Podcast: Download (Duration: 47:47 — 55.0MB)
Successful Amazon-only sellers now earn net margins in the low single digits. The money leaves through fees that never appear as a line item in most sellers’ books.
That number comes from Parag Mamnani, founder of Webgility and a former Amazon employee who managed the company’s web store developer program. His software sits on top of the finances of thousands of ecommerce sellers, so he sees the aggregate picture no individual seller can.
The three biggest costs in an ecommerce business, cost of goods, payroll, and advertising, all show up in your bank account. The leak is everything netted out before you receive your money: referral fees, storage, low-inventory penalties, payment processing, refunds, and returns.
This post covers what margins look like by seller size, the fees that surprise people, when to hire a bookkeeper, and why Parag doubts the agentic commerce hype.
Get My Free Mini Course On How To Start A Successful Ecommerce Store
If you are interested in starting an ecommerce business, I put together a comprehensive package of resources that will help you launch your own online store from complete scratch. Be sure to grab it before you leave!
Table of Contents
Key takeaways
- The most successful Amazon-only sellers now run net margins in the low single digits.
- More than half of Amazon sellers are based outside the US, and FBA erased the domestic delivery advantage.
- Cost of goods, payroll, and ads are visible in your bank account, while fees hide inside sales channels and processors.
- Treating your net marketplace deposit as revenue is the most common bookkeeping mistake.
- Set up a separate clearing account for every sales channel so fees net out as each payout arrives.
- SKU-level margin tracking is what moves a seller from six figures to seven and eight.
- Hire a bookkeeper early, and definitely once inventory, staff, payables, and receivables pile up.
- AI bookkeeping agents reconcile at roughly 83 to 84 percent accuracy and get worse over time.
What net margin do Amazon sellers actually make?
Successful Amazon-only sellers now net low single digits, and that figure describes the sellers who are winning. Parag sees it across thousands of accounts on his platform, and fewer sellers survive each year.
More than half of Amazon sellers now sit outside the US, with growth concentrated internationally. Tax loophole changes are starting to level that field, and the structural shift already happened.
FBA removed the home-field advantage. US sellers used to win on delivery speed, and now an overseas seller ships through the same warehouses at the same speed.
Every competitive tool is available to every seller. Product research, green space finders, and price optimizers give nobody an edge once everybody runs them.
Sponsored results now overtake organic on Amazon, the same way they do on Google. Winning the Buy Box on price, ratings, and reviews used to be enough, and advertising is now the price of admission.
The old playbook is finished. Finding a few products cheap, adding a markup, pricing right, and shipping on time for good ratings no longer works as a business model.
Which Amazon fees are squeezing seller profits?
Storage, aged inventory, restocking, low inventory, and holiday capacity fees are squeezing profits, and they stack on top of referral and fulfillment charges. Parag notes Amazon adds fee types faster than its own documentation covers them, so merchants learn about a fee after it hits.
Onshoring inventory raised storage costs. Sellers moved product into FBA warehouses to reduce tariff exposure, and that demand pushed local storage prices up.
Inventory that sits gets penalized. Aged inventory and restocking fees apply to slow movers.
Inventory that runs thin gets penalized too. A low-inventory fee applies when you fail to occupy enough of Amazon’s capacity.
Holiday season carries its own surcharge. You compete with every other seller for peak-period warehouse space.
Tying any of this back to a single unit sold is genuinely hard. Parag describes sorting through spreadsheets and still struggling to connect a specific fee to a specific sale.
The value keeps moving up the chain. Only larger merchants can absorb the stack, which is quietly weeding out the small businesses that built themselves around marketplaces.
Where does ecommerce profit disappear?
Ecommerce profit disappears inside your sales channels, your payment processors, and your shipping costs, which is exactly where most books stop looking.
Here is how visibility breaks down across the expense categories Parag describes.
| Expense | Where it shows up | Visible in your bank account? |
|---|---|---|
| Cost of goods | Inventory records and purchase orders | Yes |
| Payroll | Payroll system on a fixed schedule | Yes |
| Advertising | Credit card or ACH, near real time | Yes |
| Marketplace and referral fees | Netted out before the deposit | No |
| Payment processing fees | Netted out before the deposit | No |
| Storage and inventory penalties | Inside the marketplace settlement | No |
| Refund and return fees | Inside the marketplace settlement | No |
| Shipping costs | Carrier accounts, separate from sales | Partially |
Algorithmic pricing makes the hidden half worse. When your prices move constantly, your margins move with them, and the fee impact per sale becomes a moving target.
