Prep center and 3PL software: buy, build, or buy the core and build the gaps?
Most FBA prep centers and small 3PLs should buy the core (receiving, labeling, storage and a client portal) and build only the gaps a subscription cannot hold, such as client-specific billing rules, a niche prep flow or a channel it does not connect. Build the whole system only when your workflow is the product you sell. Decide job by job, and total three years of each option.
- Since January 1, 2026, Amazon no longer offers prep and item labeling for FBA shipments in the US, and AMAZON is no longer accepted as prep or label owner in the Fulfillment Inbound API.
- Amazon's current Fulfillment Inbound API is version 2024-03-20; plans created through it can be opened in Send to Amazon once placement and transportation are confirmed.
- A public SP-API app that is not listed in Amazon's Selling Partner Appstore is limited to 25 seller authorizations, and sellers must reauthorize every app every 365 days.
- The decision matrix covers 10 prep and 3PL jobs, from receiving to client authorization, with what to ask a vendor for each.
- Ecomsellertool built EzBookPrep's prep system to replace the Quickbase setup its owner had outgrown, and Little Owl's two-sided prep-center software.
This guide is for owners and operations leads of Amazon FBA prep centers and small 3PLs who are choosing between prep or warehouse software they subscribe to and software built for them. The honest answer for most is a mix: buy the core, build the gaps. The matrix below goes job by job so you can see which is which in your own building, and the worksheet after it puts three years of each option on one page. If you want to compare named prep products first, our page for 3PLs and prep centers quotes three of them in their own words.
Should a prep center buy or build its software?
Buy first. Prep and 3PL software is a crowded, mature category, and a subscription gives you receiving, FNSKU labels, storage and a client portal in weeks, with someone else paying for Amazon's API changes. A full custom build makes sense only when the way you work is part of what you sell, for example a grading flow for used books, a client experience under your own brand, or billing that no product can express.
Most prep centers end up in the middle. They run a product on the warehouse floor and pay for one or two custom pieces around it: a billing export that applies each client's contract, an order feed from a channel the product does not connect, or a report that compares their counts with Amazon's. That hybrid only works if the product has an API or reliable exports, which is why it is the first question to ask any vendor.
Which prep-center jobs should you buy, build or integrate?
Take each job your team does in a week and read across its row. "Buy" means a standard prep or 3PL product usually covers it. "Build or integrate" means custom code, either a full module or a connector around the product you already run. The last column is what to ask a vendor in the demo, so the answer comes from their screen, not their brochure.
The buy-vs-build decision matrix
| Job | Buy usually fits when | Build or integrate when | What to ask a vendor |
|---|---|---|---|
| Receiving | Clients announce shipments in advance and you receive against them by SKU and carton | You receive by something other than SKU (ISBN, lot, condition grade), or you blind-receive a lot and need photos of damage logged before prep starts | "Show me a blind receipt, damage photos and a discrepancy sent to the client." |
| Prep and FNSKU labeling | Prep steps are standard per SKU and labels print from the client's Amazon data | Each client has its own prep rules, or you need a per-unit record of who prepped what as evidence for reimbursement disputes | "Do you print FNSKU labels from Amazon's own label call, on thermal printers, and store prep steps per SKU per client?" |
| Kitting and bundling | Bundles are fixed and built to a bill of materials | Kits change per order or per client promotion, or a bundle has to stay in sync with listings on several channels | "Can a client define a kit, and is the kit's stock worked out from its components?" |
| FBA shipment creation | You create shipments inside each client's account through the product, and it is current with Amazon's inbound API | You need a step the product does not offer, such as your own carton-packing rules or placement choices made against a client's budget | "Does shipment creation run on Fulfillment Inbound v2024-03-20, and can a plan you start be finished in Send to Amazon?" |
| Client portal | Clients need to add shipments, see stock and approve work | The portal is part of your brand and sales pitch, or clients need views the product cannot show | "What can a client see and do without calling us, and can the portal carry our brand?" |
| Client billing | You bill per unit, per label and by storage volume, with a few rate cards | Contracts carry minimums, tiers, client-specific services or bundled prices the product cannot hold | "Show me an invoice line traced back to the scans that created it, and a client with a custom rate card." |
