Last updated: July 30, 2026
NetSuite optimization is the ongoing work of closing configuration gaps, retiring manual workarounds, and switching on modules and automation you already license but never deployed at go-live. It is not a second implementation. It is the cheaper half that nobody schedules.
A CFO turned a laptop around to show me a spreadsheet last spring. Inventory by SKU, color-coded, updated every Tuesday by an analyst named Marcus. Next to it, on the same screen, was NetSuite. Showing a different number.
I asked which one they trusted. The answer was Marcus.
Two years live. They were paying for a system that could have produced that number automatically, in real time, with an alert the moment it drifted, and instead they were paying a salaried human to rebuild it by hand every Tuesday. Nobody had done anything wrong, exactly. They just stopped after go-live, which is what almost everybody does, because the implementation had a partner and a budget and a finish line while the eighteen months afterward had none of those things and nobody’s name attached to it.
My bias, out front. I run an ERP and business systems consulting group, and optimization work is a thing my company sells. So read this knowing I profit if you decide your system needs attention. The counterweight is that most of what follows is stuff your own admin can do in an afternoon, and I say so where that’s true. If a section costs you nothing but a Tuesday, I’d rather you just go do it.
Also worth naming, since you’re reading this on a staffing company’s site. KORE1 places NetSuite people. They didn’t pay me to write this and they don’t sell software. Their IT staffing services practice fills a typical search in about 17 days, and 92 percent of those placements are still there a year later, which matters later in this piece when I argue that some of your problem is a headcount problem wearing a software costume.
What NetSuite Optimization Actually Means
NetSuite optimization is a recurring review of how your live instance matches how the business runs today. It covers configuration cleanup, unused module activation, workflow automation, integration repair, saved search and dashboard rebuilds, role and permission tightening, and data hygiene. Implementation gets you live. Optimization gets you value.
The distinction sounds academic until you look at where the money went. Implementation is a project. It has a budget, a partner, a go-live date, and a party. Optimization has none of those, which is why it doesn’t happen. There’s no calendar invite for it.
Gartner’s forecast is that by 2027, better than 70 percent of recently implemented ERP initiatives will miss the business case they were sold on, with as many as a quarter of them failing outright. Sit with that one. The software went live. The business case didn’t. Two different events, and companies keep celebrating the first as though it were the second. Gartner’s analysis of why ERP initiatives disappoint lands most of the blame on planning, training, and change management rather than on the platform itself.
Which tracks. The myth I spend the most time arguing with is that NetSuite is complicated. It isn’t, particularly. Your org is. The chart of accounts nobody will renegotiate, the two people who genuinely like their spreadsheet and will defend it, the VP who signed off on a process in 2019 and has since left the company without telling anyone why the process exists. That’s the hard part. The software does what you configure it to do, cheerfully and forever, including the dumb things.
The Stuff You Already Pay For and Never Turned On
Start here, because it’s free. Not free like a trial. Free like you are already being invoiced for it.
Below is what I find dormant most often in mid-market accounts. I’ve noted roughly who owns each one and how much work it is, because the honest answer for about half this list is “your admin, this week, no consultant required.” Anyone telling you otherwise wants a statement of work.
| Capability | What it usually replaces | Who can own it | Realistic lift |
|---|---|---|---|
| SuiteAnalytics Workbook | The saved search someone exports to Excel and pivots by hand | Controller or a sharp analyst | Days |
| Scheduled saved search alerts | The recurring “can someone check whether that shipped” message | Admin | Hours |
| SuiteFlow workflows | Approvals chased over email and eventually over Slack | Admin | Days |
| Reorder points and demand planning | A buyer’s instinct, plus the stockout nobody saw coming | Ops, with real historical data | Weeks |
| Bank feeds and auto-match reconciliation | Two days of somebody’s month-end | Accounting | Days |
| Custom roles and role-based dashboards | Everyone sharing one overloaded role and seeing everything | Admin | Days |
| SuiteApprovals | Purchase orders approved by forwarding a PDF | Admin | Hours |
| NetSuite AI Connector and SuiteAgents | Asking your one analyst for a number and waiting until Thursday | IT plus admin | Weeks |
None of that is exotic. It’s the boring middle of the product, and the boring middle is where the money is. The reorder point and demand planning row is the one that bites hardest, and I broke that out separately in NetSuite inventory management and stockout prevention.
Your Saved Searches Are Telling On You
Run this diagnostic before lunch. Go count how many saved searches get exported to Excel every month, and find out what happens to them after they land.
If the answer is “somebody rebuilds a report in Excel and emails it around,” you don’t have a reporting problem. You have a trust problem, and the export is the symptom. People export because the number on the screen doesn’t match the number in their head, and rather than fix the configuration that causes the mismatch, everyone quietly agreed to keep a shadow copy. I’ve seen a company run a nine-tab workbook that existed purely because two subsidiaries booked freight to different accounts. Fixing that took an afternoon. The workbook had been alive for three years.
Marcus, from earlier, was a shadow copy with a salary.
