Why NetSuite Is Eating at Your Margins
Your NetSuite renewal went up again. It is under 1% of revenue but about 8% of profit, and the contract lets that grow. Where the increase comes from.

The NetSuite renewal quote is on your desk, higher than last year's for the same users and the same modules. Against revenue the cost looks small, about 0.8% a year in our estimate for a 50-person and a 150-person manufacturer. Against profit it is about 8% a year, and the contract lets that share grow at every renewal while your power to refuse it shrinks.
Most of it comes down to 2 sentences. The first is in Oracle's standard US subscription agreement for NetSuite: "Pricing for subsequent renewal Estimate/Order Forms shall be set at then current Oracle pricing, unless otherwise agreed to by the parties." The second is in NetSuite's own article on what ERP costs: "Businesses often spend as much, if not more, on those ongoing costs than on the initial licensing and implementation."
Oracle publishes no NetSuite price list, so every price below is a third-party estimate with its publisher named. For other systems, see NetSuite alternatives for manufacturers. This post is about the contract you already have.
Renewal is priced at whatever Oracle charges then
The subscription "shall be automatically renewed for one (1) year unless and until either party provides written notice of non-renewal" at least 30 days before the term ends (Section 4.1). Miss that date and you have bought another year at then current Oracle pricing, unless you negotiated otherwise.
Nobody publishes how large those increases run. A buyer-side licensing advisor that benchmarked 30 to 40 renewals from 2024 to 2026 puts the automatic uplift "near 10 percent" where no cap was signed. Vendr's purchase data says 5% to 10% a year, and a second buyer-side advisor says 8% to 15%. Advisors sell negotiation help, so read these as ranges, not forecasts.
Customers have posted larger numbers in the NetSuite Professionals archive, a public archive of a large NetSuite admin community. One cut modules and licenses and still wrote, "Even then the estimate is 13% higher this year." (April 2024). A customer with 220 full users, larger than most shops reading this, reported 42% (March 2024). These posts are self-selected: people write in when a renewal goes badly, so they skew toward bad outcomes.
Whatever the rate, it compounds. At 10% a year the subscription is 46% higher in year 5 than in year 1, with no change in what you use. In the National Association of Manufacturers' second-quarter 2026 survey, manufacturers expected input costs to rise 5.8% over the next year and their own selling prices 4.2%. A software line compounding near 10% outruns both.
First-year discounts come back off
Most contracts start below list. Vendr's purchase data puts first-year discounts at 15% to 40%, depending on deal size. The discount is the part of the price most likely to change. "Also most discounts are being stripped away in all new renewals," a customer posted in March 2024. In December 2023, another removed about $26,000 of licenses and still saw the renewal rise from $187,683 to $206,686, about 28% more for what remained. Their NetSuite partner explained that Oracle has been "reducing discounts year after to eventually remove them from the subscription and have the accounts at list price."
The first advisor sums up the trade: "A 30 percent discount with an uncapped 10 percent uplift is a three year loan of Oracle's margin." In its own 5-year example, the offer with the bigger discount and no cap costs more from year 4 onward. The discount ends. The escalator does not.
Every seat, module and extra is its own line
NetSuite is sold as a base fee, named user seats, and modules that, in NetSuite's own guide, "can be licensed separately at any time during your NetSuite contract." A manufacturer buys a base, seats and 2 to 5 modules. The full-user seat, the largest line in most contracts, lists at $129 a month in 2026, up from $99 in 2024, by a NetSuite partner's benchmark. Customers cite $120 or $129 (March 2024), a 21% to 30% rise. Module estimates vary by a factor of 2 to 3: $600 to $2,000 a month for core manufacturing from that partner, and $24,000 to $40,000 or more a year for Advanced Manufacturing from the second advisor, who warns: "Treat these figures as negotiating reference points, not quotes."
Seats drift up more easily than down. Use more than you ordered and you "promptly must purchase and pay fees for the excess quantity" (Section 2), while signed orders cannot be cancelled. The first advisor found full-user seats overbought by 15% to 30% against login data.
Then the extras. An Oracle email a customer posted in April 2024 said "The Premium, Enterprise and Ultimate NetSuite Service Tier licenses will no longer include a sandbox account." A sandbox, the copy of your account where changes get tested, now runs about 10% of the subscription by a NetSuite partner's estimate, and rises with it. Extra API capacity costs $10,000 to $25,000 a year by the second advisor's estimate. EDI, shipping and scheduling tools come on separate contracts, each with its own renewal date. AI Units are included for most customers today at no extra cost, though RSM, a NetSuite partner, notes that "Unused units expire at the end of the term rather than carrying forward." That is a meter in the contract, not yet a charge.
The people cost starts at go-live
Implementation is billed separately, usually by a NetSuite partner, and runs $35,000 to $250,000 for a manufacturer (ERP Research; BrokenRubik). NetSuite's pricing article says the bigger cost comes later: cloud systems "cost significantly less initially but have the potential to cost more in the long run."
