Online reputation management for small businesses: what works and what’s a waste of money

A landscaping company owner I talked to last spring had just signed a $1,400-a-month reputation management contract. Twelve-month term. The pitch deck promised “negative content suppression,” “review velocity optimization,” and “24/7 brand monitoring.” Eight months in, his Google rating had moved from 3.9 to 4.0, and the one thing that changed it - a run of genuine five-stars after he fixed his scheduling backlog - had nothing to do with the agency.

That’s $11,200 for a tenth of a star he earned himself. This post is the audit I wish he’d run before signing: which reputation tactics are worthless (and sometimes illegal), which ones are real, and what the real ones cost when you buy them directly instead of through an agency markup.

Why so much of this industry sells nothing

Reputation management has a structural problem: the outcomes small businesses want most - bad reviews gone, rating up, more five-stars - are mostly outside anyone’s direct control. Google decides what gets removed. Customers decide what gets written. An honest vendor can influence those things at the margins. A dishonest one sells you the illusion of control and bills monthly.

The market rewards the dishonest pitch, because “we will make the 1-star disappear” closes deals and “we’ll help you respond well and ask more customers” doesn’t. So the loudest offers in this space are the worst ones. Here they are, by name.

Four things agencies sell that don’t work

1. Guaranteed review removal

No agency has a back channel to Google or Yelp. None. The removal process available to a $3,000-a-month agency is the same flagging form available to you, free, in your Google Business Profile dashboard. Google removed 292 million policy-violating reviews in 2025 by its own count - but the overwhelming majority of those came from automated detection, not flags. When a specific review gets flagged, the realistic removal rate is somewhere around 30%, and it depends almost entirely on whether the review actually violates a written policy. (We walk through what qualifies in what actually gets a Google review removed.)

An agency “guaranteeing” removal is doing one of three things: flagging and hoping (which you can do), filing fraudulent legal takedown requests (which can blow back on you), or planning to refund you quietly when it fails. The Federal Trade Commission’s red-flag list for reputation vendors starts with exactly this promise.

2. Review gating

Gating is the funnel trick: survey every customer privately first, then route the happy ones to Google and the unhappy ones to a “feedback form” that goes nowhere. Agencies sold this for a decade as “review filtering software.”

It’s now explicitly illegal. The FTC’s Consumer Reviews and Testimonials Rule (effective October 2024) prohibits suppressing negative reviews while soliciting positive ones, with civil penalties up to $53,088 per violation. Per violation - meaning per gated customer, in the worst reading. The FTC sent its first warning-letter sweep under the rule to ten companies in December 2025, so this is no longer a theoretical statute. If your current vendor’s software asks customers “how was your experience?” before deciding whether to show them the Google link, you are the one holding the liability, not the vendor. We covered the enforcement record in the FTC rule’s first year .

3. Negative content “suppression”

This is SEO applied to burying things: spin up directory profiles, microsites, and press releases about your business until the bad Yelp page or news story slides to page two of search results. It sounds plausible. It fails for two reasons.

First, review platforms have enormous domain authority. Your Google Business Profile and Yelp page aren’t ranking because nobody’s tried to outrank them; they rank because Google wants them there. Second, even when suppression “works” on the search page, it does nothing to the surface where buying decisions actually happen - the review panel itself. A prospect choosing a plumber isn’t paging through ten blue links. They’re reading your reviews on the map listing, where your suppressed content doesn’t exist.

The one scenario where suppression-style work has real value is a genuine crisis with news coverage - and that’s a different playbook entirely (we wrote a 72-hour version of it).

4. “Monitoring” as a paid service

Brand monitoring is real work at enterprise scale - hundreds of locations, multiple languages, social listening. For a business with one to five locations, “monitoring” means: get an email when a review lands. Google does this free. Yelp does this free. A Google Alert on your business name covers most of the rest. When monitoring is the headline deliverable in an agency proposal, you’re being billed a retainer for notification settings.

What actually moves your reputation

Strip out the theater and four activities remain. They’re unglamorous, which is why nobody builds a pitch deck around them.

Respond to every review, and write the responses for the audience

BrightLocal’s 2024 Local Consumer Review Survey found 88% of consumers would use a business that responds to all its reviews, against 47% for one that responds to none. That 41-point gap is bigger than the gap between a 4.0 and a 4.5 rating in most categories, and it’s the one lever entirely under your control.

