Growth Before Monetization: Meta’s “Muse Network” Bet and 10 Million Users

Meta kept Muse free up front to build a "Muse network", reached more than 10 million users in about six weeks, and still says it will monetize sooner than its usual playbook. Here is when founders should copy growth-first, and when compute costs say no.

Meta has a famous habit: build the audience, then sell to it. On its Oct. 21, 2026 earnings call, Mark Zuckerberg explained that Muse is free up front because “we want to establish this Muse network,” and in the same answer said Meta would “work on monetization somewhat sooner on this than our normal playbook.” That tension is the whole lesson. Growth before monetization still works for Meta, but an AI agent’s compute bill shortens how long even Meta is willing to wait.

Where it stands: Zuckerberg said Muse passed 10 million set-ups across the two launch countries, the US and Canada, within roughly its first month and a half. Meta’s CFO said revenue from new AI businesses can begin “contributing incremental revenue in 2027.” Meta has not disclosed any Muse revenue or cost figures.

What is Meta’s growth before monetization playbook for Muse?

Give the product away to most people, build a network of users and businesses around it, and charge once both sides are on board. Meta’s launch post said Muse is “free for most of what people need, with subscription plans for people who want to do more.”

Zuckerberg’s reasoning on the call was explicit: “Part of the reason why we are not, you know, just charging everyone up front to use this is we want to establish this Muse network.” A network here means the web of consumers, merchants and partner services the agent connects. The more people use it, the more businesses want to plug in, and the more useful it becomes.

That is the classic network-effect argument behind free consumer products, and Zuckerberg’s own reference to Meta’s “normal playbook” signals how often the company has used it.

Why is Meta monetizing Muse sooner than usual?

Because every free user costs compute, which a social feed mostly did not. Zuckerberg said Meta will “work on monetization somewhat sooner on this than our normal playbook.” UBS, in a note summarized by Yahoo Finance, put it in context: “Meta intends to monetize Muse faster than its historical four-year product development cycle.”

CFO Susan Li explained the cost side: “Muse’s computing intensity means that expanding its user base will require significant investment, and we expect total serving costs to rise.” In other words, growth itself raises the bill. On a social app, an extra user mostly adds server load for feeds and messages. On Muse, an extra user can mean another agent running tasks in the cloud.

Zuckerberg still leaned on the core business to bridge the gap: “In the near term, the strength of our core business enables us to scale the Muse network before then.”

Signal Figure Source
Muse set-ups reported Over 10 million Zuckerberg, Oct. 21 call
Daily use among them About 25% Zuckerberg, Oct. 21 call
Meta ad spend on Muse, October 39% (estimate) Sensor Tower via TechCrunch
U.S. 7-day retention 30% (estimate) Sensor Tower via TechCrunch

How did Meta buy Muse’s early growth?

Partly with its own advertising. According to Sensor Tower data reported by TechCrunch, “Meta has allocated 39% of its total advertising spend to market Muse in October, up from 23% last month.” 9to5Mac, also citing Sensor Tower, reported that Muse took “up to 50% of daily house impressions” during part of September. House ads are promotions a company runs for its own products on its own apps.

These figures are third-party estimates. PPC Land, reviewing the same data, stressed that “None of the figures come from Meta. Meta has not published download, user or retention numbers for Muse.”

For founders, the takeaway is that growth was not free. Meta paid for it twice: once in compute for every new agent, and again in ad inventory it could otherwise have sold. A startup that copies the free-first model without counting acquisition spend is only seeing half the bill.

Is the growth sticking?

The early signs are mixed, and that is the point founders should watch. Zuckerberg said “about 25% of people who have set up Muse use it on any given day.” Sensor Tower’s estimate for U.S. retention was “30% retention rate after 7 days and a 23% retention rate after 14 days,” per TechCrunch.

CNBC framed the challenge in early October: “the bigger challenge for the Facebook parent is to get people to keep using Muse after the magic wears off.” Analysts were cautious too. Before the results, Morgan Stanley analysts urged investors to “keep reasonable adoption rates in perspective,” and Bank of America warned that Muse “will consume capacity w/o near-term revs,” both as quoted by CNBC.

Growth before monetization only pays off if the users who arrive for free stay long enough to be monetized later.

When should a founder follow the growth-first playbook?

When three conditions hold. First, each additional user makes the product better for others, so growth compounds. Second, the marginal cost of a free user is small or tightly capped. Third, you have funding to cover the gap until revenue starts.

By Meta’s own account, Muse meets the first, and its ad business covers the third. It does not fully meet the second, which is why it added paid tiers at launch and plans a commerce fee. PYMNTS described the strain in a September report, saying the plan leaves “the $100 plan and merchant commissions to carry a product whose costs rise with every download the charts are celebrating.”

Most AI startups will meet the first condition at best. If your free users each trigger meaningful inference, cap the free tier, ask for a card early, or charge from day one. Our piece on freemium for compute-heavy products covers those guardrails, and our take-rate versus subscription guide covers how to charge once you do.

What should founders avoid copying?

Avoid copying the timeline. Meta’s normal habit of waiting years to monetize is something it is already shortening for Muse. A startup with a fraction of the resources should assume a shorter runway, not a longer one.

Avoid copying the ad push without a budget. Meta could redirect its own inventory; you would pay market rates.

And avoid treating downloads as the goal. Sensor Tower’s estimates, Meta’s 10 million set-ups and the retention numbers tell different stories. Pick the metric that predicts revenue, whether daily use, transactions or paid conversions, and grow that. For a primer on pricing models generally, see what freemium pricing is.

Meta’s own target date for new AI revenue, per Li, is that those businesses “can begin contributing incremental revenue in 2027.”

Frequently Asked Questions

What does growth before monetization mean?

It means building a large user base first and charging later, usually through ads or fees once the network is big. Meta has run that playbook for years, and Zuckerberg said Muse is free up front partly to establish a Muse network, though Meta plans to monetize it sooner than usual.

How many people use Meta Muse?

On the Oct. 21, 2026 call, Zuckerberg put Muse set-ups above 10 million across its two launch markets, the US and Canada, after roughly six weeks, and said about 25% of them use it on any given day.

How much is Meta spending to promote Muse?

Meta has not disclosed a figure. Sensor Tower estimated, as reported by TechCrunch, that Meta put 39% of its total advertising spend toward marketing Muse in October, up from 23% the month before. These are third-party estimates.

Should a startup give its AI product away to grow first?

Only if it can fund the compute until revenue arrives. Meta funds Muse from its core advertising business, and its CFO expects serving costs to rise as users grow. Most startups have to charge earlier or cap free usage tightly.

Sources

This article is general information for founders and is not accounting, legal or investment advice. Figures are as reported by the sources cited on the dates shown; analyst and third-party estimates are not company disclosures. Found Founders is not affiliated with Meta.