Marketplace dependency risk is measurable, and revenue share alone does not measure it. Six inputs do: how much of your revenue comes from one channel, how many of your customers you can reach without that channel, your margin after fees, your cash runway, how much of your business would physically move if you had to leave, and how exposed you are to a rule change. Score each from 0 to 3 and you get a number between 0 and 18 that you can recalculate every quarter. Anything above 12 means you are not just dependent, you are fragile, and those are different problems with different fixes.
Below is each input, how to compute it from data you already have, a worked example, and the honest limits of the whole exercise.
Dependence is not the same as danger
This distinction is missing from nearly every article on the subject, and it matters because it changes what you should do next.
Dependence is the share of your business that sits on one platform. It is a fact, and it is often a rational one. A seller who gets 90 percent of revenue from a marketplace that reliably brings them buyers is not being foolish. They are using a channel that works.
Fragility is dependence plus an inability to survive losing it. That is the dangerous state, and you can be highly dependent without being fragile: strong margin, months of runway, a mailing list, a portable catalog and a clean policy record together mean a bad surprise costs you a quarter rather than the business.
So the useful goal is not "reduce dependence." It is "stop being fragile." Those often point at completely different actions, and the second one is usually cheaper.
The six inputs that measure marketplace dependency risk
Score each 0 to 3, where 0 is the safest and 3 is the most exposed. Use trailing twelve-month data, not last month.
| # | Input | How to measure | 0 | 3 |
|---|---|---|---|---|
| 1 | Revenue concentration | Largest channel's share of 12-month revenue | Under 40% | Over 80% |
| 2 | Reachable customers | Share of buyers you could contact for free, with permission | Over 40% | Almost none |
| 3 | Contribution margin | One typical order after all costs and platform fees | Over 35% | Under 15% |
| 4 | Runway | Months of expenses covered if that channel stopped | 6 months or more | Under 1 month |
| 5 | Asset portability | How much of your catalog, reviews, ranking and customer list would move | Most of it | Almost none |
| 6 | Policy exposure | Rule changes and warnings that could affect your listings | No warnings, simple compliance | Prior warnings or a model that leans on one policy staying put |
These bands are working thresholds. They are judgment calls chosen so the score is testable, not findings from a study, and you should adjust them if your category behaves differently. The value is in recomputing the same number each quarter and watching the direction it moves.
1. Revenue concentration
Divide your largest channel's twelve-month revenue by total revenue. Multichannel commerce vendors publish a consistent framing here: companies including Mirakl and Nectar describe single-channel dependency above roughly 70 to 80 percent of revenue as a structural risk rather than a tactical one, as reported on their company blogs. Those are businesses that sell diversification, so weigh the source accordingly, and note that the underlying arithmetic is not controversial. The more of your revenue sits in one place, the more a single decision by someone else moves your whole year.
2. Reachable customers
Count the customers you could contact tomorrow for free, with their permission, if the channel disappeared overnight. Marketplace buyer records generally do not count. Etsy's Seller Policy limits how buyer information from an order may be used, and you should read the current wording at etsy.com/legal/sellers/ rather than rely on any summary, including this one. What counts are opt-ins people gave you deliberately.
This is usually the input sellers score worst on, and it is also the fastest one to improve.
3. Contribution margin after fees
Margin is your shock absorber. It determines whether you can afford to buy traffic during a disruption, discount to clear inventory, or simply survive a slow quarter. Compute it on one real order, and take the fee side from the current published schedule. Etsy's Fees and Payments Policy at etsy.com/legal/fees/ lists what the platform charges: listing fees, transaction fees, payment processing, subscriptions, and the Offsite Ads fee, which applies to orders that follow an ad the platform placed. The fee calculator on the tools page handles the comparison.
4. Runway
Months of business and personal expenses you could cover if the channel stopped paying tomorrow. This is the input that turns a catastrophe into an inconvenience, and it is the only one you can fix without changing anything about how you sell.
5. Asset portability
Walk your business asset by asset and ask what physically moves.
- Product data and photos: yours, and portable.
- Brand name and recognition: yours, if the name is distinctive enough for people to search.
- Reviews: not portable. They are tied to the marketplace's verified-purchase record, which is precisely what makes them worth having.
- Search position inside the marketplace: not portable, and not replaceable by anything you own.
- Customer contact rights: limited by policy, as above.
Most sellers discover their score here is worse than they assumed, because the assets they value most are the two that do not move.
