ERP Readiness Assessment: The Free Six-Check Health Check
Six counts you can finish in a week, without a vendor, that tell you whether your ERP project is ready to sign or one seat short.
The Floor Line
Six scores, one verdict. Your readiness is the lowest bar, not the average of them.
An ERP readiness assessment scores six dimensions of your business before implementation, data, process, ownership, bench, integration and adoption. Your readiness is the lowest score, not the average. This free version takes about a week.
Most companies decide they are ready the week the budget clears. That is a finance milestone. It is not a readiness one, and the distance between the two is where ERP projects quietly go wrong, usually surfacing nine months later as a go-live date that moves twice and a data migration that turns into its own project with its own budget and its own apology email.
Our 2026 mid-market ERP and AI adoption report puts a number on the gap this assessment measures, with only 16% of mid-market companies reaching a fully governed, integrated data state. Gartner’s guidance on ERP initiatives puts the odds plainly. By 2027, more than 70% of recently implemented ERP projects will fail to fully meet their original business case goals, and as many as 25% will fail catastrophically. Almost none of those are software failures. They are readiness failures.
What follows is six counts. No download. You run them yourself, with a spreadsheet and about a week, and nobody asks for your email at the end. Each check produces a number, a threshold, and the role that number names if you miss it, which is the part a vendor questionnaire leaves out because the vendor would rather sell you the seat than help you fill it. We are a staffing firm. We would rather tell you which chair is empty, and when the answer is a hire, our ERP recruiters fill those seats across NetSuite, SAP, Dynamics, Oracle and Epicor.
Readiness Is a Floor, Not an Average
Score six dimensions and the instinct is to average them. Resist it. A company that scores well on five and badly on one is not 83% ready. It is not ready.
ERP is one connected system, which is the entire reason anybody buys one, so a break in any single dimension travels. Dirty item master data does not stay in the item master. It shows up in the first physical inventory count, then in standard costing, then in a month-end close that runs eleven days because three people are reconciling by hand while the CFO explains the delay to a board that was promised the opposite. Nobody budgeted eleven days.
McKinsey and the University of Oxford studied more than 5,400 IT projects and found the large ones run 45% over budget while delivering 56% less value than predicted, with every additional year on a project adding roughly 15% to the overrun. The compounding is the whole story. A gap you carry into month one is not the same gap in month fourteen.
So the floor is the score. Find the lowest dimension, fix that one, count again. Not glamorous. It works.
Run These Six Counts
Work through them in order. Data sits first because everything downstream inherits it, and adoption sits last because you cannot train people on a process nobody has agreed to yet.
Your master data has never actually been counted
- Ask
- Export 200 random rows from your customer or item master. One person checks every row for duplicates, blank required fields, and records nobody archived when the account went dormant.
- Count
- Rows carrying at least one defect.
- Broken at
- 15 or more
What your count means
Fifteen out of 200 is 7.5%. That is the line.
Count before you estimate. Under it, cleanup is a task somebody squeezes in. Over it, cleanup is a workstream with an owner, a schedule and a budget line, and pretending otherwise is how a six-week migration becomes a five-month one that nobody planned for and everybody ends up blaming on the software. The role this names is a data migration lead, usually contract, usually the integration and migration specialist profile.
Four people draw order-to-cash four different ways
- Ask
- Ask four people from four departments to sketch how an order becomes cash. Separately, on paper, with no documents open in front of them.
- Count
- Steps that appear on all four sketches.
- Broken at
- Fewer than half
What your count means
This one embarrasses people. Run it anyway.
Sketches beat documents. A process nobody can draw is a process you are about to configure from memory, and configuring from memory is how a company automates a workflow that three of its four departments had already built quiet workarounds for. Sort it before the design workshops rather than during them. The role is a business analyst who can hold a room and write the as-is down without flattering anybody in it.
Nobody can approve a scope change alone
- Ask
- Ask five stakeholders to name the one person who can approve a scope change without escalating it to somebody else.
