Threecolts reviews — good scores, small samples, and one line that is not a subscription

The public record, every score with its count

Each figure below came from the platform that issues it, read 21 August 2026. A score without its count is not a fact, and on this vendor the counts do most of the work.

• Trustpilot, threecolts.com — TrustScore 4.4 from 67 reviews. The largest single record the company has, and the one we would weight most.
• G2, Threecolts Seller 365 — 4.8 out of 5 from 10 reviews.
• G2, Threecolts Multichannel Pro — 4.9 out of 5 from 12 reviews.
• G2, Threecolts Margin Pro — 5.0 from 1 review. Printed only so nobody quotes it at you as a five-star product.
• Capterra — a Threecolts listing exists with no rating and no reviews. Absent rather than negative.
• Trustpilot, inventorylab.com — TrustScore 3.0 from 6 reviews. Same treatment: six is not a rating.

Put plainly: the scores are good, and there are not many of them. Sixty-seven reviews across a company that has acquired and folded in a long list of seller tools is a thin public record for the size of the product surface, and the honest thing is to say so rather than lead with 4.9.

How to handle a small sample without pretending it says nothing

There are two bad ways to read ten reviews. One is to treat 4.8 as equivalent to a 4.8 across four thousand. The other is to dismiss it. Neither is right.

What ten reviews cannot do: tell you the distribution. With a sample that small a single unusual experience moves the average by half a star, so the number is unstable in a way a large record simply is not. Do not build a purchase decision on the arithmetic.

What ten reviews can do: tell you what people who chose this product were trying to accomplish, and whether that matches you. Read them individually rather than averaging them, and pay attention to the reviewer's role and business size rather than the star count.

The rule we work to is: state the count, decline to draw a conclusion from a small one, and cite the larger record where one exists. Here the larger record is Trustpilot's 67, and it lands at 4.4 — lower than the G2 figures, which is exactly what you would expect when a sample widens. That is not a contradiction between platforms. It is a small sample regressing toward a bigger one, and it is a mild point in favour of trusting the 4.4.

One thing we are deliberately not doing. The Trustpilot record is not unanimous, and one low-rated entry makes a serious allegation about the company that we cannot verify against any primary source. We are not repeating it, not hedging it, and not gesturing at it through a rhetorical question. An unverified claim that damages a company is the most expensive kind of error a page like this can make, and "some users say" is not a citation. If a specific concern matters to your procurement process, raise it directly with the vendor and ask for documentation.

What is actually being reviewed here — a bundle, not a product

This is the part that makes Threecolts reviews harder to read than most, and it is worth understanding before you weigh any score.

Threecolts is an assembly of acquired seller tools sold as bundles rather than a single application. Seller 365 packages sourcing, mobile scanning, a browser analysis extension, prep-and-ship workflow, repricing and review-request automation into one subscription. InventoryLab now sits inside it — inventorylab.com is live and trading, and its trial and plan links route to Threecolts' own hub. That combination is the same shape as several others in this market: the brand stays up, the billing moves.

The consequence for reviews is direct. A reviewer rating "Threecolts" may be rating one tool inside the bundle. Read the recent Trustpilot entries and several of them name a specific product — the repricer, a support interaction, an onboarding session — rather than the platform. That is not a flaw in the reviews; it is a signal about how the product is experienced. Somebody delighted by the repricer tells you very little about the shipment workflow.

So the practical instruction is: identify which tool inside the bundle you are actually buying it for, and read only the reviews about that. Then treat everything else in the bundle as a bonus rather than a reason. Bundles are priced on breadth and used on depth, and people who buy them for breadth are the ones who churn.

The pricing behind the reviews, both bases, checked

Threecolts publishes its reseller-bundle prices, which is more than several competitors do. From threecolts.com/pricing, read 21 August 2026, on the Seller 365 reseller bundle:

• Standard — $69 a month billed monthly, or $65 a month on annual billing. One user, full access to ten tools.
• Teams — $129 billed monthly, or $119 on annual billing. Up to ten users.
• Pro — $199 billed monthly, or $183 on annual billing. Higher limits across sourcing, repricing and review requests.

