Ecommerce businesses sell on a multiple of adjusted earnings, not revenue, and channel mix moves that multiple more than size does. Single channel Amazon brands sit near the bottom of current bands while diversified brands carrying real margin sit at the top.
The aggregators that paid 7x for Amazon brands are gone. The founders still quoting their prices are not.
In 2021, Amazon aggregators raised $12.3 billion, roughly three quarters of it as debt, and spent it buying third party brands at multiples that reached 7x earnings for assets that in hindsight did not justify the price. By 2022 that fundraising had fallen to $2.7 billion. Thrasio, which had acquired around 200 brands in two and a half years, eventually filed for bankruptcy. Marketplace Pulse laid out the sequence in its postmortem on how aggregator economics came apart.
Most founders formed their mental model of what an ecommerce business is worth during that window and have not updated it since. Ask a founder what the business is worth and you get a revenue number. Ask a buyer and you get a multiple of adjusted earnings, applied only after those earnings have been restated to show what a new owner could realistically expect to take home.
That gap is where unrealistic asking prices come from, and it is also where the opportunity sits. Whether you run $500K a year or $10M, the three factors that move your multiple most are knowable today and largely fixable inside twelve months. What follows is what buyers are paying in 2026, and how to work out which of your own numbers is capping the figure.
Get a real valuation six to twelve months before you intend to sell, not the week you decide to list, because the gaps a valuation surfaces take months of operating work to close. Before deciding whether to sell, when to sell, or what to fix first, the single most useful thing a founder can do is get an actual Business Valuation rather than working from an industry rule of thumb pulled off a blog post.
A generic multiple applied to revenue ignores the factors that move your number in either direction: channel mix, margin trend, and catalog concentration. Those are the difference between the bottom and the top of a published band, which on a business earning $500,000 is a seven figure spread. It helps to understand how a valuation is actually constructed, from SDE multiples through discounted cash flow before you react to whatever comes back.
Getting the number early also gives you time to act on it. If a valuation comes back lower than expected because of channel concentration or margin pressure, that is exactly the kind of gap six to twelve months of deliberate work can close. It is far easier to fix before a sale process starts than to explain away once a buyer’s diligence team has already found it.
The reason differs by stage. At $500K a year you are getting a valuation to know what you are building toward, and the number matters less than the list of what is holding it down. At $5M you are getting one because the number is about to be tested by someone with an accountant and an incentive to find every reason it should be lower.
Amazon FBA brands now trade at roughly 2.5x to 4x seller discretionary earnings, well below an aggregator era peak that reached 5x to 7x, and the compression reflects a permanently different buyer pool rather than a dip that reverses on its own. Sofer Advisors put the current FBA band at 2.5x to 4x SDE in August 2026. Ad Astra Equity’s June 2026 multiples reference states plainly that the 5x to 7x SDE range aggregators paid between 2020 and 2022 is gone, with multiples compressed 40 to 50 percent from peak.
It is worth understanding why, not just that it happened. The 2021 aggregator era ran on cheap capital chasing a narrow category of assets. When that capital dried up, the buyer pool shrank to people running the numbers on cash flow rather than betting on multiple expansion. FE International’s April 2026 read of the buyer landscape describes who is left: individual operators, holding companies, private equity sitting on elevated dry powder, and strategic acquirers. Financial buyer deal value surged 54 percent to $536 billion in 2025, so the money exists. It is priced by people who intend to run the business.
This matters for timing. A brand that commanded 5x in 2021 purely for being a well run Amazon storefront does not get that premium today for existing. Buyers now pay for durability and diversification, which makes the compression a signal about what is rewarded rather than a dip to wait out. Run it through the eighteen month test: nothing in the current buyer mix suggests a return of hundred plus aggregators with debt financed mandates. Plan against the market you are actually selling into.
Channel concentration is the single largest controllable lever on your eventual multiple, and it only counts when the added channel carries real margin. A brand generating 70 percent or more of revenue through Amazon is held near the lower end of its band regardless of how clean the financials are, because platform concentration is a risk buyers price automatically. Here is where the current published bands sit.
FE International describes the mix buyers treat as ideal: roughly 30 percent organic search, 30 percent paid social and search, 20 percent email and owned channels, 20 percent direct or referral. It also names the point where product concentration bites, a single product above 40 percent of sales. Buyers are not paying for a diversification label. They are paying for a revenue base where no single dependency can remove a third of it.
There is a caveat the standard advice skips, and it matters most at $500K to $2M. A second channel that loses money lowers the earnings your multiple is applied to, so diversification can lift the multiple and shrink the base at once. The move only pays when the added channel carries genuine margin. This is the premature complexity pattern that recurs in that band: operators add channels, apps and ad platforms before checkout, fulfillment and retention are solid, and end up with more revenue and less profit. If you are adding Amazon alongside a Shopify store, the question is whether it works as a genuine acquisition channel feeding your owned audience, not whether it adds a line to a pie chart.
