Per-User vs Flat-Rate ERP Pricing: What Actually Costs Less for Small Manufacturers

Most ERP software charges per user. For a 15-person insurance agency, that's fine. For a 15-person machine shop, it's a trap.

Here's why. An insurance agency has 15 heavy users. Everyone's in the system every day, running reports, writing policies, updating contacts. A $100/user/month bill for 15 people is a reasonable price for the software they actually use.

A 15-person machine shop has three or four heavy users (the owner, the buyer, the estimator, the bookkeeper) and eleven people who might touch the system for 90 seconds a day to clock into a job, mark a step complete, or print a packing slip. Paying $100/month each for those eleven people means the shop is paying roughly $13,200/year for operators to tap a barcode. Most shops don't pay it. They share a login at a kiosk. The supervisor retypes data after the shift. The floor stays on paper. The system the shop paid $300K to implement gets used by four people.

This post walks through why per-user pricing is a bad fit for manufacturing, what the alternatives actually look like, and where the math lands for shops of different sizes. If you're comparing a per-seat SaaS quote against a flat-rate alternative, this is the analysis you'd run if you had three weeks.

Why per-user pricing breaks for manufacturing

Per-user pricing assumes every seat represents roughly the same value. In a SaaS company or a law firm, that's true. In manufacturing, it isn't.

A typical 50-person shop has something like 10 heavy office users (finance, buyers, planners, customer service) and 40 occasional ones. The occasional users are machinists clocking jobs, inspectors writing NCRs, shipping clerks generating packing slips, maintenance techs logging work orders, temps during peak season, and outside processors who just need view-only PO status. Under NetSuite, Epicor, or SAP pricing, the 40 are charged like the 10. A machinist who scans a traveler once a shift is on the bill for roughly $1,500 to $2,400 a year.

The predictable consequence shows up in every manufacturing Reddit thread and Capterra review. Shops under-deploy the system they just paid millions to install. The operations manager gets a license. The shift supervisor gets one. The operators share a login at a shared kiosk, or the supervisor retypes their data after the fact. We wrote more about this pattern in How ERP Overcomplexity Tanks Profitability, but the short version is that the entire ROI case for ERP falls apart when the shop floor isn't on the system. The paperless shop, the real-time WIP, the instant job cost all require the machinists to actually be logged in.

That workaround has real problems. Shared credentials are a security risk. Traceability for AS9100, ISO 13485, ITAR, and CMMC demands unique operator logins, so regulated shops can't legitimately share accounts. And the admin cost of provisioning and de-provisioning users adds up fast when manufacturing turnover runs ~40% a year.

A few other structural patterns make the mismatch worse. Shared kiosk stations on the shop floor need a named license for every person who might touch them under strict SaaS enforcement, even though only one person is at the terminal at a time. Seasonal labor forces shops to either buy peak-headcount licenses year-round or fight SaaS no-refund clauses. Subcontractors who need view-only PO visibility land on vendor-portal licenses that are either unavailable or surprisingly expensive.

Audit risk piles on. The widely-cited Oracle JD Edwards case involved a shop with 280 licenses that Oracle audited, declared out of compliance, and forced to convert at roughly five times the original license count. SAP's 2017 UK High Court win against Diageo established that 5,800 Salesforce users indirectly accessing SAP data triggered $73.6 million in named-user fees. The doctrine from that case (that even machine-to-machine access needs a named-user entitlement) filters down through S/4HANA and creates real exposure for mid-sized manufacturers on SAP.

The four pricing models you'll actually see

Once you cut through the marketing, manufacturing ERP sorts into four pricing buckets.