- Count
- Distinct names you get back.
- Broken at
- Zero, or three and up
What your count means
Zero means no decision maker. Three means no decision.
Both stall the same way. A scope question lands in a Thursday meeting, gets deferred for context, returns the following Thursday with a new stakeholder attached, and by the third cycle the consultants are billing to wait. Then it repeats. That loop can eat six or seven weeks before anybody says it out loud. The role is an ERP implementation project manager with real authority, not a coordinator holding a RACI chart.
The core team still has a full day job
- Ask
- Ask every named core-team member how many hours a week they have actually been released from their existing role. In writing, backfilled, not verbally agreed.
- Count
- People genuinely above 20 hours a week.
- Broken at
- Fewer than two
What your count means
Everyone says yes at kickoff. Then quarter close happens.
The controller who was going to own the chart of accounts redesign spends October doing October, and the structural decision gets made without her by people who do not know why the old structure exists. Backfill costs less than rework, which sounds exactly like something a staffing company would say, so here is a test that does not require trusting us. Price the rework. Contract staffing exists for this window specifically.
The interface list is shorter than reality
- Ask
- List every system that reads from or writes to whatever the ERP replaces. Include the spreadsheet on somebody’s desktop that the warehouse actually runs on.
- Count
- Interfaces with a named owner and a documented field mapping.
- Broken at
- Under half your list
What your count means
The desktop spreadsheet counts. It always counts.
Shadow integrations surface in user acceptance testing, three weeks out from cutover, when somebody asks how freight quotes reach the order and learns that a scheduler has been pasting them in every morning at seven for four years. Four years. Each one is either an interface you build or a manual step you inherit permanently. The role is an API and integration architect, and on multi-system builds that person earns the rate back inside the first month.
Training is a line item nobody has costed
- Ask
- Ask for the number of training hours budgeted per end user, and the name of the person writing the material against your configuration.
- Count
- Hours per user, with an author attached.
- Broken at
- Under four hours
What your count means
Four hours is the floor. It is not a target.
A vendor’s train-the-trainer package is not a training plan, it is a deck built against a demo tenant that shares very little with your configured system, and the gap between the two gets filled by whichever supervisor happens to be most patient. Adoption failure looks like a clean go-live for about six weeks. Then it unravels. The workarounds start, and the data quality problem you fixed in D1 comes back with friends. The role is a change lead, which on mid-market builds usually rides with a digital transformation consultant or a fractional CIO.
Two broken dimensions is the normal first-pass result, not a red flag about your company. Count first. Argue afterwards. The counting is what keeps a readiness conversation from turning into a debate about whose department is the problem, because a number on a page is harder to take personally than an opinion in a meeting.
What Your Count Means
Lock scope now. Readiness decays, so a clean score in March is not a clean score in September.
Sign, then close the gap inside 30 days and name its owner in the statement of work.
Fix before signature. Anything you carry past contract gets rediscovered at consulting rates.
Trailing 12-month average, with 92% retention at the one-year mark.
Bands are KORE1’s ERP practice framework, drawn from the searches we run alongside live implementations. Time-to-hire and retention are KORE1’s own placement numbers.
When the Gap Is a Seat, Not a Sprint
Four of the six checks fail for the same underlying reason. Nobody owns the thing.
That is a hiring answer, and it is the one your systems integrator will not volunteer, because their incentive is to place their consultant in the seat rather than help you fill it permanently. That is the tell. Their consultant leaves at go-live. Yours does not, and the difference shows up in year two when somebody has to explain why the tax logic was built the way it was.
ERP hiring tends to split into two shapes. Two clocks, really. Contract and contract-to-hire cover kickoff through hypercare, when the work is intense and finite. Permanent hires cover everything after, because a company that just spent eighteen months and seven figures on a platform needs somebody in-house who knows why every switch is set where it is.
