For about eight months in 2023, my morning routine looked like this. Alarm at 6:30 AM. Coffee. Open a spreadsheet with a list of twenty client businesses. For each client: open Google Maps, search for their business, navigate to reviews, sort by newest, scan for anything from the last 24 hours. Repeat on Yelp. Repeat on Trustpilot if relevant. Log anything requiring attention. Draft response suggestions. Send to client.
By 8 AM I had drunk two coffees, checked sixty websites, and done nothing except monitor information that should have been coming to me automatically.
Reputation management software existed, of course. Birdeye, Podium, Grade.us, Reputation.com — a stack of enterprise tools with deep feature sets, sophisticated dashboards, and monthly pricing that started at £299 and climbed toward £2,400 for multi-location clients. Every one of those tools was designed for enterprise chains with dozens of locations and dedicated marketing managers.
Not for a local business consultant managing a dentist, two restaurants, an estate agent, and a plumber, all of whom were paying me a modest monthly retainer to keep their online reputation from going quietly bad while they were busy running their businesses.
The economics did not work. At £299 per month for the cheapest respectable reputation management tool, the tool cost more than the retainer some of my smaller clients were paying. The value I was delivering was real — monitoring and responding to reviews, identifying patterns in customer feedback, flagging urgent negative reviews before they accumulated and damaged search rankings. But the delivery mechanism was broken.
I needed a data layer I controlled, not a software subscription I could not build a business model on top of.
The Multi-Platform Review Scraper was the first actor I built for myself before it became a product.
It takes a list of business profile URLs — direct links to Google Maps business listings, Yelp pages, TripAdvisor listings, Trustpilot company pages, Facebook business pages — and scrapes the most recent reviews from each. You configure the maximum number of reviews per business, the date range, the star rating filter (useful for monitoring specifically for new negative reviews), and whether to include the owner's response if one exists.
The output is a flat, structured dataset: one row per review, with platform, reviewer name, rating, review body, publication date, sentiment classification (positive, neutral, negative — inferred from rating and basic keyword analysis), and owner response text if present.
For an n8n workflow this dataset is immediately actionable. Filter for rating ≤ 2 and date = today. For each matching row, generate a suggested response using Claude's API. Email the client with the review text and the suggested response. Log to a Google Sheet. Total workflow execution time: under two minutes. Total human time required: zero unless the client chooses to use or modify the suggested response.
For twenty clients across four platforms, the actor processes approximately 200 to 400 reviews per daily run depending on review velocity. At £0.002 per review scraped, that is £0.40 to £0.80 per day — £12 to £24 per month — to replace the two-hour manual monitoring routine I had been running every morning.
The thing that strikes me about this, looking back, is that the problem was entirely about data access. The reviews were public. The sentiment analysis was straightforward. The notification logic was simple. The only obstacle was the absence of a structured, affordable, scriptable way to get review data from multiple platforms simultaneously without a £300/month subscription to software built for enterprises.
Local service businesses — the ones who genuinely need reputation management because a run of bad reviews can materially damage their Google ranking and their walk-in traffic — are priced out of the tools built to solve their problem. A dentist practice in a market town should not need to choose between monitoring their online reputation and paying for a piece of software.
The Multi-Platform Review Scraper is £0.002 per review. A dentist with twenty reviews per month across three platforms pays £0.12 per month for the data layer. The workflow to process it costs another few pence in n8n execution. The monitoring infrastructure costs less than a cup of coffee.
My mornings look different now.