Wholesale adds a timing problem on top. A B2B order arriving is not cash arriving, so you need accrual accounting and receivables tracking or your books will show money you have not collected.
What is the biggest ecommerce bookkeeping mistake?
The biggest mistake is treating the net deposit from a marketplace as your top-line revenue. Parag calls it the rookie mistake he sees most, and it hides every fee that was subtracted before the money arrived.
The deposit shock follows a predictable pattern. Sales look strong all month in the Amazon dashboard, then the 15-day payout lands and the number is unrecognizable.
Checking your bank balance for comfort is the same mistake in a different form. The balance is true, and it tells you nothing about which channel or product produced it.
Traditional bookkeepers compound the problem. Anyone unfamiliar with ecommerce will lump every Amazon charge into a single line, which makes the data useless for deciding anything.
Webgility runs on an internal mantra for this. Almost right is not good enough.
How do you set up ecommerce books by sales channel?
Create a separate clearing account for every channel you sell on, and net out each channel’s fees as its payout arrives. Your bank account then tells you the truth about money deposited, and each clearing account tells you what happened before that.
A Shopify payout is the clean example. The gross sales land in the Shopify clearing account, the fees net out against it, and what remains is what the channel actually earned.
QuickBooks can produce this view. If it does not, Parag’s position is that the setup is wrong, since the software is capable of it.
Month-end reconciliation is too slow to manage by. Waiting until the 10th or 15th of the following month means running blind for two weeks at a time.
I built my own version of this for Bumblebee Linens. A real-time contribution margin system, accurate short of the penny, that shows me what is making money and what is losing it without waiting for the monthly close.
Why does SKU-level profit tracking matter?
SKU-level tracking matters because it is the visibility that separates six-figure sellers from seven and eight. You cannot decide which products deserve more inventory and which quietly lose money when every fee arrives as one lump.
Bookkeepers rarely go this deep, and Parag is direct about why. They do not get paid enough to work at the SKU level, so they summarize.
Business owners set that expectation. Most are used to paying for compliance-grade books, which satisfy the tax return and nothing else.
Decision-grade books cost more and pay for themselves. Knowing profit by channel, order, customer, and SKU is where the competitive edge lives.
Sellers describe themselves by revenue tier. Profit is the number that determines what the business can become.
What should sellers focus on at each revenue stage?
Sellers should focus on product-market fit at six figures, channel optimization at seven, and scaling the engine past eight. Each stage has a different job and a different accounting need.
| Stage | Primary focus | Channel move | Accounting need |
|---|---|---|---|
| Six figures | Product-market fit and consistent revenue | Validate on a marketplace, or own site plus social if a brand is possible | Software syncing orders and inventory, a bookkeeper early |
| Seven figures | Repeat customers, CAC, LTV, catalog expansion | Second and third channels | Channel-level clearing accounts and SKU margins |
| Eight figures and up | Distribution, fulfillment footprint, bundling, campaigns | International, retail, big box | Multi-entity complexity, full outsourced bookkeeping |
Many six-figure sellers are still moonlighting. Parag sees plenty who have not fully committed, which is a fine place to start and a hard place to grow from.
Amazon is still the best validation channel. It is brutally competitive and remains the easiest place to test an early product and collect real feedback.
Treat that early Amazon phase as validation rather than income. This matches how I teach it: validate on Amazon, then build where you own the customer.
At seven figures the questions change. Where does the repeat customer base live, what does acquisition cost, what is lifetime value, and which second channel fits.
Past eight figures the levers are structural. Fulfillment footprint, product bundling, campaigns across media, and whether international or big-box retail fits the category.
Where are Amazon sellers expanding to?
Amazon sellers are expanding into Walmart, TikTok, social selling, retail pop-ups, and eventually big-box distribution. Parag sees all five across his platform, with Walmart showing tremendous growth as an expansion marketplace.
TikTok is producing real order volume. It ranks among the key channels for order growth across Webgility’s customer base.
Influencer and social marketing run mostly on Instagram and TikTok. That is where the top-of-funnel visibility work is happening.
Pop-ups replaced the permanent storefront. Brands take temporary retail locations without committing to a mall footprint.
Big box comes after seven figures. Brands aiming for eight and nine figures start building shelf relationships with major retailers.
My own store treats Amazon as a checkout option. I drive traffic from other channels and let buyers purchase wherever they prefer, which makes Amazon a fine fulfillment channel and a weak acquisition channel.