Buyers apply the same scepticism to the label. A brand where most of the diversification is a second Amazon storefront under a different name gets no credit for the mix it claims, and a single channel brand with a clean record and minimal SKU concentration can still reach the top of its band. The bands are gravity, not law.
Two businesses with identical current margins get very different offers depending on which way those margins are moving, because buyers price three to five years out rather than the trailing twelve months. Platform costs are the usual reason the trend is unfavourable. Marketplace Pulse reported in February 2023 that Amazon’s cut of seller revenue had passed 50 percent, up from around 40 percent five years earlier, counting referral fees, fulfillment and the advertising that is no longer genuinely optional. That figure is now several years old and includes ad spend, so treat it as a direction of travel rather than your own current number.
The annual increments look small in isolation. Amazon’s own 2026 referral and FBA fee announcement raises fulfillment fees by an average of $0.08 per unit sold, less than half a percent of an average item’s selling price, effective January 15 2026, with no referral fee increase. One year of that is noise. Six years of it compounding against flat retail prices is the reason so many brands have the same revenue and thinner margins than they did in 2022.
What a buyer sees in flat or shrinking margins with no offsetting plan is a business worth less a year from now than today. What tells a different story is visible effort the other way: diversifying fulfillment away from one provider, renegotiating supplier costs, shifting mix toward higher margin channels, or lifting repeat purchase rate so acquisition cost is amortized across more orders per customer. Retention is the lever most founders underuse, and the economics of repeat purchase rate at $50K to $2M are worth understanding before you conclude your margin problem is a pricing problem.
This is where a founder’s own instincts tend to mislead them. A healthy current year margin feels like the number a buyer will value the business on. It is not. The trend is, and a trend heading the wrong way gets discounted even when the current snapshot looks fine.
A business where one product generates half or more of total earnings gets priced near the bottom of its band, not because buyers dislike hero products but because one supply disruption, competitor knockoff or listing policy change can remove most of the earnings at once. Sofer Advisors is explicit that risk factors like single product and single supplier dependency move a brand within the 2.5x to 4x range rather than triggering a fixed percentage deduction, so treat concentration as band position rather than arithmetic. Anyone quoting you a precise number of turns knocked off for SKU concentration is estimating.
The fix does not require launching an entirely new product line under time pressure. Splitting a hero product into two or three genuinely distinct variants, or building out a real second bestseller alongside the first, materially changes how concentrated a buyer perceives the business to be without a ground up catalog rebuild. What matters is whether earnings survive the loss of the top listing, not how many SKUs appear in the export.
Supplier concentration belongs in the same conversation and gets checked just as reliably. A single supplier with no documented agreement and no qualified alternative is the same structural risk as a single product, and cheaper to fix. Sourcing a backup and getting terms in writing is a few weeks of work that removes a diligence finding permanently.
Both sit inside a larger pattern. Bootstrapped founders scaling from $100K to $5M tend to underestimate the operational architecture the business actually requires, and concentration risk is what that gap looks like when a buyer prices it. The businesses that reach the top of their band are rarely the ones with the best marketing. They are the ones that do not depend on any single thing continuing to work.
Two brands with the same adjusted earnings can be worth close to double one another, and the difference is almost entirely channel mix and product concentration. The comparison below is an illustrative benchmark built from the published bands above, not a record of two transactions, so treat the dollar figures as arithmetic rather than comparables.
Brand A earns $500,000 in SDE, sells exclusively on Amazon, has one product generating 60 percent of revenue, and margins that have compressed slightly over two years. That profile sits at the bottom of the single channel band. At 2.5x, the business is worth around $1.25 million, and a buyer who prices the margin trend conservatively may come in lower.
Brand B earns the same $500,000. It does 55 percent of revenue on Amazon and the rest across DTC and wholesale, no single product above 30 percent, and margins holding steady or improving. That profile earns the top of its band. At 4x it is worth around $2 million. Same earnings, roughly $750,000 of difference, none of it explained by how hard either founder worked.
That gap is exactly what diligence teams are trained to find, and exactly the kind of gap a founder can close with a year or two of deliberate work before ever listing. It is also why the order of operations matters. Fixing these things after an offer is in hand means renegotiating from a weaker position than waiting twelve months would have produced.
Clean financials do not change the valuation methodology, they change how well the resulting number survives contact with a real offer. A valuation run on messy books produces a softer, less defensible figure than one run on clean books for an identical business. Separate personal and business expenses. Document supplier agreements rather than relying on informal arrangements. Have two to three years of consistent financials that match your tax returns, because the first thing a serious buyer does is reconcile the two.