Most of the long run is people: report requests, integrations, and retesting through 2 major releases a year. A NetSuite administrator's US median base salary is about $111,000 in 2026, by a recruiter's salary guide, before benefits. Shops without one pay retainers of $2,500 to $5,000 a month for small companies and $5,000 to $25,000 or more for mid-sized ones, according to Folio3, a NetSuite partner. In our estimate this is the second-largest line after the subscription, and it does not end.
The exit terms set how hard you can push back
Every renewal comes down to what it would cost to say no, and the agreement sets most of that. The customer owns its data: "all title and intellectual property rights in and to the Customer Data is owned exclusively by Customer." But "each Estimate/Order Form and SOW is non-cancellable and all sums paid are non-refundable," termination is for uncured breach only, and afterward "Oracle will deactivate the applicable Customer account(s) and delete any data therein." The US text states no window for getting data out.
So leaving happens inside the term. You export through saved searches, CSV, SuiteQL or APIs, rebuild reports, workflows and scripts, and usually carry over only 1 to 2 years of transaction detail, according to 2 NetSuite partners (Jade Global; BrokenRubik). Every script your team adds makes leaving cost more. When a customer posted a 40% renewal quote in August 2024, a reply the same day was: "They have all the power since it's the heart of your company most likely." That is one customer's view, not a measurement, but the direction holds: your bargaining power peaks at the first signature and falls every year after.
What it adds up to over 5 years
Oracle publishes no prices, so this is a KoldOps estimate built from the third-party prices above, not a quote. The assumptions:
- 15 full users at 50 employees and 40 at 150. The ratio is ours; no source supplies one.
- Revenue from Census receipts per employee: $321,302 at 50 to 74 employees and $369,065 at 150 to 199, in 2022 dollars.
- Estimated list prices, no first-year discount, people and add-on costs held flat, and administrator benefits and staff time left out.
- 3 renewal paths: a 3% cap; an uncapped 10% uplift; and a flat 3-year term followed by a 40% step like the 2024 posts, then 10%.
- A 9.9% net margin, the average for small Canadian machine shops. It is the only recent small-manufacturer benchmark we found, and a weak one: Canadian, with its tax basis unstated.
| KoldOps estimate, from third-party prices | 50 employees | 150 employees |
|---|---|---|
| Revenue used | $16.1 million | $55.4 million |
| Year-1 subscription | $56,808 (base, 15 users, manufacturing, Advanced Inventory) | $196,812 (base, 40 users, 4 manufacturing modules, analytics access, sandbox) |
| Implementation and EDI setup, one time | $81,250 | $162,500 |
| Administrator or retainer, per year | $36,000 | $171,000 |
| Add-ons, per year | $10,688 | $22,688 |
| 5 years, 3% renewal cap | $616,291 | $2,175,841 |
| 5 years, uncapped 10% uplift | $661,508 | $2,332,496 |
| 5 years, flat term then a 40% step | $652,129 | $2,300,003 |
| Average share of revenue per year | 0.77% to 0.82% | 0.79% to 0.84% |
| Average share of net profit per year | 7.7% to 8.3% | 7.9% to 8.5% |
| Outer bounds from the same price ranges | $350,000 to $1.75 million | $1.06 million to $4.12 million |
The 50-person first year, $138,058 of subscription, implementation and EDI setup, sits inside Vendr's $60,000 to $150,000 range for small deployments, the main check that the model is not inflated. The administrator or retainer line carries much of the total. Strip it out and the 3% cap totals fall to about $436,000 at 50 people and $1.32 million at 150, and the 50-person figure sits level with the high case in our Business Central model. The subscription is still the line that grows on its own.
Uncapped, the 10% path costs $45,000 more over 5 years than a 3% cap at 50 people and $157,000 more at 150. At a 9.9% margin, each recurring dollar takes about $10 of new sales to replace. At the 5.9 to 7.2 times EBITDA that GF Data recorded for lower middle-market manufacturing deals in 2025 and early 2026, as reported by an M&A advisor, each is $5.90 to $7.20 off the sale price. The escalator alone, in year 5:
| Year 5, uncapped 10% against a 3% cap | 50 employees | 150 employees |
|---|---|---|
| Subscription, 3% cap | $63,937 | $221,513 |
| Subscription, uncapped | $83,172 | $288,152 |
| Gap in that year | $19,235 | $66,639 |
| Year-5 total cost as a share of net profit, capped and uncapped | 7.0% and 8.2% | 7.6% and 8.8% |
| New sales needed to replace the gap | About $194,000 | About $673,000 |
| Off the sale price at 5.9 to 7.2 times EBITDA | About $113,000 to $138,000 | About $393,000 to $480,000 |
Read these as orders of magnitude, not appraisals. The prices are estimates, the user counts and margin are assumptions, and the multiple matters only if the cost is still recurring when a buyer looks. The gap in the second table is what a 3% cap is worth.