The fix is rarely clever. Find the exports, ask what the person does to the data after it lands in Excel, and then go make NetSuite do that part. Nine times out of ten it’s a grouping, a formula column, and a filter that nobody knew SuiteAnalytics could handle. The tenth time it’s a genuine gap and you build something. Fine. Build it. Start by assuming the platform can already do the thing, though, because it usually can, and because the alternative is commissioning custom work to replicate a feature you have owned since the day you signed the contract.
Stop Paying People to Be Middleware
The other place value leaks is between systems.
Every growing company ends up with a stack. Shopify or BigCommerce on the front, a 3PL or a warehouse system in the middle, Avalara for tax, a bank, maybe EDI with a big-box retailer who will fine you for getting a label wrong. NetSuite sits in the center of that and it is supposed to be the system of record. What actually happens is a person becomes the integration. They download a file from the 3PL portal, massage it, and import it. Every morning. Forever.
That person is middleware. Expensive, slow, and they take vacations.
Real integration, through Celigo or Boomi or a straight SuiteScript build, is not a big scary project at this point. It’s well-trodden. The reason it doesn’t get done is that the manual process technically works, and things that technically work don’t get funded. Then the person quits. Then it’s an emergency, and emergencies cost about four times as much as the same work scoped calmly in a quarter when nothing was on fire and you had the luxury of picking the right approach instead of the fastest one.
I watched this go the right way at justingredients.com. They were growing hard, past $150 million, running disparate systems that had each made sense individually and made no sense together. We built the integrated stack around NetSuite, ecommerce and wholesale both. They’re north of $250 million now with actual visibility into financials and operations at the same time, which sounds obvious and is genuinely rare. Same pattern at zamp.com, where we designed, built, and launched an integrated NetSuite product in about five months and they can now sell it in the NetSuite marketplace. And at accuserve.com, project-level profitability for retail construction used to take months of manual spreadsheet work. It lives inside NetSuite now and it runs itself.
The largest one I’ve personally stood up was a $250 million wholesale and ecommerce business, full tech stack, roughly seven months start to finish. I mention the timeline because people assume this class of work takes years. It doesn’t, when the org decides quickly.
That last clause is doing a lot of work. The org deciding quickly is the whole variable.
The AI Part, With the Hype Removed
You are going to get sold AI for NetSuite this year. Probably this quarter.
My mental model is crude and it has outlasted every framework anyone has handed me. AI is a high speed idiot. Fast, tireless, happy to do a thousand of something before you finish your coffee, and possessed of no judgment whatsoever. So you point it at simple, well-defined, low-stakes work, the way you’d hand a task to an intern on their second day. You don’t hand the intern the audit.
Practically, in a NetSuite context, that means three things worth doing right now.
The MCP connector is the one people ask about, and it deserves the attention. MCP is standardization, nothing more mystical than that. Think ramen noodle soup. Every ramen needs noodles and broth, and past that you add whatever you want. The standard part means your AI tools can talk to NetSuite in a predictable way instead of everyone building bespoke plumbing. Oracle documents what’s included in the NetSuite AI Connector Service, and the useful bit is that a lot of what used to be a five-hour saved-search-and-pivot exercise becomes a question you ask in plain English.
Second, SuiteAgents for the tasks that are annoying rather than hard. Document handling, summarizing, routing. Third, and this is the one my own team runs on, AI-augmented development against your existing codebase. Our technical group hasn’t grown in headcount while demand has climbed, and we’ve held delivery quality by tuning against nine years of our own NetSuite code. That’s not a product pitch. It’s just what worked.
What I would not buy is an “AI readiness assessment.” Look at the apps you already own, find the AI features already shipped inside them, and turn one on. Congratulations. You’re AI ready.
A 90-Day Optimization Plan You Can Actually Run
Sequence matters here, so this one is genuinely numbered.
- Weeks 1 and 2, inventory the workarounds. Not the software. The workarounds. Every spreadsheet, every export, every “we just know to check that.” Sit with finance, ops, and the warehouse separately, because they will each tell you a different story and all three are true.
- Weeks 3 and 4, audit what you’re licensed for versus what’s live. Pull your account’s feature list and walk it against reality. This is where people find modules they’ve been paying for since 2022 and never enabled. It’s a slightly humiliating meeting. Have it anyway.
- Week 5, fix the data before you automate on top of it. Duplicate vendors, dead items, customers with three records. Automating on bad data just gets you wrong answers faster, which is the high speed idiot problem in miniature.
- Weeks 6 through 9, kill the top five manual processes. Rank them by hours consumed, not by how annoying they are. Those are different lists and the second one is a trap.
- Weeks 10 and 11, rebuild reporting around roles. The CFO, the ops lead, and the buyer should each land on a dashboard that answers their question without a click. If they’re still exporting after this, you missed something. Go back.
- Week 12, decide what recurs. Optimization is not a project you finish. Put a quarterly review on the calendar with a named owner, or it silently stops happening again.
Everybody skips step six. It’s also the only one that decides whether you are sitting in this exact same position eighteen months from now, reading a different article, recognizing a different spreadsheet, and wondering how the drift happened again so quickly.
Sometimes It Isn’t the Software. It’s That You’re One Person Short.
I’ll be honest about the limit of what a consultant can fix.
