Last updated: August 6, 2026
By Tom Kenaley, Co-Founder and President, KORE1
Build in-house when your NetSuite instance changes most weeks and somebody has to answer on Tuesday morning. Bring in a consultant when the work is project-shaped, deep, and finite. Nearly every company treats this as a budget question. It is a cadence question. The budget falls out of the answer, and running that sequence backwards is how a controller ends up carrying a $9,500 monthly retainer nobody uses and a change backlog nobody owns.
I have watched it break in both directions.
Where I sit should shape how you read this. KORE1 has placed ERP talent since 2005 and we bill on both halves of this question, the full-time search and the contract consultant, so no outcome here is neutral for us. One section below tells you to promote somebody already on your payroll and send them to training. That is a search we do not run and cannot invoice, which is the only reason I fully trust it as advice, because everything else on this page has our thumb resting somewhere on the scale. It stays because I have watched the alternative fail on a predictable schedule, usually around month seven.
If you are sizing the outside market, our NetSuite consultant staffing desk covers that half, and NetSuite administrator staffing covers the inside half. This piece is about which one you are actually shopping for, because most companies asking the question have not decided yet and do not know it.
Three Things Get Called Consultant and Only One Competes With a Hire
A NetSuite consultant is an outside specialist who configures, extends, or repairs an Oracle NetSuite instance under a contract instead of a payroll line. That one label covers three engagement shapes with completely different economics, and comparing any of them to a full-time hire without naming which one produces an answer you cannot act on.
Shape one is the project consultant. Implementation, a module rollout, a OneWorld subsidiary going live, an ASC 606 revenue recognition build in Advanced Revenue Management. Fixed scope, real end date, and the work honestly requires somebody who has done it eleven times. No employee competes with that. You are not choosing between two options. You are buying something your team has never built and will not build again for four years.
Shape two is managed services. A monthly retainer, a block of hours, a named person who answers tickets and occasionally does not. This shape competes with a hire head-on, it is where most of the wasted money in the category lives, and it is almost always what somebody actually means when they tell me they are weighing a consultant against an employee.
Shape three is staff augmentation. A contractor in your standups doing the job an employee would do, billed hourly through a contract staffing arrangement. Same work. Different paper. Very different exit.
When a controller asks me consultant or in-house, they mean shape two roughly nine times in ten, and they describe shape one to justify asking. Sort that out first. Everything below assumes you have.
What Each One Costs Across Three Years
Here is where the conversation usually goes sideways, because people compare a salary to an hourly rate and call it analysis. A salary is not a rate. A retainer is not a salary. The only comparison that survives contact with a CFO is total cost across the period the instance actually has to keep running, which is never one year.
Assume a mid-market company, single subsidiary, roughly 60 NetSuite users, a Celigo pipeline into Shopify and an Avalara tax integration. Common shape. We staff it constantly.
| Model | Year 1 | 3-Year Total | Hours You Get Per Year | Effective Cost Per Hour |
|---|---|---|---|---|
| In-house admin, $111,000 base | $144,000 loaded, plus $22,000 to $28,000 to find them | $460,000 to $470,000 | ~1,750 productive | $83 to $90 |
| Managed services retainer, 20 hours monthly | $96,000 | $288,000 | 240 contracted | $400 |
| Senior functional contractor at $145/hr, half time | $150,800 | $452,400 | 1,040 billed | $145 |
| Project consultant, one 14-week build | $180,000 to $220,000 | $220,000 to $290,000 | ~560, then zero | $320 to $395 |
Loaded cost assumes base plus 30 percent for payroll taxes and benefits. Search cost assumes a contingency fee at 20 to 25 percent of base. Contract rates reflect what we are placing senior functional consultants at in mid-2026, detailed further in our NetSuite consultant salary guide.
Read the last column, then read the second one again.
The retainer has the smallest invoice and the most expensive hour on the board by a factor of nearly five. Finance approves it anyway, every time, because $8,000 a month clears a signature threshold that $144,000 a year does not, and because a retainer has no severance conversation attached to it. That is a real advantage and I am not pretending otherwise. It is just not a cost advantage, and the deck someone sold you almost certainly framed it as one.