| Multichannel order sync | Clients sell on the channels the product already connects | A client's channel, marketplace or wholesale buyer is not on the list, or orders arrive as email or spreadsheets | "Which channels do you connect today, who maintains each connector, and how is tracking written back?" |
| Inventory reconciliation | You only need your own on-hand counts per client | You must match your shipped counts against what Amazon received, and chase the gap per shipment and client | "Can I see shipped versus Amazon-received units per shipment, with the open gap for each client?" |
| Reporting | The standard dashboards answer the questions clients and your team ask | Owners need margin per client, labor per job or a report in a format a client requires | "Can I export raw events by API on a schedule, not only CSV by hand?" |
| Client onboarding and Amazon authorization | The vendor's app is already authorized by sellers and handles renewals | You run your own Amazon app for many clients, which brings registration, authorization limits and yearly renewals (see below) | "Whose Amazon app do my clients authorize, what happens when they must renew it, and what data does it keep?" |
Read the matrix by counting rows. If every job you run sits comfortably in the "buy" column, buy, and spend nothing on custom software. If two or three rows land in "build or integrate", buy the core and price those rows as integrations. If most rows land there, or the rows that do are the ones your clients pay you for, a full build deserves a quote. Billing and multichannel order sync are where custom work most often pays first, because a gap there costs money every month.
What changed at Amazon that affects the decision?
Prep moved fully onto your side. Amazon's changelog notice on the end of US FBA prep and labeling services says that from January 1, 2026, prep and item labeling services for FBA shipments are no longer available in the US, including inventory that reaches FBA through AWD, Amazon Global Logistics, Amazon SEND or Supply Chain Portal (checked October 9, 2026). The same notice says AMAZON is no longer an accepted value for prepOwner or labelOwner in the Fulfillment Inbound API. Any software you buy or build must send the seller as prep and label owner.
Shipment creation also changed shape. Amazon's Fulfillment Inbound API page lists version 2024-03-20 as current and v0 as legacy (checked October 9, 2026). A send-in now moves through an inbound plan, packing options, placement options and transportation options, and two v0 operations, getLabels and getBillOfLading, are still needed to print box labels, pallet labels and bills of lading. Plans created through the API can be opened in Send to Amazon once placement and transportation are confirmed; switch earlier and the options chosen so far are discarded.
Send to Amazon itself is the workflow your team already knows from Seller Central. Amazon's Send to Amazon article describes reusable case pack templates that save box contents, weight, dimensions and prep and labeling details for single-SKU boxes (checked October 9, 2026). For a prep center with a handful of steady clients, doing shipments there, with your software only tracking the work, is a reasonable choice. Software earns its keep when volume makes clicking through each client's account the bottleneck.
Item labels can come straight from Amazon. The createMarketplaceItemLabels operation creates labels for a list of seller SKUs, on standard sheets or on thermal labels (checked October 9, 2026). Ask any vendor whether its labels come from this call or from a barcode it builds itself, because a label generated from Amazon's own data is one less place for a retyped number to go wrong.
What does a custom build have to handle that a subscription already has?
A prep center serves many sellers, so a custom system needs its own Amazon app that each client authorizes. That is where many build plans get more expensive than expected. Amazon's SP-API registration overview says apps for other sellers are public apps authorized through OAuth, that your security team answers Amazon's security control questions, and that you must follow its Acceptable Use Policy and Data Protection Policy (checked October 9, 2026). Our SP-API glossary entry explains the terms.
Authorization limits matter at prep-center scale. Amazon's application authorization limits page says a public seller app that is not listed in the Selling Partner Appstore can hold at most 25 OAuth authorizations, while a listed app has no limit, and a private app is limited to 10 self-authorizations (checked October 9, 2026). A prep center with 40 clients that builds its own app either lists it in the Appstore or runs into that ceiling. Plan for it before the build, not after client 25.
Authorizations also expire. Amazon's reauthorization page says sellers must reauthorize an app every 365 days, and again whenever the app adds a role (checked October 9, 2026). Your system needs to warn you, and the client, before a renewal lapses, or shipments stop at the worst moment. Roles matter too: Amazon's roles page lists an Amazon Fulfillment role for fulfillment operations and marks Direct-to-Consumer Shipping as restricted because it involves buyers' personal data (checked October 9, 2026).
