Businesses embraced AI in review replies faster than customers, study finds

A Revfer analysis of about 88,000 Google reviews across 400 Greater Boston and MetroWest businesses found AI-like writing markers rose much faster in owner replies than in customer reviews. The study also found reviews became more positive overall, with 5-star ratings climbing to 86% by 2026.

Why it matters: - Local businesses appear to be adopting AI writing tools faster in public replies than customers are in the reviews themselves. - The shift could affect how consumers judge authenticity, especially as more reviews trend positive and negative feedback stands out more. - Revfer says the findings matter because reply tone can shape how businesses respond to customer sentiment.

What happened: - Revfer analyzed the public Google review histories of a random sample of 400 businesses in Greater Boston and MetroWest. - The sample included 100 hair salons, 100 restaurants, 100 dentists and 100 auto repair shops. - The dataset covered about 88,000 reviews written between 2019 and 2026. - The research found AI-like writing markers increased far more in owner replies than in customer reviews after generative AI tools arrived in late 2022.

The details: - The em dash appeared in 0.2% of reviews in 2019 and 2.9% in 2026. - In owner replies, the em dash rose from 0.1% in 2019 to 4.2% in 2026. - Chatbot-associated words such as “delve” and “testament to” remained almost absent from reviews. - The typical review length stayed around 140 characters since 2021. - People who have written 100 or more Google reviews now write reviews about 60% longer than before ChatGPT's release. - People who have written only one Google review now write shorter reviews than before. - The share of 5-star reviews rose from 69% in 2019 to 86% in 2026. - The average rating increased from 4.40 to 4.60. - Hair salons rose from an average rating of 4.58 to 4.79. - Restaurants rose from 4.29 to 4.48. - Words such as “amazing,” “incredibly” and “genuinely” became more common. - Words such as “good” and “nice” declined. - The share of reviews with stars but no written text fell from 47% to 24%.

Between the lines: - The study treats AI-like patterns as signals consistent with writing tools, not proof. - Phone keyboards and Google prompts may also have influenced the language patterns. - All figures are aggregated, and no business or reviewer is identified. - The stronger rise in polished owner replies suggests businesses may be using AI to keep pace with customer expectations for fast, on-brand responses. - A more positive review landscape can make the rare negative review more visible and potentially more influential.

What's next: - Revfer says the full analysis is available on the Revfer blog. - The research report with aggregate data is available at the full report. - Revfer says its software drafts social media posts and Google review replies for small businesses and startups, then publishes them after approval. - Revfer is based in Ashland, Massachusetts, and says its system learns the business once and writes in the owner’s voice.

The bottom line: - Businesses appear to have adopted AI-assisted writing in Google review replies faster than customers adopted it in reviews, while the overall review mix became more positive.