Base fee plus named users is the dominant model by revenue. NetSuite runs roughly $999 to $4,999/month for the base plus $99 to $199/user/month for full users, with $15 to $30/month Employee Center seats sold in five-packs. Epicor Kinetic runs $1,500 to $2,500/month base plus $125 to $200/user/month with a 10-user minimum. SAP Business One is $94 to $108/user/month Professional, $56 to $112 Limited. Microsoft Dynamics 365 Business Central is $80 Essentials, $110 Premium, and $8 Team Member after the October 2025 price increase. Infor CloudSuite Industrial runs about $150/user/month. Rootstock runs $150 to $300/user/month on top of a mandatory Salesforce platform license that can double the effective per-seat cost.

Tiered per-user is nominally per-seat but priced aggressively for shop-floor roles, so it behaves more like flat-rate. Cetec ERP publishes $50/user/month for base users and $25/user/month for shop-floor "Production Only" users. ProShop ERP uses three role tiers (Shop, Office, Executive) starting around $500 to $715/month covering 7 to 8 employees. MRPeasy is straight per-user at $49 to $149/user/month. Odoo is per-user per-app at about $31/user/month Standard.

Consumption or unlimited-user has one mid-market pure play: Acumatica. Since its April 2024 license refresh, Acumatica prices on transaction volume tiers plus module selection. Users are effectively free once you're past the Essentials edition (which caps at 5 to 10 named users at roughly $7K to $13K/year). Typical mid-market deployments run $1,800 to $5,000/month regardless of login count.

Flat-rate with caps is the smallest category but the most relevant for small manufacturers. DBA Manufacturing sells on-premise at $105/user/year plus a $660/year flat support subscription and a one-time ~$2,995 startup fee. Plex (Rockwell-owned) sells per enterprise or site. PAX sells tiered flat-rate: $350/month for up to 5 users, $900/month for up to 20 users, $1,500/month for up to 50 users.

Where the math actually lands

The honest way to evaluate these is to build three-year license cost at representative user counts for a realistic manufacturer mix (roughly 60% shop-floor, 40% office). All numbers below are list-partner figures. Actual deals run 10 to 30% below list on multi-year commits.

Three crossover points fall out of this math.

At 5 users, the per-user hypothesis is confirmed. DBA at $6.5K and a Dynamics 365 BC mix at $5.9K crush ProShop's $18K floor by roughly three times and Acumatica Essentials at $21K by four times. If you're a genuine 3-to-5-person micro-shop, cheap per-user or a stripped-down BC deployment is the cheapest real option. The flat-rate minimums that make those products attractive at 40 users make them uncompetitive at 5.

Between 10 and 25 users, tiered per-user wins with cheap light-user tiers. Cetec's $25 shop-floor rate and BC's $8 Team Member tier turn the table. At 15 users, a BC heavy-Team-Member mix lands at $17.7K over three years versus $45K for Acumatica.

Between 25 and 75 users, user mix drives everything. A 50-person shop that is 70% shop-floor pays roughly $232K over three years on NetSuite full-user pricing, $194K on SAP B1, $120K on Acumatica Prime, and $63K on Cetec. A 50-person shop that is 70% office flips the ranking: BC Essentials comes in around $59K, Acumatica around $120K. This is the gray zone where marketing claims on both sides are simultaneously true for different buyers.

Above 75 to 100 users, unlimited and consumption pull ahead. Acumatica at 100 users ($180K to $225K) beats NetSuite ($265K), SAP B1 ($277K), and ProShop scaling ($216K). This is where the per-user growth tax compounds until it becomes the single biggest line on the five-year TCO. It's also not the small-manufacturer band, so we'll stop there.

Implementation cost usually matters more than licensing

One big caveat. Licensing is rarely the biggest number on a five-year ERP TCO. Panorama Consulting's benchmark data puts license fees at 20 to 30% of total ERP cost, with consulting, integration, and change management making up the rest. Implementations run 1x to 2x annual license for straightforward mid-market deployments, 2x to 4x for Epicor, IFS, and Sage X3, and 3x to 5x for SAP S/4HANA and Oracle Cloud. In absolute terms, a typical mid-market cloud ERP runs $150K to $750K to implement, and more than half of companies exceed that budget.