6. Policy exposure
Ask two questions. Have you received warnings, listing removals or intellectual property notices? And does any meaningful part of your model depend on a specific rule staying exactly as it is today? Platform rules move regularly, and a business built on the current wording of one policy is carrying a risk that does not show up anywhere in its financials.
A worked example, clearly hypothetical
The numbers below are invented round numbers used to show the arithmetic. They are not a real seller.
Suppose a shop does $120,000 a year, with $108,000 of it from one marketplace (90 percent, score 3). About 200 of roughly 1,400 buyers are on an opt-in email list (14 percent, score 2). Contribution margin after fees runs 28 percent (score 1). There is about two months of runway (score 2). The catalog and photos move but the reviews and search position do not (score 2). No warnings, straightforward compliance (score 1).
Total: 11 of 18. Dependent, but not yet fragile, and the two cheapest points to move are runway and the email list. Neither requires leaving the marketplace, and both would still be worth doing if this seller never leaves.
What happens when a marketplace actually goes away
Most sellers think about suspension. The rarer event is the one that removes the argument entirely: the marketplace itself closing.
It happens. Ecommerce News Europe reported on July 17, 2026 that Rakuten France, the platform previously known as PriceMinister, is closing at the end of the year after failing to find a buyer, having seen active customers fall 33 percent and traffic fall 42 percent since 2016, as reported. French outlet JustGeek reported a shutdown date of September 30, 2026 for sales on the platform, after which seller shops would no longer be accessible to buyers.
For a seller on that platform, none of the usual advice applies. There is no appeal, no reinstatement, and no version of "improve your listings" that helps. Everything you can do about that scenario has to be done before it is announced, which is the entire argument for scoring yourself while nothing is wrong.
The everyday version is less dramatic and much more common: a fee change, a policy update, a ranking change, a category rule. Same lesson, smaller scale.
What this score does not capture
- It does not measure how good the channel is. A high score on a channel that is growing and profitable is a very different situation from a high score on one that is shrinking.
- It ignores seasonality. A shop doing most of its year in one quarter carries a timing risk that a flat annual average hides completely.
- It cannot price your relationship with the platform. A clean ten-year account with no history of problems is genuinely safer than the number suggests, though it is not a guarantee of anything.
- It says nothing about whether you should leave. Reducing fragility and leaving are different projects. Most of the fixes above make you better off on the marketplace too, which is the point.
What to fix first, by band
- 0 to 5, low fragility. Keep recomputing quarterly. Nothing urgent.
- 6 to 11, moderate. Fix runway and reachable customers, in that order. Both are cheap, and neither requires changing where you sell. Then re-read signs you should start your own ecommerce website to see whether the store question is even live for you yet.
- 12 to 18, fragile. Treat it as an operations problem this quarter, not a migration project. Build runway, start collecting permission-based contacts, and fix margin. If leaving is genuinely on the table, work through the timing gates in is it too early to leave Etsy and the full decision in should I move from Etsy to Shopify before you spend anything.
One more thing worth checking alongside the score: whether your products would find buyers at all without the marketplace introducing them. That is a separate question with a separate answer, in products that only sell on Etsy.
Last checked: August 18, 2026. Platform fees, seller policies and company figures change. Verify rules on the platform's own page and company metrics in the company's own filings before acting.
FAQ
What counts as too much marketplace dependency?
There is no universal line, and revenue share on its own is the wrong test. A seller at 90 percent concentration with six months of runway, a healthy margin and an email list is in better shape than one at 60 percent with none of those. Score the six inputs and watch the total move over time.
Does diversifying to a second marketplace reduce my risk?
Partly. It reduces exposure to one company's decisions, which is real. It does not fix the deeper issue, because on any marketplace the customer relationship, the reviews and the search position still belong to the platform. Two rented channels are safer than one, and neither is an owned channel.
Can Etsy really close my shop without warning?
Platforms do suspend and close accounts under their own policies, and appeals processes exist. Rather than trade horror stories, read the current Seller Policy and the account termination terms at etsy.com/legal/sellers/ so you know what the actual rules are for your shop. Then plan for the risk with runway rather than with worry.
How often should I recalculate this score?
Quarterly is enough for most sellers, and the direction of travel matters more than the absolute number. If the score has not moved in a year, nothing you did that year reduced your fragility.
Is high dependence always a problem?
No. It is a fact about your business model that becomes a problem only when combined with thin margin, no runway and no way to reach your customers. Fix those three and you can stay highly dependent on a channel that works for you, deliberately and with your eyes open.