Client stores bring the same work on other platforms. On Shopify, a 3PL connects as a fulfillment service app; Shopify's fulfillment service documentation says the app receives fulfillment requests from merchants and can accept or reject them (checked October 9, 2026). Every channel you connect yourself is a connector you maintain. A subscription spreads that upkeep over all of its customers; a build puts it on your invoice. That is the strongest argument for buying, and it is a fair one.
How do you compare the total cost of ownership?
Put three options side by side over three years: buy, build, and the hybrid of buying the core and integrating the gaps. Three years, because a build front-loads its cost and a subscription spreads it, so one year flatters the subscription and five years flatters the build. Every option also carries the hours your team still spends working around what the software cannot do; those hours are usually the line that decides.
3-year cost = up-front cost + monthly running cost × 36 + workaround hours a month × 36 × loaded hourly cost
| Input | Buy | Build | Hybrid (buy core, integrate gaps) |
|---|---|---|---|
| Up-front cost | Setup, onboarding and data migration fees | The fixed build price | Setup fees plus the integration build price |
| Monthly running cost | Subscription, including fees that grow with clients, orders or users | Hosting, monitoring and maintenance hours at the builder's rate | Subscription plus connector upkeep |
| Workaround hours a month | Manual work in the jobs the product does not cover | What stays manual by design, such as approvals | Manual work left after the integrations |
| Your team's time | Training and setup | Writing the requirements, testing on real data, running beside the old process | Testing the integrations |
| Exit cost | Moving data out when you switch | Changing builders if the first one leaves | Both, smaller |
Your loaded hourly cost is pay plus benefits, divided by hours worked. Price fees that grow with clients or orders at the size you expect in year three, not today's, because that is where subscriptions climb. Amazon's guide to 3PL pricing lists receiving, volume-based storage, picking, packing, shipping and returns among the charges 3PLs bill (checked October 9, 2026); count how many of your billable events the product records on its own, and treat the rest as workaround hours.
A worked example (illustrative)
The numbers below are illustrative, not a client result and not anyone's price. Picture a prep center with 30 clients that already runs a prep subscription and is asking whether to replace it. Its workarounds, timed for two weeks:
| Workaround today | Hours a week (illustrative) |
|---|---|
| Rebuilding month-end invoices for clients with custom rate cards | 5 |
| Retyping orders from two client channels the product does not connect | 4 |
| Comparing shipped units with Amazon's received counts in a spreadsheet | 3 |
| Answering client emails about stock the portal does not show | 2 |
| Total | 14 |
Fourteen hours a week is about 60 hours a month, or 2,160 hours over three years. At a $25 loaded hourly cost, a placeholder, that is $54,000 of workaround time over three years. If integrations for billing, the two channels and the reconciliation report remove 12 of those 14 hours, they free about $46,000 of time over three years. That figure is the ceiling for the hybrid: if the integrations plus three years of their upkeep cost less, the hybrid wins against staying as you are.
A full rebuild has to clear a higher bar. It saves the same workaround hours and the subscription, but takes on hosting, maintenance, Amazon app registration, yearly reauthorizations and every channel connector. On these illustrative inputs, the hybrid usually wins unless the subscription itself is large or the product blocks the integrations. Replace every number with your own; the shape of the comparison is the useful part.
When is buying the better choice?
Often. Say it plainly to yourself before anyone quotes you a build: if your work is standard Amazon prep, your billing fits the product's settings, your clients sell on channels it already connects, and nobody on your team would own a custom system, buy. A build in that case mostly buys you maintenance. The same holds if you are still finding your service mix, because software built around this year's workflow may not fit next year's.
- Your prep, labeling and storage steps match what the product does in a demo with your own SKUs
- Your rate cards fit the product's billing settings without a spreadsheet beside them
- Every channel your clients sell on is already connected and maintained by the vendor
- The vendor's Amazon app is authorized by your clients and handles renewals
- You have no developer or operations lead who could own custom code
- Workarounds take less than a few hours a week across the team
If most of these are true, buy and revisit the question in a year. If billing or order sync fail the test, keep the product and price the gap as an integration first. Our guide to what ecommerce automation costs covers how to get that quoted, and 3PL invoice reconciliation shows the billing gap from your clients' side.