The buying journey has split into pieces. Discovery happens on TikTok and Instagram, research happens in ChatGPT, and the purchase happens on Amazon or wherever the buyer trusts.
When should you hire an ecommerce bookkeeper?
Hire a bookkeeper early, because your plate is full from day one and the books are the first thing that slips. Parag recommends putting a bookkeeper in place before complexity arrives.
Software carries the first stretch. Syncing orders, inventory, and prices into QuickBooks Online handles a business that is still simple.
Operational complexity is the trigger for a person. Running inventory and purchasing, employing staff, and managing payables and receivables is the point where you need software and a human together.
Webgility added full bookkeeping services this year. Parag describes it as the company’s biggest learning: putting all the accountability on the business owner was not working.
Can AI do ecommerce accounting?
AI cannot do ecommerce accounting reliably yet, and Parag builds accounting software for a living. AI reconciliation agents top out around 83 to 84 percent accuracy in the research he cites, and they degrade over time as they hit cases they cannot follow.
Financial data resists probabilistic tools. Run the same reconciliation twice and you can get two different answers, because there is no deterministic path.
Books that are 90 percent right are unusable. Accounting has to be fully correct or it is wrong.
Coding moved first for a reason. Code is already written in a language machines understand, and accounting is a mix of codified standards and human judgment on every new tag and category.
The useful layers are translation and workflow. Natural language lets non-accountants query their books, and agents handle categorization, reminders, and routine tasks, which is why Intuit and Xero are both adding them.
What matters to a small business is simpler than the AI question. Is it accurate, who is accountable, and what new value does it produce.
Insights come after clean data. The recommendation I want, “this product sells a ton and makes no money, raise the price,” only works when the books underneath are right.
Is agentic commerce worth chasing?
The catalog work is worth doing, and the shopping behavior is mostly hype. Parag’s position is that nobody asked for agentic commerce, because buying online was never the hard part.
Selling online is easy and making money doing it is hard. Agentic commerce optimizes the part that did not need optimizing.
The motive is advertising. The industry needs the ad pie to grow, and agents are how the narrative gets there.
Shopping is a human and social act. People want to see, touch, read reviews, and compare, which Parag credits for the return of the mall.
Repeat utility purchases are the real use case. Reordering kitchen towels and toilet paper fits an agent, and discovering a new brand does not.
My numbers back the skepticism. Sales attributable to AI tools sit in the low single digit percentages and are difficult to separate in analytics, so Shopify’s 3X AI traffic claim is tripling a very small base.
Clean the catalog anyway. Product pages and meta tags need to be parseable, site speed has to be solid, and you need a presence across social so agents and humans both find you.
Brand is the durable answer. Uniqueness, community, and product stickiness are the things an agent cannot commoditize.
Frequently asked questions
What is a realistic net margin for an Amazon seller?
Successful Amazon-only sellers now run net margins in the low single digits, based on aggregate data across thousands of accounts. Sellers earning more than that typically operate across multiple channels.
Why is my Amazon payout so much smaller than my sales?
Referral fees, fulfillment fees, storage, inventory penalties, payment processing, refunds, and returns are all subtracted before the deposit. The payout is what survived, and treating it as revenue hides every one of those charges.
Which Amazon fees surprise sellers most?
The biggest surprises are storage fees driven by onshored inventory, aged inventory and restocking fees, the low-inventory fee, and holiday capacity surcharges. Amazon adds fee types faster than its documentation covers them.
How should an ecommerce business structure its books?
Set up a separate clearing account for each sales channel, so fees net out per channel as payouts arrive. The bank account shows money deposited and each clearing account shows what happened to revenue before it got there.
Why track profit by SKU instead of by channel?
Channel totals cannot tell you which products make money and which lose it. SKU-level margin visibility is what separates six-figure sellers from seven and eight.
When does an ecommerce seller need a bookkeeper?
Hire one early, and without question once the business is managing inventory, purchasing, staff, payables, and receivables. Software handles the simple stage, and complexity is what makes a person necessary.
How accurate is AI bookkeeping?
AI bookkeeping agents reach roughly 83 to 84 percent accuracy on reconciliation at best, and that accuracy degrades over time. Books need to be fully correct to be usable, so human judgment still closes the gap.
Should sellers prepare for agentic commerce?
Clean your catalog and product data so AI can parse it, and keep investing in brand and community. Outside of repeat utility purchases, shopping remains a human decision.
