Storage fees climb, holiday peak fulfillment fees kick in, returns creep up, and capital gets tighter just as the buying windows get shorter. The cost of guessing wrong multiplies fast.

A SKU that quietly loses money in Q3 becomes an SKU that loses money at volume in November and December. That’s a much more expensive mistake to make twice.

This is exactly the moment where “revenue looks great” and “profit looks great” have to describe the same SKU, on the same report… not two different stories hiding in two different spreadsheets.

Replens are where this gap shows up the most. A good replen keeps paying you all season. A bad one keeps tying up cash and racking up fees.

This article is a working system for deciding which SKUs actually earn a reorder, especially heading into Q4.

Quick refresher: what are “replens” in OA?

In Amazon online arbitrage, a replen is any SKU you can:

• Source repeatedly at a similar landed cost
• Sell consistently at a profit
• Restock before you fully run out, without needing a brand-new “hunt” every time

Replens are the opposite of one‑off flips. Instead of constantly scanning for “the next deal,” you’re building a base of SKUs you know, trust, and can rebuy.

Why they matter:

• Predictability: You can plan cash flow around SKUs that sell again and again.
• Speed: You spend less time hunting and more time placing smart, repeat orders.
• Compounding: Each good replen you add is another brick in a business that doesn’t reset to zero every month.

But “replen” doesn’t mean “always reorder.” It only means “can be reordered.” The system below helps you decide which ones should be.

Before you reorder: what you actually need to track

You don’t need a PhD in analytics to manage replens. You do need a single, honest view per SKU.

At minimum, for each candidate replen, you should be able to see:

• Units sold (last 30 / 60 / 90 days)
• Average sale price
• Buy cost per unit (including tax, shipping to you / prep center, and prep cost)
• Amazon’s per‑unit fees (referral, FBA fulfillment, any extra surcharges)
• Storage impact (is it small and fast… or bulky and slow?)
• Return rate and the cost of those returns
• Net profit per unit
• Total net profit over the last period
• Rough “days of cover” (how fast it sells vs how much you keep on hand)

That’s it. If you can get those nine things in one place, you have enough to stop guessing.

A lot of “busy but broke” OA businesses are just running a sales report, not a profit report. Replen decisions should be made from the latter.

The Replens Reorder Framework: from sales to reorder priority

Here’s the simple ladder we use to get from “this SKU sold” to “this SKU actually earns a reorder in Q4.”

1. Sales: the starting line, not the finish line

This is your gross revenue: units sold × sale price.

It’s the number that shows up first, feels the best, and tells you the least.

A SKU that did $5,000 in sales might be a hero or a villain. Sales tell you where to look, not what to do.

How to use it:

Filter for SKUs that actually moved units in the last 60–90 days. That’s your shortlist, not your winner circle.

Questions to ask:

• Did this SKU sell consistently, or in one weird spike?
• Is demand still there, or did you just ride a promotion that’s gone?

2. Costs: what it really took to land the product

Now we get into what it actually cost to get each unit sitting in an Amazon bin:

• Product cost
• Sales tax (if any)
• Inbound shipping to you or your prep center
• Inbound shipping into FBA (including placement quirks if they apply)

This is where “cheap” products quietly become expensive. A $4 item with ugly prep and high inbound can behave more costly than an $8 item that ships clean.

How to use it:

Compute a true landed cost per unit. Not a guess. Not “about five bucks.” A real number.

If you don’t like what you see at this step, don’t bother going further. Reordering something that’s already expensive to land is how you end up cash‑starved when Q4 fees jump.

3. Fees: Amazon’s cut of the action

Next, layer in Amazon’s tolls:

• Referral fee
• FBA fulfillment fee (by size/weight)
• Storage (especially as Q4 peak storage and fulfillment fees ramp up)
• Any extra category‑specific or program fees

Amazon’s fee structure is not a rounding error. For third‑party sellers, this is a major, structural cost of doing business, and it moves around based on size tier, category, and time of year.

How to use it:

• Run each candidate SKU through a profit calculator with current fees.
• Pay special attention to: bulky items, borderline size tiers, and SKUs that sit more than 60–90 days.

Some SKUs are fine in Q2 and Q3 but get hammered when Q4 storage and peak fees turn up the pressure. Those often lose their right to be a replen in Q4.

4. Returns: the revenue that boomerangs back

A return doesn’t just erase a sale. It often brings:

• A refund processing fee
• A restocking fee
• A damaged or unsellable unit that you now have to liquidate or dispose of
• Extra labor/time managing tickets, photos, and appeals

High‑return SKUs can look like winners on a sales report and behave like liabilities on a P&L.

How to use it:

For each candidate replen, ask:

• What percentage of units sold came back as returns?
• Of those returns, how many were:
• Resellable at full price?
• Must be moved through other channels (eBay, etc…)
• Total write‑offs?

If an SKU’s “real” margin collapses once you account for returns, it’s not a Q4 replen. It’s a Q4 landmine.

