Ask any consultant what derails ERP projects and you will hear the same answer. It is not the software. It is not the budget. It is the data.
Data migration is the least glamorous part of a NetSuite implementation and the most common source of delays, cost overruns, and painful go-lives. Companies spend months evaluating software and days thinking about the data that will run through it.
The good news is that data preparation is entirely within your control, and you can start long before the project kicks off. At Anchor Group, the clients who arrive with clean, well-understood data consistently go live on time and under stress. This guide covers what they do differently.
Why Data Preparation Matters More Than You Think
NetSuite is only as good as the records inside it. Migrate duplicate customers and your AR aging is wrong from day one. Migrate items with bad costs and every margin report lies to you. Migrate open transactions incorrectly and your first close in NetSuite becomes a forensic exercise.
There is a second, less obvious cost. Every hour your implementation team spends untangling data is an hour not spent on configuration, testing, and training. Dirty data does not just create bad records. It starves the rest of the project.
The rule we give every client is simple. Garbage in QuickBooks becomes garbage in NetSuite, except now the garbage drives your entire operation instead of just your books.
The Six Data Domains You Need to Address
Every implementation moves data in roughly six categories. Each has its own cleanup priorities.
1. Customers
Start with duplicates. Legacy systems accumulate them for years: “Acme Corp,” “ACME Corporation,” and “Acme Corp.” all holding fragments of the same relationship. Merge them before migration, because merging after is far harder.
Then standardize. Decide on naming conventions, address formats, and how you will structure parent-child relationships for customers with multiple locations or subsidiaries. NetSuite handles customer hierarchies well, but only if you define the hierarchy before loading.
Finally, decide what not to bring. A customer who last ordered in 2017 probably does not need to come along. Set an activity cutoff and archive the rest.
2. Vendors
The same duplicate and standardization work applies. Add one more step: verify remittance details, payment terms, and 1099 status for every active vendor. Migration is the perfect forcing function for vendor master hygiene that finance has postponed for years.
3. Items
Items are usually the messiest domain, especially for product companies. Work through these questions:
- Which SKUs are active, and which are dead weight from discontinued lines?
- Are your item types correct? Inventory items, assemblies, kits, non-inventory items, and services behave very differently in NetSuite.
- Are costs accurate? Standard cost, average cost, and last purchase price often disagree in legacy systems.
- Do you have consistent units of measure, or does the same product appear as “each” in one place and “case” in another?
- Are bills of materials complete and current for anything you assemble?
Item cleanup takes longer than anyone expects. Start it first.
4. Chart of Accounts
Do not lift and shift your old chart of accounts. This is a once-a-decade chance to rebuild it around how you want to run the business, not how a bookkeeper set it up in 2011.
NetSuite adds dimensions your old system likely lacked: departments, classes, and locations that segment every transaction. Many accounts that existed only to separate divisions or product lines can collapse into a cleaner structure with segmentation handled by dimensions. Design the chart and the segment structure together.
5. Open Transactions
Open AR, open AP, open sales orders, open purchase orders, and unbilled commitments all need to move. The standard approach is to migrate open balances and open documents rather than full history, then keep the legacy system available read-only for lookups.
Reconcile everything before cutover. Every open invoice in the migration file should tie to the AR aging on the same date. If the numbers do not tie in the old system, they will not tie in the new one.
6. Historical Data
Decide early how much history you actually need inside NetSuite. Full transactional history sounds appealing and is rarely worth the cost. Most companies land on two years of summary financials for comparatives, plus complete open transactions, and keep deeper history in the legacy system or a data warehouse.
The Timeline: When to Do What
Data preparation should start 60 to 90 days before your implementation kicks off. A realistic sequence looks like this:
- Days 1 to 15: Audit. Export your masters and profile them. Count duplicates, blanks, and inconsistencies. You cannot plan a cleanup you have not measured.
- Days 15 to 45: Clean masters. Customers, vendors, and items, in that order of difficulty reversed. Items last is a mistake; they take the longest, so run them in parallel.
- Days 45 to 60: Design decisions. New chart of accounts, segment structure, naming conventions, and the keep-or-archive rules for history.
- Days 60 to 90: Trial extracts. Produce migration-format files and review them as if they were already in NetSuite. Every error found here is ten times cheaper than one found during testing.
Who Should Own This
Assign a single internal data owner with real authority. Committees do not clean data. This person makes the judgment calls: which duplicate record survives, what the naming convention is, where the history cutoff sits.
Your implementation partner supplies the templates, the validation, and the loading, but they cannot make your business decisions for you. This is one reason the choice of partner matters so much. Experienced NetSuite implementation partners have migrated data from every legacy system you can name, and they will tell you upfront which decisions are yours, which are theirs, and in what order the work has to happen. A partner who hands you a spreadsheet template and disappears until load week is a warning sign.
Common Mistakes to Avoid
A few failure patterns come up so often they deserve explicit warnings:
- Cleaning data in the migration files instead of the source system. Fix records at the source so every subsequent extract is clean. Fixing files creates version chaos.
- Letting perfect block good. You will not fix every record. Prioritize by transaction volume: the 500 items that drive 95 percent of revenue matter more than the long tail.
- Skipping the reconciliation. Trial balances, AR, AP, and inventory value must tie between systems on cutover date. No exceptions.
- Freezing changes too late. Set a cutoff date after which no new items or customers enter the legacy system without also entering the migration plan.
Getting Help Before the Project Starts
If your data situation feels overwhelming, that is normal, and it is also solvable. Many companies engage NetSuite consulting services for a data readiness assessment before committing to a full implementation. A short engagement produces a data audit, a cleanup plan, and a realistic migration scope, which in turn produces a more accurate implementation quote. It is a modest investment that removes the single largest unknown from your project.
How to Validate the Loads
Once data starts moving, validation is what keeps small errors from becoming systemic ones. Build these checks into every load cycle:
- Record counts. The number of customers, vendors, and items extracted should equal the number loaded, minus documented exclusions. Any unexplained gap gets investigated before the next load.
- Financial tie-outs. Trial balance, AR aging, AP aging, and inventory valuation must match between systems as of the cutover date, to the penny. Print both reports, staple them together, and have the controller sign off. This document becomes your audit trail.
- Spot checks by the people who know. Have your best customer service rep pull up ten accounts they know cold and confirm the records look right. Have the warehouse manager check twenty items. Domain experts catch errors that scripts miss.
- Transaction tests. Enter a sales order, a purchase order, and an invoice against migrated records in the sandbox. Data that loads cleanly can still behave badly in transactions if item types or account mappings are wrong.
Run the full validation on every trial load, not just the final one. The goal is a cutover weekend with no surprises, and the only way to get one is to have already done the cutover twice in rehearsal.
The Payoff
Clean data does not just make go-live smoother. It compounds. Every report you run, every reorder NetSuite calculates, every invoice it generates is built on the records you migrate. Companies that invest in data preparation spend their first year in NetSuite improving the business. Companies that skip it spend their first year fixing records.
Start early, assign an owner, clean at the source, and reconcile relentlessly. Do those four things and data migration becomes what it should be: a milestone on the schedule instead of the reason you miss it.