If you were to ask fulfillment providers when you should outsource and you’ll get many different answers.
3PLGuys puts the line at 50 orders a day. Warpspeed says 150 a day. DCL Corp says 200 to 500 orders a month. Bolian Logistics says 500 a month. Lite Fulfillment argues that below 1,500 monthly orders, outsourcing almost always wins on cost. Atomix Logistics has published two pieces with two different thresholds, one at 100 to 300 orders a month and another at 300 to 500. This a fifteenfold spread on the same question, and every source quoting a number sells the service.
Order count is a lagging indicator. It tells you how busy you were last month. It says nothing about whether your operation is about to buckle, whether your shipping costs are drifting out of your control, or whether you’re losing orders at checkout to a competitor who ships faster. Those are the things that decide it.
Who are we? Speed Commerce is an end-to-end provider of outsourced customer experience solutions for eCommerce retailers (including for BigCommerce & more) as well as manufacturers, for close to 20 years. We grow our clients’ businesses by providing winning customer experience strategies such as 24/7/365 eCommerce customer service, order fulfillment, and warehousing – get a free quote from a fulfillment expert. Refer to our guide on crowdfunding fulfillment, and east coast fulfillment updated for 2026.
Why does everyone quote a different threshold?
Because the threshold moves depending on what you sell.
A company shipping 200 lightweight, single-SKU orders a month out of a spare room has a different math problem than one shipping 200 orders that each contain six components, cross a border, and require serialized tracking. Same volume, completely different operation.
Product size, SKU count, order composition, return rates, channel mix, and seasonality all shift the point where in-house fulfillment stops paying for itself. One number can’t hold all of that, which is why the published numbers disagree so wildly. You should treat volume as the trigger to run the analysis, not as the answer.
Your cost per order should fall as you grow
This is the most reliable signal available, and most companies never track it.
Fulfillment has economies of size built in. As you push more orders through the same space, the same equipment, and the same software, fixed costs spread across more units and cost per order comes down. That’s how it’s supposed to work.
When it goes the other way, something has broken. Rising cost per order at rising volume means you’ve passed the point where your setup fits your business. You’re paying overtime to cover shifts you can’t staff. You’re expediting because you missed a cutoff. You’re eating replacement costs on mis-picks. You’re renting overflow space by the month at rates you’d never sign a lease at.
Pull twelve months of total fulfillment cost with everything in it: labor including payroll taxes and benefits, rent and NNN charges, packaging, software, equipment, damages, and your full carrier spend. Divide each month by orders shipped. If that line climbs while volume climbs, you have your answer.
Worth checking: industry estimates put warehouse labor at 45% to 57% of total operating costs. That’s before recruiting, training, and the productivity loss while new hires get up to speed, none of which show up on a payroll report as fulfillment expense.
Can you reach your customers in two days from where you sit?
Consumer patience has settled onto a hard floor. Around 74% of online shoppers now expect delivery within two days, and 86% define fast shipping as two days or fewer. Roughly 43% say they’ve abandoned a cart or walked away from a retailer entirely because shipping took too long.
DHL’s 2026 E-Commerce Trends Report, built on responses from 29,000 shoppers across 29 countries, found that 67% have abandoned a cart because of the delivery options in front of them. Cost, speed, and choice all count, and friction in any one of them can end the sale.
There’s a structural problem. One warehouse can’t cover the United States in two days by ground. Depending on where you sit, a single location reaches maybe half to two-thirds of the population inside that window. Everything past it either arrives late or moves by air at a cost that eats the margin on the order.
Two-day coverage means multiple nodes, and multiple nodes multiply every fixed cost you already carry: lease, staff, systems, supervision. That’s a different order of commitment than renting one building, and it’s usually the moment a distributed network you rent by the order starts looking sensible next to one you build.
Another wrinkle is that accuracy moves returns, not just speed. Capital One Shopping’s research found that every day a delivery arrives late correlates with a 1.1% rise in returns, and every day it arrives unexpectedly early correlates with a 1.2% rise. Hitting the promised date is its own line item.
How much of your shipping bill do you still control?
Less than you did, and less every year.
UPS and FedEx both applied general rate increases of about 5.9% for 2026. That headline undersells what most shippers felt. Once accessorial charges stack up, effective increases have been landing in the 8% to 20% range. USPS layered an 8% surcharge on Priority Mail, Priority Mail Express, and Ground Advantage running April 26, 2026 through January 17, 2027.
The accessorials are where the damage happens. Residential delivery fees run $4 to $6 per package, and more than 90% of ecommerce parcels go to homes. Both carriers tightened dimensional and oversize thresholds so they track cubic volume more closely, which punishes anyone using one box size for everything. A single shipment triggering residential, additional handling, fuel, delivery area, and declared value charges can carry $40 to $300 or more in accessorials alone, depending on size and zone.
Volume buys leverage in carrier negotiations, and a provider aggregating thousands of shippers has more of it than you do on your own. That advantage is real. Pressure-test any specific savings figure a provider quotes you against your own zone and weight profile, though, rather than taking a headline number on faith.