The number that never makes the comparison is context. An employee in month eighteen knows that the Anaheim warehouse posts receipts a day late, that the sales team abandoned three of the custom fields nobody deleted, and that the Celigo error email on the second of the month is always the same duplicate SKU and can be ignored. A retainer consultant re-learns some fraction of all that on every single ticket, sometimes the same fraction twice in a quarter because a different person picked it up, and you are paying $400 an hour for the re-learning. Over three years that gap is not a rounding error. It is most of the value.
The Threshold Everybody Uses Is Measuring the Wrong Thing
Search this question and you will get the same two triggers from a dozen sources. Fifty users. Twenty support tickets a week. Cross either line and hire somebody.
Both are support-volume metrics, and support volume is the least interesting thing about a NetSuite instance. A company can generate forty tickets a week and need no admin at all, because thirty-eight of them are password resets and saved searches that a trained power user in accounting closes before lunch. Another company files four tickets a month and desperately needs a full-time person, because all four are “we just acquired a company and need it in NetSuite by close.”
Count changes instead. Not tickets. Changes.
Go through last quarter and count every time somebody asked NetSuite to do something it was not doing before. A new approval routing. A field added to the sales order form. A workflow in SuiteFlow. A new saved search that feeds a board deck. A subsidiary, a currency, a tax jurisdiction, a warehouse, a revenue schedule.
Under six a quarter, a retainer is genuinely the right instrument and hiring is premature, because an admin with nothing to change spends most of the year producing reports nobody asked for and maintaining a documentation wiki nobody opens. Six to twenty is the contested zone where the answer depends on the next section. Above twenty a quarter, you needed the hire two quarters ago and the retainer is currently functioning as an expensive queue.
There is a second test and it takes about four seconds. Something breaks at 8:40 on Tuesday morning during close. Does the fix require a person who is already in your Slack, or can it wait for a ticket queue with a four-hour SLA? Answer honestly rather than aspirationally. Plenty of businesses can wait. If yours genuinely can, that is worth about $150,000 to you and you should stop reading and go renew the retainer.
Four Questions That Actually Decide It
- How often does the business change shape? That is the cadence count above. It outranks the other three and it is the one nobody runs.
- How much of your instance is SuiteScript rather than configuration? Pull up the custom scripts list. If there are more than about fifteen deployed scripts and none of your people can read JavaScript, you have a technical dependency, not an administrative one, and the fix is a different person entirely. Our breakdown of the functional versus technical consultant split covers which one you are looking at.
- Structural complexity. OneWorld across three subsidiaries, four currencies, intercompany eliminations, ASC 606 revenue schedules, and an external audit every spring, and each one of those pushes toward a depth of experience you cannot grow internally inside a single year no matter how sharp the person is.
- The last one is uncomfortable, so it goes last. When the auditors ask who controls administrator access to the system of record for revenue, is the honest answer a vendor?
Three or four pointing the same direction and the decision is already made. A genuine split, two and two, means you are in the blend and should skip ahead.
The First In-House Hire Is Usually Not a Consultant
This is the part that costs us money, so here it is with no hedging.
When a mid-market company decides to bring NetSuite in-house, the instinct is to hire a NetSuite consultant into the seat. Implementation background, four go-lives, certifications on the resume. It reads like buying the best available. It fails often enough that I now talk clients out of it before they post the req.
An implementation consultant is optimized for a job that ends. New instances, new clients, a fresh problem every quarter, and a professional identity built on the interesting part. Then you sit them in a stable single-subsidiary instance whose main demand, four days out of five, is knowing why the Costa Mesa distribution team keeps overriding the item receipt date and which supervisor to talk to about it. The skills that made them expensive atrophy. Around month seven they start answering recruiter messages, and I know because ours are among the messages they answer.
The hire that works more often is somebody you already employ. The senior accountant or ops analyst who has quietly become the person everyone asks. Send them to NetSuite training, back it with a real retainer for the first year, and give them the title and the band. You are paying $85,000 to $100,000 plus maybe $9,000 in training against $130,000 for the external hire, and you are getting somebody who already knows the business, already has the relationships, and does not need six months to learn that the Tustin site does inventory differently for reasons that predate everyone.