What should you check before you sign with a prep software vendor?
Whatever you buy will be the core you integrate around later, so test the parts that decide whether a hybrid is possible. Ask for each answer to be shown, not described.
- An open API or scheduled exports for orders, inventory, scans and invoices, so custom pieces can be added later
- Fulfillment Inbound v2024-03-20 for shipment creation, with seller as prep and label owner
- Billing traced to events: an invoice line you can follow back to the receive, label or storage record behind it
- Your clients' channels: each one connected today, with tracking written back
- Amazon authorization: whose app your clients authorize, how renewals are handled and what data is kept
- Data export on exit: your clients, inventory history and invoices in a format you can load elsewhere
- Price at year-three size: the fee at your expected number of clients, orders and users
Who builds the custom parts for prep centers?
Ecomsellertool is a tech agency that builds the pieces prep centers and 3PLs cannot buy: billing rules, channel connectors, reconciliation reports and full prep systems on the Selling Partner API. Send one workflow through our project brief and a person replies within 2 business days with a written plan: what to build, what to keep buying, the timeline and a fixed price.
The clearest buy-then-build story we have is EzBookPrep. Its owner, an experienced book seller, started offering prep services to other book sellers on Quickbase and outgrew it. We built a system that kept the Quickbase functions he relied on and added what a book prep business needs: cataloging by UID, title, author and ISBN, condition grading and labeling, Amazon's inbound and outbound shipment flow through its APIs, billing, and notifications for the next action on each book, which replaced the follow-ups that had made him the bottleneck.
For Little Owl, a 3PL and FBA prep center, we built software with two sides: an admin view for the team and a client view where sellers add shipments, returns and listings and generate FNSKUs themselves. We have built on Amazon's seller APIs since 2017. If your gap is one integration rather than a system, our workflow automation service prices it on its own, and it can run on Ecomsellertool Growth OS so a single connector does not need its own servers. You keep your accounts, your data and the custom code we build; the Growth OS base is licensed to you.
Frequently asked questions
Should a small prep center buy software or build its own?
Buy first in most cases. A subscription built for prep centers already handles receiving, FNSKU labels, storage and a client portal, and it costs far less up front than a build. Build, or pay for an integration, only where a rule or a channel the product cannot hold costs you more each month in manual work than the custom piece would cost over three years.
What is a hybrid approach to 3PL software?
You buy a warehouse or prep product for the floor work and build small custom pieces around it through its API: a billing export that applies your contract rules, an order feed from a channel it does not connect, or a reconciliation report that compares your counts with Amazon's. The product vendor maintains the core, and you maintain only the connectors.
Can a prep center build its own Amazon integration for all its clients?
Yes, but it is a real software project. Serving many sellers means registering as a public SP-API developer, answering Amazon's security questions, and handling authorization for each client. An app not listed in the Selling Partner Appstore is limited to 25 seller authorizations, and every seller must reauthorize it every 365 days. Off-the-shelf prep software has already done this work.
Which Amazon API do prep centers use to create FBA shipments?
The Fulfillment Inbound API, version 2024-03-20. It splits a send-in into an inbound plan, packing options, placement options, transportation options and labels. Two older operations, getLabels and getBillOfLading, are still used to print box labels, pallet labels and bills of lading. Ask any vendor whether its shipment creation runs on this version.
How do I compare the cost of prep software with a custom build?
Total three years for each option, not the first month. For a subscription, add setup, fees that grow with clients or orders, and the hours your team still spends on workarounds. For a build, add the build price, hosting, maintenance hours and your team's time on testing. The hours spent on workarounds are usually the number that decides it.
When is buying prep software the better choice?
When your work is standard Amazon prep, your billing fits the product's settings, your clients sell on channels it already connects, and you have nobody to own a custom system. In that case a build mostly buys you maintenance work. Revisit the decision when workarounds start taking several hours a week.
How we research, fact-check and compare: our editorial standards. Spot an error? Email hello@ecomsellertool.com and we will correct it.
