Seller Account Health

01

What standing actually controls

Seller metrics are not a report card. They are inputs to systems that decide, continuously and without notification, how much of the platform you get access to.

• Search ranking. Marketplaces are optimising for completed, problem-free sales. A seller whose orders generate cases is a worse bet for the platform, and the ranking reflects that regardless of listing quality.
• Fees. Some platforms apply discounts to sellers in top standing and surcharges to those below standard. The same sale can net differently for two sellers.
• Payout timing. New or troubled accounts often have funds held longer, which is a cash-flow problem rather than merely an inconvenience.
• Selling limits. How many items and how much value you may list at once. This is a direct cap on growth.
• Whether the account survives. Sustained poor performance leads to restriction and eventually suspension.

02

What gets measured

The names differ by platform, but the underlying measurements are consistent because they all proxy the same thing: how often does buying from this seller go wrong.

Notice how few of these are about the goods. The majority measure your operations — how fast you reply, how fast you post, whether your stated quantity was real. That is good news, because operations are the part you fully control.

03

The behaviours that damage accounts quietly

Sellers rarely get into trouble through a dramatic failure. They get there through four ordinary habits.

1. Cancelling your own orders The most damaging routine action available, because a cancellation for lack of stock is generally counted against the seller rather than treated as neutral. Its usual cause is not carelessness but structure: the same physical unit listed in two places with two independent quantities. Overselling is an inventory architecture problem that presents as a metrics problem.
2. An optimistic handling time Promising same-day dispatch and shipping the next afternoon converts every order into a late one. Stating a handling time you can hit on your worst week costs you almost nothing in conversion and removes an entire metric from your risk surface.
3. Letting messages sit A buyer whose question goes unanswered for two days opens a case instead. Cases closed without seller resolution are among the heaviest-weighted measures on most platforms, and they are entirely a response-speed artefact.
4. Grading generously Every optimistic condition grade is a not-as-described claim with a delay on it. Rounding down costs you a little price and removes the most damaging dispute category from your account.

All four are process defects, and all four are fixable in an afternoon — which is why account health tends to be a solved problem for sellers who take it seriously once, and a recurring crisis for those who treat each incident individually.

04

Policy violations are a separate risk

Performance metrics degrade gradually. Policy violations do not — they can restrict an account immediately, and the seller is often genuinely unaware they broke a rule.

• Restricted and prohibited items, which include categories most people would not guess. Check before listing anything unusual rather than after.
• Intellectual property claims — counterfeits, unauthorised reproductions, and using a rights holder's images. Repeat claims escalate quickly.
• Keyword stuffing with brands the item is not, which is a policy breach and not merely bad practice.
• Moving transactions off-platform, or including contact details to encourage it.
• Operating multiple accounts in ways the platform does not permit, particularly after a restriction on one of them.

The safest posture is to read the policy pages for the categories you actually sell in, once, properly, and to treat "everyone does it" as evidence of nothing. Enforcement is increasingly automated and does not weigh custom.

05

Monitoring on a schedule, not on alarm

Metrics are calculated over trailing windows, which means by the time a dashboard shows a problem, the events causing it are weeks old and already locked in. Checking monthly is early warning; checking when you get a warning email is too late to prevent it.

1. Once a month, open the seller dashboard on each channel and read the actual rates rather than the summary badge.
2. Look at direction, not just level. A defect rate that has doubled while remaining under the threshold is the signal; the threshold breach is the consequence.
3. For every problem order in the period, write down the cause in one line. Patterns become obvious after two months and are invisible in individual incidents.
4. Fix the process the pattern points at — handling time, response speed, grading, stock accuracy — rather than the individual orders.

Keep a note of each channel's current thresholds, since they differ and they change. Knowing you are at half the allowed defect rate is useful; knowing you are "fine" is not.

06

If you are restricted or suspended

The instinct is to send an urgent message insisting there has been a mistake. That is the response least likely to work, because appeals are assessed on whether you have identified and fixed a cause.

1. Read the notice properly Find the specific policy or metric cited. Appealing the wrong thing wastes your one good attempt.
2. Fulfil every open order Whatever your standing, complete what you owe. Outstanding orders during a restriction make everything worse.
3. Identify the actual cause Not "a few unfair buyers". Something concrete: stock quantities were wrong across two channels, handling time was unachievable, condition grades were inconsistent.
4. Fix it before appealing, and say what you changed A plan of action that names the cause and the specific change you have already made is what gets reinstatements. Assertions of innocence generally do not.
5. Appeal once, calmly, and wait Repeated messages restart queues rather than accelerating them. One clear, factual appeal.