They do have a ceiling. A promoted power user will not architect a multi-subsidiary consolidation and should not be asked to. Pair them with outside depth for that work, which is the blend, which is the section after next.
One warning on comp. NetSuite admins nearly always report into finance, and finance benchmarks the band against the senior accountant sitting next to them rather than against the platform market. That is how a $111,000 job gets posted at $86,000 and sits open for five months. We wrote the full pattern up in the NetSuite administrator salary guide. Public data does not help much either. Salary.com put the national average at $76,758 as of August 1, 2026, while ZipRecruiter had it at $110,278 the same summer. Same role, same country, a $33,000 gap. Pick the low one and your req is decoration.
Where the Handover Actually Breaks
Whichever direction you go, you will eventually move work between an outside party and an inside one. That transfer is where the real damage happens, and almost nobody writes it into the contract.
A Newport Beach client came to us after their implementation partner’s engagement ended. Forty-one custom SuiteScript deployments across order management and billing. No inline comments. No script inventory. Two saved searches feeding the CEO’s weekly dashboard that pulled from a custom record nobody could explain. The partner had done competent work, had been paid in full, and behaved honorably from start to finish, and not one bit of that mattered on the Monday, because everything that made the work usable had walked out in a single person’s head on the Friday.
They spent eleven weeks and about $54,000 having a second consultant reverse-engineer the first consultant’s work. Then they hired.
Put these in the statement of work before anybody starts. Not after.
- A written inventory of every deployed script, workflow, custom record, and integration, refreshed at the end of each engagement phase
- Inline comments and a plain-English purpose line on every SuiteScript deployment, treated as a deliverable rather than a courtesy
- Administrator credentials owned by your company, never by the vendor. This one sounds obvious. We see it violated maybe a third of the time.
- Two recorded walkthrough sessions in the final two weeks, with your people asking the questions
- A named internal person who attends design decisions from week one, even if they understand only half of what is said in the first month
That last item is the whole thing, honestly. The other four are artifacts and artifacts go stale. A person in the room from day one is the only knowledge transfer that has ever reliably worked, and it costs you a few hours a week of somebody’s attention.
What the Blend Looks Like on a Calendar
Most mid-market companies land here, and “hybrid” gets recommended so often that it has stopped meaning anything. So here is the actual split we see working, by work type rather than by philosophy.
| Work | Owner | Why |
|---|---|---|
| Roles, permissions, user provisioning | In-house | Audit exposure, and it needs same-day turnaround |
| Saved searches, dashboards, SuiteAnalytics reporting | In-house | Requires business context more than platform depth |
| Form and field changes, approval routing, SuiteFlow | In-house | Highest-frequency request type by a wide margin |
| SuiteScript development and code review | Outside, retained | Rarely enough volume to justify a full-time developer |
| New subsidiary, currency, or tax jurisdiction | Outside, project | You do this twice a decade and it is unforgiving |
| Integration builds via SuiteTalk, Celigo, or Boomi | Outside to build, in-house to run | Building is a project. Monitoring is a Tuesday. |
| Version upgrade regression testing | Both, twice a year | Your people know what to test. Theirs know what breaks. |
Notice the shape. The employee owns everything that is frequent and context-dependent. The outside firm owns everything that is rare and unforgiving. Nothing in that table splits the work by seniority, which is exactly how these arrangements get drawn on a whiteboard most of the time, junior work inside and senior work outside, and it is why most of them quietly stop working by the second quarter.
Budget for the blend runs one salary plus $2,500 to $5,000 a month in retained hours for a single-subsidiary instance, then project money on top when something structural happens. That is more than either pure model on paper. It is less than either one after the second unplanned event, and there is always a second unplanned event.
Worth knowing before you plan around a partner as a permanent fixture. Oracle moved to cut somewhere between 20,000 and 30,000 roles globally in 2026, close to 18 percent of its workforce, against a $2.1 billion restructuring charge disclosed in an SEC filing, according to reporting in CIO, and NetSuite’s own development organization took reductions. Partner bench depth is thinner in some practices than it was eighteen months ago. Ask who specifically is assigned to you and what happens when that person rolls off, because “the firm” is not an answer.
What Controllers Ask Before They Pick
We already have a consultant. When does the first hire make sense?