A large share of the accounts I walk into don’t need a bigger project. They need a competent NetSuite administrator who owns the system on Tuesdays when nothing is on fire. Companies that grow past $50 million and refuse to hire someone smart enough to manage that growth end up buying consulting hours instead, forever, at a much worse rate, and they will tell you with a straight face that they can’t justify the headcount while writing checks that would cover two of them. It’s the pattern. Nobody names it out loud.
What that person looks like depends on where the pain is. If your problem is configuration drift and nobody owning the instance, that’s an administrator. If your problem is that finance processes don’t match how the business runs, that’s a functional consultant. If your problem is that four systems don’t talk, that’s an integration specialist, and that’s a different animal from either of the other two.
KORE1 staffs all three, which is the reason this piece is on their site. NetSuite administrator staffing for the ownership gap, NetSuite consultant recruiting for the process gap, and NetSuite integration specialists when the stack is the problem. Two newer specialties matter here too, because both showed up earlier in this piece. NetSuite MCP and agentic AI consultants for the AI work, and WMS integration consultants when the warehouse is the bottleneck. If the work is a defined burst rather than a permanent seat, contract staffing is the cheaper shape and there is no shame in it.
On the money question, McKinsey’s research with the University of Oxford, spanning more than 5,400 IT projects, found that large builds ran 45 percent over budget and delivered 56 percent less value than predicted. That study is older now and the number that has aged best is the value one. Budget overruns get noticed. Value shortfalls don’t, because nobody is watching the second number after the party.
If you want the hiring side of this in more depth, KORE1’s NetSuite implementation guide covers who sits in which seat, and their NetSuite implementation cost breakdown handles the budget version. This piece assumes you’re already live.
Things People Ask Me About This
How do I know if my NetSuite instance actually needs optimization?
Count the spreadsheets. If your team exports data out of NetSuite to make a decision more than a few times a month, the configuration is behind the business and you have work to do.
The other tell is the month-end close. If it takes longer than it did a year ago, something got bolted on that shouldn’t have been. Closes should get faster with time, not slower. When they slow down, it’s almost always accumulated customization nobody documented plus data that’s gone soft.
Is this something my own admin can do, or do I need a partner?
Probably half of it is internal work. Dashboards, saved searches, workflows, approvals, and role cleanup are all admin-level tasks if you have a real admin with time carved out.
Where outside help earns its rate is integration architecture, anything touching SuiteScript at volume, multi-subsidiary or multi-currency structure, and any change that affects the general ledger. Those have blast radius. The rest doesn’t, and paying consultant rates for dashboard cleanup is a waste of your money and frankly a boring use of ours.
We just went live eight months ago. Isn’t it too early for this?
Eight months is close to ideal. You’ve been through enough closes to know what’s actually broken, and you haven’t yet built two years of workarounds that people are emotionally attached to.
The worst time is year three. By then the shadow spreadsheets have owners, and the owners have opinions. At that point you’re not running a systems project. You’re running organizational therapy.
What does optimization work typically cost?
It ranges enormously and anyone quoting a number without seeing your instance is guessing. A focused engagement on a few processes is a fraction of an implementation. A full stack rebuild approaches one.
More useful than a range is the comparison you should actually be running. Take the fully loaded cost of the people currently doing manual work, annualize it, and put that against a fixed-scope engagement. In the freight-account example I mentioned earlier, an afternoon of configuration retired a workbook that had been consuming several days a month for three years. That math is not close.
Everyone is pitching us AI features. What’s worth turning on now?
Start with AI capability already bundled into apps you own, since most of it is implemented and waiting. Then look at the MCP connector for querying NetSuite in plain English, and SuiteAgents for repetitive document and routing tasks.
Skip anything sold as a readiness or maturity assessment. Those exist to bill you for a slide deck. Pick one simple, well-scoped, low-risk task, point the high speed idiot at it, and see what happens. You’ll learn more in a week of that than in a quarter of strategy.
Our data is a mess. Do we clean it first or automate first?
Clean first, always. Automation applied to bad data produces wrong answers at scale, faster than a human would have produced them, and with more confidence.
You don’t need perfect. You need the records that feed decisions to be trustworthy, which in practice means items, customers, vendors, and your chart of accounts. Everything else can be cleaned as you go.
How often should we be doing this?
Quarterly, with a named owner and a standing meeting. A half day every three months prevents the eighteen-month drift that turns a tune-up into a rescue.
The companies who do this well treat it like equipment maintenance rather than a project. Nobody schedules a heroic annual initiative to change the oil.
The Short Version
Every business is underutilizing tech. Yours included, mine too on the days I’m honest about it.
The gap between what NetSuite can do and what your team does with it on a Wednesday is the cheapest performance improvement available to you, and closing it needs no new platform, no new vendor, and no second selection committee. Let’s graduate our tech stack from the 90’s to at least the 2000’s. Then go find the spreadsheets. They know where the problems are.
If you’re staring at a shadow workbook right now and recognizing yourself, hit me up on LinkedIn. If the honest answer is that you need a person in the seat rather than another project, talk to the KORE1 team instead. That’s their lane, not mine.
