In-house fulfillment versus a 3PL: the real cost per order, on your numbers.

In-house fulfillment is cheaper than a 3PL while the labor and the space are close to free: a founder packing from a spare room, a few hundred orders a month. It stops being cheaper when the hours have to be paid for, the room has to be rented, and a wrong pick or a missed cutoff starts costing customers. To know which side you are on, price every in-house line per order, your own hours and error rate included, and set it next to a 3PL's rate lines on your real volume.

What in-house fulfillment really costs per order

Most founders price in-house fulfillment as postage plus boxes. That is two lines out of seven. Some of the seven grow with orders. Some sit there whether you ship or not.

• Labor: receiving, shelving, picking, packing, labeling, carrier handoff. Grows with orders, in whole shifts.
• Space: the room for stock, racks, the pack bench and the boxes. Fixed per month.
• Supplies: cartons, mailers, tape, dunnage, labels. Cheaper by the pallet than by the case.
• Software: labels, shipping rules, inventory counts. Fixed per month.
• Your own hours: the weekend pack, the lost parcel, the stock count. On no invoice, so left out.
• The error cost: a wrong size means a return label, a replacement, postage twice and an apology.
• Postage: per parcel, by service, weight, box size and destination.

Labor: the line most founders undercount

Labor is the heaviest in-house line after postage and the one priced most loosely. The wage is only the start. On top of it the employer pays Social Security tax at 6.2% of wages and Medicare tax at 1.45% (IRS, Social Security and Medicare withholding rates, checked October 2026), plus state unemployment tax and workers' compensation premiums.

For a baseline, the Bureau of Labor Statistics puts the May 2025 median hourly wage for hand laborers and material movers at $18.38, and the median annual wage for hand packers and packagers at $36,280 (BLS Occupational Outlook Handbook, checked October 2026). Say you pay $20 an hour. With employer taxes on top, each hour costs more than the wage before a carton is packed.

Then count the hours honestly, including the ones nobody books: training a new packer, the shift that runs out of orders by early afternoon. The pack itself takes minutes. The rest goes to receiving, counting, printing, opening returns and walking parcels to the carrier. Track every paid hour for a month and divide orders shipped by those hours.

Space, supplies and software

Space is fixed. The room costs the same in a slow month as in a busy one, and stock grows faster than sales when you buy ahead of a season. Storage billed per cubic foot falls as you sell through. In-house, it falls when you move.

Supplies cost more by the case than by the pallet, and the box you choose sets part of the postage. USPS prices Priority Mail parcels larger than 1 cubic foot on the greater of actual weight and dimensional weight, length times width times height in inches divided by 139 (USPS Domestic Mail Manual, Priority Mail, checked October 2026). A carton with air in it is a postage mistake.

Software is the smallest line. You can buy, print and manage shipping labels from Shopify's supported carriers in the Shopify admin (Shopify Help Center, checked October 2026). Bin locations, scan verification, cycle counts and reorder triggers come from a warehouse system: a monthly line plus scanners and a thermal printer.

Your hours and the error cost

Your own hours never appear on an invoice, and they decide most of these comparisons. Count them for a month: the weekend pack, the stock count, the return opened on the kitchen table. USPS picks up eligible packages free during your regular mail delivery and charges for a pickup at a time you choose (USPS Schedule a Pickup, checked October 2026). Someone has the parcels ready when the truck comes, every business day.

The error cost is the sharpest line. Say 20 of every 1,000 orders go out wrong. That is 20 reships a month, each a return label, a replacement unit, postage twice and the time. Across our floor pick and pack error runs under 0.3%, because every unit is scan-verified against the order before it goes in the box. On the same 1,000 orders that is fewer than 3. The gap is a line in the model.

When in-house still wins

In-house wins in four situations, and a 3PL is not the answer for every brand.