What happens at your next lease renewal?
Fixed costs don’t rise smoothly. They move in steps.
You absorb growth in your current building until you can’t, and then you sign a bigger lease, add a second shift, and buy a warehouse management system more or less at once. That step is where in-house fulfillment gets expensive fast, and it’s worth modeling well before you’re standing in front of it.
The current market makes that step harder than the headlines suggest. National industrial vacancy has climbed to roughly 7.5% to 8.8%, which reads like a tenant’s market. For most growing companies, it isn’t. That number is driven by big-box space delivering after the 2023 to 2024 construction wave. Small-bay product under 50,000 square feet sits at 4.8% or lower, the tightest of any size segment. Atlanta shows the split clearly: overall vacancy around 8.1%, small-bay near 4%.
CommercialCafe’s May 2026 data puts national in-place rents at $9.12 per square foot, up 5.2% year over year, with asking rents closer to $10.20. NNN charges add $1 to $3 on top. Regional spread is enormous, from roughly $5.28 in Kansas City and $6.57 in Chicago up to $16 to $22 in Los Angeles.
Software is the other half of the step. Cloud warehouse management systems run roughly $100 to $500 per user per month. CPCON puts first-year all-in cost for a small operation under 5,000 SKUs at $25,000 to $75,000 once you count implementation, training, hardware, and twelve months of subscription. On-premise systems run $100,000 to $500,000. Scanners, label printers, and handhelds add thousands more.
If your growth curve puts you in front of that combined step inside the next twelve to eighteen months, run the comparison now instead of after you’ve signed.
Customs paperwork stopped being optional
This trigger is new, and plenty of companies haven’t priced it in yet.
The de minimis exemption, which let imports valued under $800 enter the United States duty free, has been suspended since August 29, 2025. The February 2026 Supreme Court ruling on IEEPA tariffs didn’t restore it, because the suspension rests on separate legal authority. On June 24, 2026, CBP moved the suspension from executive-order policy into permanent regulation, and the One Big Beautiful Bill Act codifies permanent elimination starting July 1, 2027. A 10% Section 122 global surcharge took effect February 24, 2026.
Operationally, that means every commercial import now needs formal customs entry and 10-digit HTS classification regardless of value. Landed cost has to carry duties on every unit. Classification errors carry real consequences.
If you import anything, customs competence just became part of your fulfillment operation. Most in-house teams don’t have it, and hiring for it is expensive. This is a column in the analysis that didn’t exist two years ago, and it pushes the answer toward outsourcing for anyone with meaningful import volume.
Where does your leadership team spend its hours?
This is the softest signal, and sometimes the most telling one. If your ops lead spends mornings on labor scheduling, your founder packs boxes during promotions, and somebody senior handles carrier claims every week, fulfillment has stopped being a back-office function. It’s now the constraint on everything else you’re trying to do.
Count those hours honestly for two weeks, interruptions included: the stockout call, the mis-ship escalation, the carrier dispute, the seasonal hiring push. Then price them at what that person’s time is worth to the business. Most teams find the total uncomfortable.
There’s a related signal on the customer side. When order problems rise, support volume rises with them, and shipping questions are among the most expensive contacts you’ll handle. If your team spends its days answering “where is my order,” you’re paying for fulfillment failures twice, once in the warehouse and once at the service desk.
When is staying in-house the right call?
Plenty of the time, and you won’t hear that from most providers.
Stay in-house if your volume is low and flat. Below a couple hundred orders a month with no real growth trajectory, per-order rates won’t beat what you’re doing now and you’ll add coordination overhead on top.
Stay in-house if the unboxing is the product. Hand-written notes, custom assembly, influencer kits, personalization at the pack station. When the way an order arrives is part of why customers buy, handing that to a third party costs you the thing you’re selling.
Stay in-house if you hold the licensing. Certain regulated categories, controlled substances, and specialized handling requirements can make outsourcing slower and costlier than doing it yourself.
Stay in-house if your order profile is genuinely simple: few SKUs, one channel, predictable volume, regional customers, light packages. Simple operations run cheaply out of a small space with two people, and no provider will beat that.
A hybrid setup suits a lot of companies too. Route standard orders to a provider and keep a small in-house operation for VIP shipments, returns triage, and anything that needs a human touch.
Running the comparison
Build your in-house number honestly before you compare it to anything. Total twelve months of rent and NNN, labor with taxes and benefits, software, equipment, packaging, damages and shrinkage, and your complete carrier spend including accessorials. Divide by orders shipped.
Then get quoted on your real order profile rather than an average. Hand a provider your SKU list, order composition, seasonality curve, and destination mix. A quote built on your own data looks nothing like the one on a pricing page.
Compare on total landed cost per order, not on pick-and-pack fees. Then put a number on what doesn’t appear in either column: the delivery speed you’d gain, the leadership hours you’d get back, the step costs you’d sidestep.
If the two figures land close together, the tiebreaker is usually what you’d do with the capacity you free up. We’d rather you run those numbers and decide to stay in-house than sign with anybody, us included, because of a threshold you read on a blog.
