Does reputation management work?

The three jobs, and their separate odds

Almost every argument about whether reputation management works collapses because the arguers mean different jobs. There are three, they have very different success profiles, and a provider quoting one number across all three is quoting a number that cannot mean anything.

The job	What it is	What makes it succeed	What makes it fail
Removal	Getting a specific URL or review taken down at the source	The content violates a published platform policy, or a specific law reaches it directly	The content is lawful, policy compliant and merely unwelcome
Suppression	Changing which results appear for a search of your name	The query is not heavily covered, and enough credible material can be built and earned	Strong, established domains hold the results and will not be displaced
Rating and review improvement	Raising a visible average and the volume behind it	You have real customers and a consistent, compliant way of asking	You have very few transactions, or the underlying service problem is real

A fourth thing is often bundled into the same quote and behaves differently again: public relations. PRSA's definition of public relations describes a strategic communication practice concerned with relationships between an organisation and its publics. Whether that works is a fair question and it is not the question this page answers. If it appears on your proposal, ask for it to be priced separately, because its results cannot be checked the way a removed URL can.

The third job works most reliably, because it is close to arithmetic. If you serve customers and ask them properly, the volume of recent positive reviews grows and the visible average moves. The first job is binary per URL. The second is a matter of degree and time.

What decides whether a removal succeeds

Removal is not a negotiation about how much you want it. It is a test against someone else's published rule. The variables that actually matter are these.

• Whether the content breaks a stated policy. Review platforms remove reviews from non customers, reviews containing personal attacks or personal data, and reviews posted for competitive reasons. They do not remove reviews that are unflattering and accurate.
• Which platform holds it. Some run a real appeals process. Some publish a form and answer nothing. This varies more than anything else and it is most of what experience buys.
• Whether a specific statute reaches it. Copyright, certain categories of personal and financial data, and in some jurisdictions statutory privacy rights, all bypass the ordinary policy route. Most content is reached by none of them.
• Who posted it and where else it exists. Content that has been syndicated to other sites has to be worked host by host. One successful removal on the original does not clear the copies.

Suppression: realistic outcomes and what decides them

Suppression works when there is room to work in. Two questions settle most of it. How strong are the pages currently ranking for your name, and how much credible material can realistically be created and earned around you?

An individual with an unflattering post on a small forum and a strong professional presence has a very good chance. A person whose name returns coverage in national outlets does not, and any provider implying otherwise is describing an ambition rather than a plan. This is ordinary search work at bottom rather than anything proprietary, and a provider who cannot explain their method in terms you could check for yourself is worth a second look.

What cannot be fixed

Any page that hedges this is not being straight with you, so here it is plainly.

• A truthful negative review is not a removal candidate. It does not violate policy and no law requires its deletion. It can be replied to and outweighed by volume. It cannot be taken down for being true.
• Lawful news reporting is not removable. Publishers exercise editorial discretion, and a court order against an author does not oblige a newspaper to unpublish. Some outlets will consider an update or an unpublishing request on their own terms, and that is a request, not a route.
• An expunged record does not clear the article about the arrest. Expungement changes the official record. The news coverage is a separate publication with its own First Amendment protection.
• Deleting your own account rarely deletes the content. On most forums the posts remain, reattributed to a deleted user.
• Nothing is removed everywhere, permanently. Copies, archives and screenshots exist outside any provider's reach.
• A real service problem cannot be reviewed away. If the pattern in the reviews is accurate, the reputation work is operational before it is digital. That is not a slogan, it is the cheaper fix.

How to read claimed success rates

Any percentage a provider offers you has a denominator, and the denominator is where the number is made. A firm that only accepts cases it expects to win reports a high rate honestly and tells you nothing about your case. A rate calculated per request rather than per client looks very different from one calculated per client. A rate that counts a deindexed URL and a suppressed URL as the same outcome is combining two things a buyer would price differently.

There is a second reason to be careful with outcome claims: they are advertising claims, and advertising claims need substantiation. Where a provider is a law firm or works alongside one, the lawyer is additionally bound by the ABA Model Rules of Professional Conduct, which restrict claims that create unjustified expectations about results. That is a real constraint a marketing firm does not operate under, and it partly explains why law firm adjacent providers describe outcomes more cautiously.

The useful question is not what is your success rate. It is: what happens in my case if this does not work, and what do I pay then.

Evidence of results that is worth something

Ask for the artefacts rather than the summary. Dated before and after screenshots of the actual search, run logged out. The URL of a removed page, so you can confirm it now returns nothing. A platform's own confirmation message. A monitoring export showing the position of a specific result over time. None of that identifies another client, so client confidentiality is not a reason to withhold it in redacted form.

Measurement, and what actually tells you it worked

Decide what you are measuring before the work starts, because the metric chosen afterwards is always the flattering one.

For an individual, the honest measure is the first page of results for your name, captured logged out at the start and re run monthly. For a business, it is the same plus the visible rating, the count of recent reviews, and whichever inbound signal you already track. The SBA's guidance on managing a business makes the general point that applies here: measure the operational thing you can observe rather than the outcome you wish for.

Reputation work is one input among many, so beware of attributing every change in enquiries to it. Establishing which of the three jobs your situation actually needs, and whether it falls on the workable side of each, is the whole question, and it is the first thing a reputation audit answers.