• Low volume with free hands. A few hundred orders a month, packed by a founder or family in time nobody would otherwise pay for. Nothing beats a labor line near zero.
• Products that need the maker. Made-to-order or hand-finished pieces, inspected by the person who made them.
• A building you already run for wholesale or retail, with spare hours on the floor. Space and people are already paid for.
• Early days. Your first thousand orders teach you what breaks in transit and what comes back, lessons that become your packing standards later.

When in-house stops winning

The switch point is not one order count. It is the month one of these becomes true.

• You are hiring for it. The first packer on payroll makes labor real money, with employer taxes on top, in whole shifts.
• You are renting for it, or the room you ship from could earn money another way.
• The cutoff slips. Morning orders ship the next day because nobody got to them, and the customer emails start.
• Peak is a problem rather than a month to enjoy. A sale or a creator post becomes a backlog you dig out of for days.
• You sell on more than one channel and keep inventory straight by hand between your store, Amazon and TikTok Shop.
• You can name the hours of your week that fulfillment takes, and you want them back for product, buying or marketing.
• Returns are a pile. They sit unopened because opening them is not an order, and refunds wait.

300 or more orders a month is where our rate card works well, and it is near where the in-house lines stop being free. Below that, run the model anyway. The answer may still be in-house.

How to run the comparison on your own numbers

1. 01Pull one normal month: orders shipped, units per order, returns received, and the cubic feet your stock takes, from carton dimensions times carton count.
2. 02Log every paid hour that touched fulfillment, receiving through returns. Multiply by the wage plus the employer's share of Social Security and Medicare tax.
3. 03Log your own hours on the same tasks. Price them at what you would pay someone to take them, not at zero.
4. 04Price the space at what the room rents for, or would rent for.
5. 05Add the monthly software line and the supplies you bought, from receipts.
6. 06Count the orders reshipped. Cost each at the postage twice, the replacement unit and the time.
7. 07Divide each line by orders shipped for an in-house cost per order, with postage on its own line at what you paid. The cost per order guide runs the same formula on our rates.
8. 08Put the same month into a 3PL rate card. The cost calculator runs ours in a minute, and the pricing guide explains each line.

On direct cost the in-house month comes in close, and in a room you already pay for it can come in lower. Missing from the in-house total: 20 hours of your own time, 20 reships and a peak weekend. Missing from the YNK total: receiving hours and packaging as used. Run your own inputs through the cost calculator. If it tips toward in-house, you have a number instead of a feeling. Weighing more than one 3PL? Compare the quotes on the same month.

The same month at YNK

At YNK the in-house lines map onto a rate card with seven lines and nothing behind them: handling $3.00 per order, storage $0.65 per cubic foot per month, returns $3.00 per parcel, receiving billed hourly, packaging and supplies billed as used, special-project labor $35 per hour only when you ask for it, and postage billed semi-monthly per shipment at extremely competitive carrier rates. One handling fee per order is the baseline, whatever the unit count. There is no setup fee and no seasonal or peak surcharge on handling or storage. The rates are estimates, confirmed in your quote; the full rate card has every line.

The lines in-house cannot price are the ones this floor was built around. Orders in by 12pm ET ship the same business day, and the SLA is one business day, two to three in peak. Pick and pack error runs under 0.3% across 1.2M+ orders a year. Shopify, Amazon and TikTok Shop connect natively and draw on one inventory pool, and the live dashboard shows inventory by SKU, order status and SLA. Labor is a team of 50+, so a sale weekend is staffed rather than survived: Nominal's last-minute sale put 10,000+ orders on the floor in one weekend, every one shipped within SLA with no surcharge on handling or storage.

We have run both sides of this table. YNK began as the in-house fulfillment of Veiled, the founders' own brand, which still ships from this floor in Elmwood Park, NJ, about 15 miles from Manhattan. Ground from here reaches the Northeast in 1 to 2 business days and the rest of the country in about 5 business days, one more line a spare room cannot change.

Questions founders ask about in-house versus a 3PL