Platform choice changes who you hire, not whether the six checks apply. We staff SAP S/4HANA, Dynamics 365, Oracle Cloud ERP, Epicor and Sage Intacct and Workday teams, plus manufacturing ERP builds where the plant floor is in scope, all off the same IT staffing bench. Still costing the project? The NetSuite implementation cost calculator is the companion tool to this page.
Three Things to Do Before the Next Vendor Call
Readiness work has a reputation for being slow and abstract. These three are neither. Each takes an afternoon.
Count 200 rows
Pull the sample this week and put the defect count in the project charter before anybody estimates the migration.
Name one approver
Write a single person’s name against scope authority, then say it out loud at kickoff where the whole room hears it.
Book the backfill
Release your core team in writing and cover their day jobs before design workshops start, not after they slip.
Run It Before the Demo, Not After the Contract
When you run this changes what it is worth. Timing is the difference. Before vendor demos, it shapes your requirements. After signature, it is a list of things you are now paying consultants to discover on your behalf.
There is a second reason to go early. A demo is engineered to be impressive, and an impressive demo makes readiness gaps feel like implementation details, which is how a company with a broken ownership check talks itself into a fixed-price contract that quietly assumes decisions get made in 48 hours. This is not only a mid-market problem. The GAO’s 2025 high-risk review of federal IT found the same pattern at national scale, where investments too frequently fail or slip while contributing little to the mission they were bought for.
Re-run the counts at three points. Before selection, at design sign-off, and 30 days out from cutover. Same six questions each time. The numbers move before the confidence does, and having them written down is what lets a project sponsor push a go-live date without the conversation turning into an argument about who is being negative.
Common Questions
What is an ERP readiness assessment?
It is a structured check of whether your business, rather than your software, can absorb an ERP implementation. This version scores six dimensions, data quality, process agreement, scope ownership, internal bench, integration inventory and adoption planning. Your readiness is the lowest of the six, never the average of them.
How long does this ERP health check take to run?
About a week of calendar time and roughly six hours of real work. The data count is the longest piece at two to three hours. The rest are conversations. You can hold them between other meetings. Nobody needs to block a sprint or buy a tool, and you do not need to tell your vendor you are doing it.
Do we have to give you an email address to see the results?
No. The whole assessment is on this page, thresholds included, with no form, no gate and no PDF to download. We publish it ungated because readiness conversations are where staffing gaps surface, and those gaps are our business. The assessment itself is not. Same deal with our free engineering velocity assessment and our AI readiness scorecard.
What if we fail two or three of the six checks?
Two broken dimensions is the normal first-pass result, so treat it as a starting position instead of a verdict on your company. Fix the lowest one, count again in four to six weeks, then move to the next. Three or more broken usually means the project is being driven by a budget cycle rather than a business case, which is worth naming before anybody signs.
Can we run this before we have picked a platform?
Yes. It is the best time to run it. None of the six checks depend on which vendor you choose. Platform affects who you hire and what the integration work costs, not whether your master data is clean or whether anyone owns scope. Running the counts first also gives you much sharper questions for the demos.
How is this different from a vendor’s readiness questionnaire?
Two things. Every check here ends in a number you count rather than a scale you rate yourself on, and every failed check names the role that closes it. A vendor questionnaire is scoped to the vendor’s implementation, so it tends to skip the parts a systems integrator would rather bill for than fix. Gartner’s own ERP guidance puts more than 70% of recent implementations short of their original business case by 2027, and the misses cluster in the six areas above. McKinsey and Oxford’s study of 5,400 IT projects found the same pattern in cost terms.
Counted All Six? Bring Us the Lowest One.
Tell us which dimension came back broken and we will tell you whether it reads as a hire, a backfill or a process fix, in writing, before you commit to anything. No pitch deck.
Ready, but undecided on the platform? The ERP comparison tool scores NetSuite, SAP, Dynamics 365 and Epicor against your constraints and shows which two are worth a demo. For the evaluation itself, ERP vendor selection advisory puts an independent advisor on requirements, the RFP and the scripted demos.