5. Profit: the number that actually matters

Now we can calculate what matters:

Net profit per unit= sale price – landed cost – Amazon fees – expected return cost per unit

Then:

Total profit = net profit per unit × units sold

This is the only number that tells you whether reordering this SKU makes you money or just makes you busy.

Two SKUs with the same ROI can behave very differently in Q4:

• A $4 profit that turns 10 times per month
• vs a $12 profit that turns once every two months

In a fee‑heavy, time‑compressed part of the year, the faster payback often wins.

How to use it:

• Flag SKUs that deliver strong net profit per unit and solid total profit in a recent window.
• De‑prioritize SKUs that only look good because they had one weird pop or required deep discounting to move.

6. Reorder Priority: where the decision actually gets made

Once you’ve done the work above, you can finally do the simple thing that most sellers skip:

Rank your SKUs by how much money they actually make you.

Not by sales.

Not by estimates.

Not by “I really like this brand.”

By profit.

A simple Q4 rule of thumb:

1. Sort your SKUs by total net profit over the last 60–90 days.

2. Start at the top of the list.

3. Working down, decide:

• “Yes, reorder and consider ordering a bit deeper.”
• “Yes, reorder but keep the same depth.”
• “No, pause this SKU for Q4.”

4. Stop when your Q4 inventory budget is spoken for.

Turning this into a real system (instead of a one‑time spreadsheet)

Everything above works in a spreadsheet. But if you find it too tedious and time-consuming, you won’t keep doing it.

The goal is to make this easy enough that you run it every cycle, not just once a year when something breaks.

Here’s one way to put it on rails using two tools many in our community already use: Saturn 44 and ArbiSource.

Step 1: Keep a steady stream of replen candidates (Saturn 44)

Replens start as good leads. You can’t reorder what you never find in the first place.

That’s where an OA lead list like Saturn 44 helps:

• 10+ fresh OA leads daily
• Mostly ungated or lightly gated brands in the most common categories for replens (Grocery, Beauty and Personal Care, Health & Household,etc…), designed for newer and intermediate accounts
• Filtered for realistic profit and velocity, not just pretty ROI screenshots

In practice, Saturn 44 becomes your candidate pool:

1. You cherry‑pick the leads that match your criteria and testing budget.

2. Some of those tests become winners.

3. A subset of those winners become true replens you want to keep on the shelf.

Over time, you’re not just chasing today’s deals. You’re building a bench of SKUs that qualify for the framework you just read.

Learn more about Saturn 44

Step 2: Put your replens on “radar” instead of manually re‑hunting (ArbiSource)

Once you know which ASINs are worth reordering, the next problem is simple and annoying:

• The supplier goes out of stock.
• Or the price drifts up.
• Or you just don’t feel like checking twenty tabs every week.

This is where ArbiSource becomes the star of the system:

• You upload a sheet of ASINs (hundreds or even thousands at a time).
• ArbiSource ties those ASINs to source links and tracks:
• Supplier in‑stock / out‑of‑stock status
• Current price vs your target buy price
• When a replen pops back into your buy zone, you get notified.

Instead of re‑sourcing the same products over and over, you:

• Use a tool like SellerAmp (fully integrates with ArbiSource) in your normal workflow to vet leads up front.
• When a SKU passes your tests and fits your Q4 framework, you add its ASIN to your ArbiSource list.
• ArbiSourcewatches your “replen radar” for you and pings you when it’s time to buy again at your preferred price points.

No more “I forgot to check that” or “I didn’t realize it was back in stock until it was gone again.”

If you want to try it, there’s a coupon code for our readers:

FBALEADLIST30 for 30% off your first month of ArbiSource.

They also offer a free trial.

Learn more about ArbiSource

How to put this to work this week

If you’re looking to jumpstart Q4 prep with replens and wondering what to do next, here’s a simple 3‑day sprint:

Day 1: Pull the truth.

Export your last 60–90 days of sales and build a simple SKU‑level view with:

• Units sold
• Average sale price
• Landed cost
• Amazon fees
• Return rate
• Net profit per unit and total

Day 2: Rank and cut.

Apply the framework:

• Eliminate obvious losers (high returns, thin or negative profit, slow movers).
• Highlight your top 10–20 SKUs by total profit.
• Decide where each one sits on the “order deeper / same / pause for Q4” spectrum.

Day 3: Put winners on rails.

• Add your top replens into a system like ArbiSource so you stop manually babysitting them.
• Keep using a daily lead source like Saturn 44 to find and test new candidates, knowing you now have a clear way to decide which lead graduates into your replen list.

Do that once, and you’ll feel more in control.

Do it every cycle, and you’ll build a Q4 that isn’t just big on revenue… it’s honest on profit.

That’s the whole point of replens. Not just “more sales,” but more sales that deserve to come back into your inventory again and again.

Some links may be affiliate links. We may get paid if you buy something or take an action after clicking one of these. We appreciate the support.