When your change count crosses roughly six a quarter, or the day somebody on your team starts collecting and rewriting everyone else’s requests before they reach the consultant. That second signal usually arrives before the first one does.
The informal coordinator is the tell. Somebody in finance has quietly become the person who decides what is worth a ticket, writes the requirements, and tests the result. That person is already doing something like forty percent of the job for free, on top of the full-time role they were actually hired to do, and they will keep doing it right up until the quarter they burn out or take a call from somebody like us. You are not deciding whether to create the position at that point. You are deciding whether to pay for it.
Our whole instance lives in one contractor’s head. How bad is that?
Bad, and worse than the version you are imagining, because the risk is not that they leave. It is that you cannot negotiate with them, cannot audit them, and cannot get a second opinion on their work.
Rate conversations only ever go one direction when the alternative sitting on the table is eleven weeks and $54,000 of reverse engineering, and both sides understand that perfectly well without anybody needing to say it out loud. Start the fix this month, and start it small. A script inventory and two recorded walkthroughs cost about a week of their time and shift the balance of that relationship further than anything else you could buy for the money. Most good consultants will do it without complaint. The reaction to the request tells you which kind you have.
Is a NetSuite certification worth paying extra for?
Somewhat, and mostly as a floor rather than a signal. A NetSuite Certified Administrator credential proves familiarity with the platform’s vocabulary and standard behavior, which eliminates a specific category of bad hire, and proves nothing about judgment.
The interview question that actually separates people is not about certification content. Ask them to describe a customization they argued against building. The strong ones have three, with reasons, and can tell you what the business wanted instead. Resume-padders describe things they built. There is more on how we weigh credentials in our guide to hiring a NetSuite implementation team.
Two admins, or one admin plus a retainer?
One admin and a retainer, in almost every case under about 250 users. Two admins is the right answer far less often than companies think, and the second seat is usually solving a specialization gap rather than a capacity gap.
The exception is coverage. One admin means one person’s vacation, one person’s illness, and one person’s resignation are all single points of failure on the system that closes your books. A retainer with a named backup consultant who has actually logged into your instance solves that for a fraction of a second salary. Set it up before you need it, because the month you need it is the month nobody has bandwidth to onboard anyone.
How long before an in-house hire pays for itself?
Fourteen to twenty months in the cases we can measure, against a comparable retainer, assuming the change cadence justified the hire in the first place. If cadence did not justify it, the answer is never, and that is not a hypothetical.
The payback math is straightforward on the hours and misleading in both directions on everything that actually determines the outcome, which is why I distrust anyone who hands you a clean number for this without asking a single question about your instance. Faster answers and accumulated context bring it in sooner. A bad hire, or a good hire banded $25,000 below market who leaves at month fourteen, pushes it past never and takes the search cost with it. BLS projects 15 percent growth for computer and information systems managers through 2034 with about 55,600 openings a year, against a median of $171,200 as of May 2024. Retention is the variable that decides this, not the offer.
Run the Cadence Count This Week
Open last quarter’s request history and count the changes. Not the tickets, the changes. It takes about forty minutes and it answers the question more reliably than any framework, including the one above.
Then be honest about the Tuesday morning test, decide who owns administrator access when the auditors ask, and get the knowledge-transfer language into the statement of work regardless of which way you go. Those three cover most of what goes wrong.
One more thing worth saying, since public data will not tell you much here. The Bureau of Labor Statistics does not track NetSuite consultants as an occupation. The nearest category, management analysts, showed a median of $101,190 in May 2024, which averages platform specialists in with general business consultants and reads low against every aggregator. Treat it as a floor. Never as a band. And on the project side, Panorama Consulting Group’s 2026 ERP Report found more than a quarter of organizations exceeded their project budgets, with additional technology needs leading the causes. Budget the overrun into the decision rather than discovering it.
If you want a second read on which shape your situation calls for, or a sanity check on a band before the req goes live, send our ERP recruiting desk the details. We have been placing this talent since 2005, our average time-to-hire across IT roles is 17 days, and 92 percent of the people we place are still in the seat at twelve months. If the honest answer is that you should promote your accounting analyst and buy them a training package, we will say so.
