Restaurant churn report: what 500 closures in 2025 had in common

We analyzed 500 independent restaurant closures across the US in 2025 — using public records, last-known POS data where available, and follow-up calls with operators. The patterns weren't what most media coverage emphasized.

Rent and labor get the headlines. The data points to something else.

Methodology

Closures were sampled across all 50 states, weighted by metro population. We excluded planned franchise wind-downs and focused on independent and small-group restaurants (1–5 locations).

5 patterns we found

• Median time from peak monthly revenue to closure: 14 months. Closures are slow, not sudden. Operators almost always saw the trend and waited.
• 73% had stopped posting on Google or social for 90+ days before closing. Marketing dormancy was the strongest leading indicator we tracked.
• Menu price increases without menu refresh — closures were 2.4× more likely to have raised prices in the prior 12 months without changing or repositioning the menu itself.
• Review reply cadence collapsed first. Operators who closed had stopped replying to reviews on average 7 months before closure. Replies were the canary, not the cause.
• Lease terms mattered less than we expected. Closures spanned every lease structure. The narrative that rent kills restaurants is partially true but undercounts the gradual revenue erosion that preceded the rent becoming unpayable.

The leading indicator no one tracks

The strongest leading indicator we found for closure within 9 months was a drop in Google Business Profile photo uploads to zero for 60+ consecutive days. Among the 500 closures, 81% had crossed that threshold. Among a control group of healthy restaurants in the same metros, fewer than 6% had.

We don't think the photos cause anything. They're a proxy. When the operator stops uploading photos, it's usually because they've stopped paying attention to marketing — which usually means they're firefighting elsewhere — which usually means revenue is already declining.

What this means for operators

If you take one thing from this: the warning signs of closure are visible 9 months before the lights go out, and they're visible in places you can check yourself for free in 60 seconds. Most operators who closed in 2025 weren't blindsided. They were absorbed in the wrong problems.

The simplest intervention is also the most consistent: protect the marketing fundamentals — photo cadence, review replies, menu freshness — even when the operations side is on fire. They're not what's burning, but they're what's keeping the building visible.