Review removal vendor due diligence: the checklist nobody published.

Somewhere between the RFP template and the first purchase order, an enterprise reputation buyer runs into a question with no published answer: how do you actually evaluate a review removal subcontractor? Software listicles cover monitoring platforms. Agency-law content explains NDAs and SLAs with no reputation context at all. Enterprise procurement content covers security questionnaires written for SaaS. None of it addresses the thing you are buying, which is fulfilment on a workstream where the outcome is controlled by a third party nobody at the table employs.

So here is that document. Take the sections below into a vendor call and score the answers. We sell removal, which you should factor into how you read the parts where we talk about ourselves. The framework works whoever you end up routing volume to.

Start with the question vendors dodge

Google reported that in 2024 it blocked or removed more than 240 million policy-violating reviews, took down more than 12 million fake Business Profiles, and placed posting restrictions on more than 900,000 accounts. That is Google's own reporting on its enforcement against fake engagement, and it is the most important context in any removal purchase.

Enforcement at that scale runs continuously, in the background, against reviews nobody flagged. Which means that when a review you submitted disappears three weeks later, the cause is genuinely ambiguous. Was it the vendor, or the same sweep that caught 240 million others? A competitor analysis of this category gave the problem a name: credit-claiming opacity. No vendor page in the market addresses it, ours included until now.

If you want harder evidence than risk allocation, there is a way to ask for it. Request a time-to-removal comparison between submitted cases and a matched set of unsubmitted ones, or a control group held back for a defined window. Most vendors will not have run one. How a vendor responds to that request tells you more than any figure on their pricing page.

While you are there, retire the success-rate conversation. Rates in this category are self-reported, with no auditor, no standard denominator and no agreed definition of a case. Two vendors can both publish a number in good faith and mean completely different things depending on whether declined cases sit in the denominator, whether rating-only reviews count, and whether a review that comes back still counts as removed. Ask for the denominator in writing. Nobody in this market is independently audited, including us.

The contract stack

Two instruments, and the failure mode for each one if procurement lets it slide.

Non-circumvention deserves its own paragraph, because it is the clause agencies under-draft. The scenario you are protecting against is your end client contacting the vendor directly, which happens more often than people expect once a removal lands and somebody gets curious about who did the work. Agency trade content commonly cites 12 to 24 months post-termination for non-solicitation and non-circumvention, confidentiality surviving 5 years for general confidential information, and indefinite survival for client identity and for the existence of the white-label arrangement. Treat those as market conventions rather than legal standards. They are useful anchors in a negotiation, not something you can cite at anyone.

Two more items that get skipped. Work product and reporting belong to you or your client, not the vendor: a vendor retaining rights over the case file or the evidence package has given itself leverage in a renewal conversation. And write down off-boarding before you sign, meaning return of credentials and Business Profile access plus a defined transition window, with 30 days the figure most commonly cited.

White-label integrity is a list, not a promise

The highest-anxiety question this audience actually types into a search box is whether the end client can find out a subcontractor is involved. Fair question, boring answer: get a written list before signing of every surface that stays vendor-branded. Specifically the sending domain on any email that could reach the client, help links, widget footers, and terms and privacy links.

Then ask for a sample report before contract. Not a redacted case study. A real report in the format you would forward to a client. Leaks in white-label arrangements are rarely dramatic. They are a footer, an auto-reply, or a link buried in a PDF.

Operational terms that matter at volume

Ask for a response and reporting SLA with an escalation ladder attached, not a general commitment to be responsive. One provider in this market publishes a one business day response commitment, which makes a usable reference point when you ask others to match it. Our own numbers, for comparison: accept or decline within 24 hours of receiving a case, median turnaround 4 days, range 3 to 7.

Pricing structure matters more at scale than unit price does. Ask whether the model holds from 20 accounts to 200, and whether it prices per location, per seat or per case. Per-case is the easiest to reconcile against your own retail. Our agency tiers are custom by volume and set on a 30 minute call, modelled against your retail and projected throughput rather than quoted off a rate card. Direct retail, for comparison, is $450 per review. Coverage is US, Canada, UK and EU, in English. Payment runs by SEPA, wire or stablecoin, always after confirmed removal.

Method, client exposure, and three red flags

Ask what the vendor actually does. Then ask the second question, which is the one your client's legal team will eventually ask you: what exposure does the end client carry if the method is one that could get their own Business Profile flagged, or surface publicly as review suppression? The first question gets you a description. The second gets you a risk profile.

Three red flags are worth naming plainly, because they are what buyers in this market are taught to watch for.

1. Anyone guaranteeing removal of any review. Rating-only reviews, off-platform content and reviews with no policy hook all exist. A blanket guarantee means the scope was never read.
2. Large upfront fees before case evaluation. Payment ahead of assessment prices your risk into their revenue.
3. Anyone claiming insider access at Google. There is no such channel to sell, and a vendor willing to claim one will claim other things.

Our method is published as a proprietary, direct removal process and no further than that. Our declared limit is worth stating because it is the kind of thing a vendor hides: reviews containing text are in scope, rating-only reviews (a star with no text) are not, and we say so before accepting a case rather than after. Published track record is 1,427 reviews removed between January and May 2026, 100% success on reviews containing text, 0 of them returned, with a lifetime re-removal warranty written into every invoice and into partnership agreements. We have no contact with a partner's end client and a published policy of not competing with partners for their clients.

Run a pilot instead of an RFP

Procurement practice in this category keeps moving toward a scoped pilot on real cases rather than a paper RFP, and the reasoning holds up. An RFP tests a vendor's proposal team. A pilot of five to ten live cases tests acceptance criteria, turnaround, reporting quality, and vendor behaviour on a case that goes badly. Score the pilot against the sections above and weight the failed case heaviest, because that is where vendors stop looking alike.

For the commercial mechanics underneath all of this, see our white-label primer for agencies, or why most removal services fail.

For the wider picture, see what Google's 292 million removals mean for businesses in 2026.