That reshapes the licensing comparison in one important way. A 25-user shop choosing Acumatica over Cetec might save nothing on total cost after a $40K to $100K partner-led implementation versus Cetec's $10K to $25K self-guided path. Any vendor that charges $0 implementation (DBA, Statii, PAX) starts with a $30K to $100K head start before the per-user math even begins.

If you're a small manufacturer picking ERP, that's the first question to ask: what does it actually cost to get live, not what does a user seat cost. We dig into this more in How Much Does Small Manufacturing ERP Cost?

The honest case against flat-rate

Flat-rate has its own failure modes.

"Unlimited" is rarely literally unlimited. Acumatica's transaction tiers can escalate unexpectedly for high-throughput distributors. ProShop's tiers actually scale with employee count. Cetec is transparent per-user, not flat. Plex's "unlimited" is priced per enterprise and is correspondingly expensive. Small shops pushed into a flat-rate vendor's minimum tier by its floor are subsidizing the price point's breakeven. If you're 3 people, you don't belong on a product whose pricing was designed to make sense at 40.

Flat-rate also doesn't fix a bad software fit. A cheap ERP that doesn't do what your shop needs is more expensive than an expensive ERP that does. Pricing model is one variable among many.

How to think about this for your shop

The useful framing isn't "per-user versus flat-rate" but three zones.

Under 15 users, cheap per-user SaaS (DBA, MRPeasy, Odoo, or a BC Essentials-plus-Team-Member mix) is almost always cheapest, and flat-rate vendors with minimums above $5K/year price themselves out.

Between 15 and 75 users, the winner depends on mix. Shop-floor-heavy operations want Cetec, ProShop, Acumatica, or a flat-rate alternative like PAX. Office-heavy operations often still win on BC or SAP B1 with a smart Limited-user mix.

Above 75 to 100 users, the per-user growth tax compounds enough that unlimited and consumption models dominate. That's also the scale where enterprise implementation cost (3x to 5x license) becomes the bigger discussion.

A few specific things to negotiate into any ERP contract, regardless of model. Cap annual escalators at 3 to 5% and ideally pin them to CPI. Negotiate true-forward rather than retroactive true-up clauses. Insist on ramp-down rights, because most standard SAP and Oracle ELAs explicitly exclude them. For any SAP customer, add explicit indirect-access language post-Diageo. None of this shows up on pricing pages but all of it can swing five-year cost by 20 to 40%.

And before you sign anything, count your actual users by role. Not headcount. How many people need to be in the system every day, how many need to be in it occasionally, and how many never need to log in at all. That list is the input to every pricing comparison you're going to run.

Where PAX fits

PAX sells tiered flat-rate: $350/month for up to 5 users, $900/month for up to 20 users, $1,500/month for up to 50 users. All features unlocked in every tier. No module-gating, no user-type tiering, no implementation fee, free data migration.

That's a good fit if you're a 5 to 50 person manufacturer with simple to moderate BOMs who wants shop-floor users on the system without paying $100/user/month for machinists. It's also a good fit if you want to go live in days rather than spending three months on an implementation.

It's not the cheapest absolute option if you're a 3-person shop. DBA Manufacturing at $105/user/year plus $660 support will beat us on year-two and onward cost if you're willing to run on-premise software. And it's not the right tool if you're a 75+ user operation with complex multi-plant scheduling, regulated aerospace finite capacity, or heavy engineer-to-order configuration. At that scale, you want Plex, Acumatica, or a tier-one ERP that was built for that complexity.

If you're somewhere in between, or if you're still trying to figure out whether you even need ERP yet, When Does a Small Manufacturer Actually Need ERP? is a better starting point than this post.

And if you're staring at a $69K Epicor quote and trying to figure out whether flat-rate would actually save you money, reach out. We're happy to run the numbers with you. If PAX isn't the right fit, we'll let you know and point you in the right direction.

Written by

Matthew Obey
April 23, 2026