[14]. The share of its users with at least eight rose from 62% to 71% in a year, a gain of 9 points ^[3], while members became 1.8 times as likely to start a cancellation ^[12]. Mackler said cancellations at a single merchant can jump as much as 50% when prices rise ^[15]. "More of life is becoming a recurring bill. Those bills cost more every year, and it becomes increasingly difficult for consumers to keep track of everything they are paying for," Jordan Mackler, the company's co-founder and chief executive, said ^[13].

The Muse effect could run three ways. It could concentrate in digital services such as credit monitoring, which Mahoney said can keep billing long after a consumer stops thinking about it, while pet food arriving at the door stays hard to forget ^[10]. It could lift cancellations while new sign-ups keep outpacing them, the pattern already in ScribeUp's numbers ^[12][14]. Or the larger sums could move in bank accounts. I think the first is the likeliest, because the evidence places the forgettable revenue in digital services. Mackler said the rise in cancellations predates ScribeUp's newer AI features ^[12], so nothing in the record yet ties a change in spending to agents. If Bank of America's subscription growth holds at 7.7% or above in the months after Muse's September rollout ^[1][5], the agent is not cutting into the base.

The bank version comes from Apollo. "Muse and similar agentic AI assistants could soon sweep household cash automatically into accounts paying 3.3% to 5.0%, instead of the 0.1% national average on checking accounts. If every household used AI agents to optimize the return on their cash balances, banks could lose a large share of the cheap deposits they rely on to make loans, which would be a problem for the entire financial system," Torsten Slok, Apollo's chief economist, wrote last week ^[11]. His range puts the average checking account 3.2 to 4.9 percentage points below the alternatives ^[4].