When NetSuite is the right call
NetSuite is a capable system, and for some shops its cost fits what it does.
- The hard problem is financial. NetSuite is strongest in discrete, multi-entity and distribution-heavy manufacturing (ERP Research). On Capterra, "Integrated financial process automation" draws 90% positive mentions in 209 reviews.
- You signed a cap. A 3% to 5% cap and a multi-year price hold written into the first contract avoid the worst of this post (Vendr; BrokenRubik).
- The worst cases are not typical. The 40% and 42% renewals are self-selected posts. A 150% figure in the March 2024 thread was an opening quote still under negotiation, and first quotes in the first advisor's sample ran 20% to 35% above the final price.
- Your data is yours on paper. The agreement says so. The risk is timing and export, not title.
The fit is weakest in high-mix job shops, where a comparison site says NetSuite "struggles with deep ETO complexity." Sources disagree on its scheduling (Delfoi, a competing scheduler vendor, against Anchor Group, a NetSuite partner). The likely reading is that finite-capacity scheduling needs the paid Advanced Manufacturing module.
Read your own contract
Before the next renewal, pull the agreement, the order form and the last invoice, and check 8 things.
- The renewal clause. Find "then current Oracle pricing" and any amendment that overrides it. No override means no cap: plan for about 10%, plus a step if your first-year discount was large.
- The cap. Check that it covers every user type, every module and the sandbox, for every year of the term, and applies to the price you pay rather than list. A cap on list alone does not stop a withdrawn discount.
- The notice date. 30 days before the term ends, under the standard agreement. Calendar it, and calendar a date 6 months earlier, when a buyer-side advisor says renewal work should start. The first advisor also points to Oracle's May 31 fiscal year end as a window for discounts.
- Users against logins. At $129 list, each full seat nobody uses costs $1,548 a year.
- Modules against use. A module nobody opened last quarter is a line to drop or to trade.
- The discount column. List price minus your price, line by line, is what a renewal at list would add.
- Extras. Sandbox, API capacity, analytics access, storage and every third-party app, each with its renewal date, on one list.
- Your data. Run a full export now through saved searches, CSV and SuiteQL, open it outside NetSuite, list what would need rebuilding, and ask in writing for an export window after termination.
Questions we get
Is NetSuite a big share of a manufacturer's revenue?
No. In our estimate it averages about 0.8% of revenue a year, and 1.0% to 1.15% in year 1. It shows up in profit instead: about 8% of net profit a year at a 9.9% margin.
How much does NetSuite go up at renewal?
No one publishes a distribution. Without a cap, a buyer-side advisor's sample puts the automatic uplift near 10%, and Vendr's purchase data says 5% to 10% a year. Customers have posted quotes 13% to 42% higher, but those posts skew toward bad outcomes. A 3% to 5% cap written in at signing avoids most of it.
Can I still negotiate if my renewal is close?
Yes, with less room than at first signature. In one advisor's sample, first renewal quotes ran 20% to 35% above the price finally signed. Bring login data and know your notice date, because once it passes the contract renews for a year at then current Oracle pricing unless you agreed otherwise.
Do I own my data in NetSuite?
Yes, the standard US agreement says so. The catch is access: after termination Oracle deactivates the account and deletes the data, and the agreement states no window for getting it out. Export and test your data before you give notice.
Will cutting users or modules lower the renewal?
Not reliably. One customer removed about $26,000 of licenses and the renewal still rose from $187,683 to $206,686 as discounts were withdrawn. Cut unused seats anyway, since one advisor found 15% to 30% overbought, but settle the discount and the cap in the same conversation.
Is owning a system always cheaper than renting NetSuite?
No. In our worked example, a $100,000 build with $24,000 a year of upkeep costs $220,000 over 5 years, so it wins only where it removes more than about $44,000 a year of rent. Our 50-person NetSuite estimate runs $103,000 to $134,000 a year after year 1. The upkeep figure is our assumption and the least-sourced number in that comparison; why software is part of your means of production shows the full break-even.
What's next
Find your notice date this week; it is the last day you have a real choice about next year. If the system fits and only the renewal hurts, negotiate a cap, a multi-year price hold and a right-sized seat count. If the fit is the problem too, read NetSuite alternatives for manufacturers and how much a custom ERP costs. The same arithmetic for Microsoft is in why Microsoft Dynamics is eating at your margins.
KoldOps builds the system its partners run on, and they keep the code and the data. A build is live in under 6 months for under $100k. Packaged NetSuite deployments are quoted at about 100 days (Oracle Licensing Experts), so the go-live date is not the difference; the rent after go-live is. For a second read of your renewal, write to Kol with the quote and the notice date. Kol reads every message.