The craft matters as much as the coverage - a defensive reply under a 1-star does more damage than silence. The full method is its own post, the negative review response playbook, but the one-line version: you’re writing for the hundred prospects who’ll read the exchange later, not the one reviewer. And if staring at a blank reply box is the bottleneck, there’s a tone-matched set of negative-review templates that works as a starting draft - edit it until it sounds like you, or it defeats the purpose.

Ask for reviews - every customer, no filter, no incentive

The single most reliable finding in this field: businesses that ask get reviews, businesses that don’t ask get only the angry ones. BrightLocal’s 2026 survey has 83% of customers writing a review when asked, and the share who always do jumped from 16% to 28% in a year. Asking is free. The compliant version is: ask everyone, at the moment the job wraps, with a direct link, and offer nothing in exchange. (Offering a discount for a review is its own legal mess - short answer no, and here’s the FTC rule that made it official .)

Flag the reviews that actually violate policy - carefully

Fake and off-topic reviews do come down, but the flag is closer to a legal filing than a complaint form: you cite the specific policy, you get roughly one shot, and vague “this is unfair” flags burn it. Budget an hour per genuinely problematic review, not five minutes.

Fix the thing the reviews keep saying

The least marketable advice in reputation management: if eleven reviews this year mention hold times, the reputation problem is the hold times. An agency will never tell you this, because the fix isn’t billable to them. Read your last 25 negative reviews, tally the complaints by theme, and take the top theme to your next staff meeting. That tally beats any sentiment-analysis dashboard on the market.

A worked example: firing the agency

Riverbend Family Eye Care in Fort Wayne - two optometrists, one location - spent 14 months paying $1,150/month for “full-service reputation management.” Deliverables: a monitoring dashboard, monthly PDF reports, and canned responses posted under reviews within an hour of each one landing. Rating over those 14 months: 4.1 to 4.2.

Then a 2-star landed that the agency answered on autopilot:

“Waited 55 minutes past my appointment time, then felt rushed through the exam in 10 minutes. Front desk didn’t even acknowledge the wait.”

“Thank you for your feedback! We’re sorry your experience didn’t meet expectations. Please contact our office so we can make it right!”

The reviewer replied publicly: “This is a bot. Case in point.” That exchange - visible to every prospective patient - is what $1,150 a month bought. The practice manager cancelled the contract, and replaced it with a routine that costs about 90 minutes a week: a text with a review link sent when each patient checks out, and every review answered by an actual human within a day, worst complaints first. She also rebuilt the double-book slot in the scheduling template that caused the waits - the thing four other reviews had already mentioned.

Eleven months later: 4.2 to 4.6, review volume up from 3 a month to 11, and $12,650 in agency fees back in the budget. Nothing in that outcome required a vendor. At higher volume it requires tooling - there’s a real threshold where a paid response tool earns its keep, which we map in free vs. paid review response tools - but tooling and a $14k retainer are different purchases.

When paying for help is actually right

This isn’t an argument that every dollar spent on reputation is wasted. Three purchases hold up:

Software, once volume justifies it. Past roughly 30 reviews a month or three locations, spreadsheet-and-memory breaks down and an aggregation tool pays for itself in coverage alone. The decision logic is in review response at scale .

A lawyer, for the narrow cases. Defamatory factual claims, extortion attempts, reviews disclosing private information - an hour of actual legal advice beats any amount of agency “escalation.”

A person, if the constraint is honestly time. A part-time employee or VA who owns the ask-and-respond routine is often better value than software or agencies - they can also answer the phone. What you’re buying is execution of the boring loop, and it should be priced like admin work, not consulting.

The part nobody tells you: the trust floor is dropping

BrightLocal’s trust numbers have been falling for years - the share of consumers who trust reviews as much as personal recommendations dropped to 42% in 2025, from 79% in 2020. Fake reviews, AI-generated sludge, and incentivized five-stars did that.

Most owners read that stat as bad news. It’s the opposite. When readers stop trusting star counts, they start reading for signals that can’t be faked cheaply - and the hardest thing to fake is a year of specific, non-defensive owner responses. A wall of unanswered five-stars now reads as suspicious. A 4.4 with visible, human engagement under the 1-stars reads as real. The declining trust environment is a transfer of advantage from businesses that accumulate ratings to businesses that demonstrably show up, and it makes the agency-theater model - volume, gating, suppression - worth less every year.

Reputation management for a small business in 2026 is four habits and maybe one piece of software. Everything else on the invoice is decoration.