The page states the annual savings as $48, $120 and $192. Run the standing check — take the annual rate, multiply by twelve, and compare it against twelve monthly payments. Standard: $780 against $828, a $48 difference. Teams: $1,428 against $1,548, $120. Pro: $2,196 against $2,388, $192. All three reconcile exactly against what the page advertises, which is rarer than it should be and worth crediting. Check which billing tab you are reading, on any vendor including us.

The Brands bundle, aimed at brand owners rather than resellers, is quote-only — the page says the plan is built around your team and routes to sales. Stated neutrally: they quote on a call for that side of the business.

The trial is a 14-day structure with a condition worth knowing: seven days on signup, plus seven more when you connect a seller account. Plan the connection for day one or you have a week, not a fortnight.

For scale, Sellerboard sits below all of this at $19 to $79 a month billed monthly, or $15 to $63 on annual billing. It does a narrower job. If profit analytics is all you want, the cheaper tool is genuinely the right answer and we would rather say so.

The 15% line that is not in the subscription

Here is the number that decides the real cost for a lot of accounts, and it is not on the tier ladder. Seller 365 carries a 15% reimbursement commission on recoveries.

That is a perfectly normal way to price recovery work — it aligns the vendor with the outcome, and a percentage of money you would not otherwise have had is an easy thing to pay. But it is a variable rate on top of a fixed subscription, and variable rates are where bills surprise people. On a good recovery year the commission can exceed the subscription several times over, so model it separately rather than folding it into the monthly figure.

For context on where 15% sits in the market: GETIDA publishes performance-based pricing starting at 25% — and note the wording carefully, because that is a published floor rather than a flat rate, and several articles report it as a flat 25% when the vendor's own page does not say that.

One structural fact dates every reimbursement comparison written before it: Amazon's updated FBA inventory reimbursement policy took effect on 31 March 2025, moving valuation toward manufacturing and sourcing cost. That shrinks the recoverable pool for every recovery service in the market. If you are reading a comparison of recovery vendors that predates it, the arithmetic in it no longer holds.

What decides whether you recover anything — and it is not the tool

The thing nobody puts in a review of recovery software is that the software is not the constraint. Cadence is.

On a beauty brand's account, running reimbursement claims weekly rather than quarterly was the difference between recovering money and missing the window entirely — one fortnight recovered $6,500, including refunds on shipping labels for shipments that were never sent, and a separate week on the same account recovered $4,591. That is one account rather than a benchmark, and nobody should read it as a number to expect. The mechanism is what generalises: claim windows expire, and the categories people forget are the unglamorous ones — unused shipping labels, lost-in-transit units on cancelled shipments, warehouse damage that was never reconciled. The practice that produces recovery is calendar discipline, not a clever tool.

So the question to ask of any bundle, including this one, is not what it can find. It is who runs it, how often, and what happens on the weeks nobody logs in. A tool that surfaces a hundred candidates and a person who reviews them quarterly will recover less than a plain spreadsheet reviewed every Friday.

That is the distinction between buying software and buying the work. Dr. Stock is the second one: reorder timing and stockout risk, storage and aged-inventory surcharges, the removal-versus-liquidation call, fulfilment fee errors and dimensional-weight misclassification, shipment discrepancies and reimbursement recovery, run to a calendar rather than a dashboard. It is run by Fable 5 and supervised by operators from Full Circle, a full-service Amazon management company with more than $500M in managed revenue across 100+ brands, and 70+ brands live across the group today. Purchasing decisions always come to a human. Orbit is included at no additional cost. We publish no price, which Threecolts does for its reseller bundle — a point in their favour.

Two redirects. If reimbursement recovery is the whole reason you are here, the reimbursement services comparison covers who each vendor suits. And if you want the cheapest credible profit analytics rather than a bundle, the Sellerboard review is the honest starting point.