Most of that is bookkeeping discipline rather than accounting sophistication, and it is easier if your profit and loss statement already reconciles across channels instead of being assembled from platform exports the week a buyer asks. Expect scrutiny on add backs. Buyers generally accept owner salary and genuinely discretionary expenses when calculating SDE, and push back on anything that looks like a real cost of the business dressed up as a personal one. Every rejected add back comes straight off your earnings base before the multiple is applied.
One more thing to plan for: the headline multiple is not the whole number. FE International reports deal structures increasingly featuring 60 to 75 percent upfront cash with earnouts over 12 to 24 months. A 4x offer with 60 percent at close and the rest contingent on performance you no longer control is a different proposition from a 3.5x all cash offer. Model both before anchoring on a multiple.
Timing the valuation matters as much as the preparation. A number six months to a year before you intend to sell gives you room to respond to what it tells you. A valuation that comes back lower than hoped because of channel concentration or a shrinking margin is not bad news if there is still time to fix it. It only becomes bad news when it arrives too late to act on. If you are still weighing whether this is even the right window to sell, the signals that tend to show up before founders decide to sell are worth reading alongside your own numbers.
Your multiple is set mostly by three things you control, and the aggregator era numbers many founders still quote sit roughly 40 to 50 percent above where the market actually is. Channel mix, margin direction and product concentration are the three, and all three respond to twelve months of deliberate work in a way that revenue growth alone does not.
Founders who anchor to 2021 headlines tend to be disappointed, or worse, waste months in a process that stalls once a real offer lands and the gap between expectation and market becomes undeniable. Understanding where your specific business sits on those three factors before you ever talk to a buyer is what separates a number you can defend from one that falls apart under scrutiny.
If you only do one thing after reading this, pull your revenue by channel and your earnings by product for the last twelve months and write down the two percentages: what share comes through your largest channel, and what share of earnings comes from your best product. Those two numbers tell you more about your eventual multiple than your revenue figure does, and most founders have never written them down side by side.
Most ecommerce businesses sell between 1.5x and 7x earnings in 2026, with the band depending on business model and size rather than revenue alone. Dropship and undifferentiated catalog businesses under $2M typically trade at 1.5x to 3x SDE. Amazon FBA brands sit at roughly 2.5x to 4x SDE per Sofer Advisors in August 2026. Branded Shopify DTC businesses between $3M and $15M reach 3.5x to 5.5x EBITDA, and genuinely hybrid brands above $15M with DTC, marketplace and retail distribution reach 5x to 7x EBITDA. The aggregator era band of 5x to 7x SDE for Amazon only brands no longer exists, having compressed 40 to 50 percent from its 2021 peak.
Profit is used, specifically adjusted earnings, and revenue only matters as context for which earnings measure applies. Businesses below roughly $2M to $3M are usually valued on seller discretionary earnings, which adds back the owner’s compensation and genuinely discretionary expenses to net profit. Larger businesses are valued on EBITDA, which does not add back a market rate salary for the owner’s role because a buyer will need to pay someone to do it. This is why two founders with the same revenue can hear very different numbers. The multiple is applied to earnings, and the definition of earnings changes as you scale.
Amazon concentration above roughly 70 percent of revenue holds a business near the bottom of its multiple band rather than triggering a specific percentage discount. Buyers price platform dependency as a durability question: a policy change, suspended listing or fee increase can compress earnings without warning, and no amount of clean bookkeeping offsets that. Brokers working this market are generally clear that concentration moves a brand within its range rather than applying a fixed deduction. Practically, moving Amazon below 40 percent of revenue with channels that carry genuine margin is what shifts a business from the bottom of a band toward the top.
Get a valuation six to twelve months before you intend to sell, because that is the minimum runway needed to act on what it finds. A valuation is only partly a price. It is mostly a list of the specific things holding your number down, and channel concentration, margin trend and product concentration all take one to two quarters of operating work to move. Getting the number the week you decide to list turns it into information you cannot use. If you are earlier than that, a valuation is still worth running as a baseline, because knowing which lever matters most changes how you spend the intervening year.
Yes, a single product generating half or more of earnings holds a business near the bottom of its multiple band, because one supply disruption, knockoff or listing suspension can remove most of the earnings at once. Buyers check this on every deal. The useful news is that the fix is cheaper than founders expect. Splitting a hero product into two or three genuinely distinct variants, or building a real second bestseller, changes perceived concentration without rebuilding the catalog. What buyers are testing is whether earnings survive the loss of the top listing, so the same logic applies to supplier concentration.
