The short version: Yes. I build the new app to match the old one screen for screen first, and only then change anything. For a firm under 20 staff that is a fixed $4,000 to $10,000 into SharePoint lists or $8,000 to $18,000 into Dataverse. Most screens can be identical. A few things cannot, and I list them before I start.
The rule: same first, better second
A migration that also redesigns the workflow is two projects wearing one price tag, and the second one is where jobs overrun. So I split them. Phase one is a like-for-like copy that your staff can pick up on a Monday without a training session. Phase two is a short, priced list of improvements you choose once the copy is live and trusted. Nothing from phase two gets into phase one, however sensible it sounds in week two.
The old system stays running and linked while the copy is built, so there is no deadline forcing a rushed switch. How long that takes by source and destination is on how long moving off an old database takes; this post is about what "the same" means and how you prove it.
What can be identical, and what cannot
| Part of the old system | In the copy | Why |
|---|---|---|
| Tables, fields, field names, row counts | Identical | They move with the data, so the vocabulary your staff use stays |
| Data entry forms, in the same field order | Identical in layout and tab order | Muscle memory is most of "how we work" |
| The three to five reports people actually print or export | Identical in columns, sorting and totals | The totals are the test that the data came across intact |
| Lookups, drop-downs and default values | Identical | Rebuilt from the old definitions, not from memory |
| Macros, VBA and stored logic | Same result, different mechanism | They become flows, business rules and formulas, because the new platform does not run the old code |
| Speed and look of the screens | Close, not pixel-identical | A web-style app draws lists and buttons its own way |
| Keyboard shortcuts and a right-click menu | Rarely identical | Ask the two heaviest users which ones they really use, then decide case by case |
| Anything that only worked on the office PC | Replaced | A label printer, a linked spreadsheet or a shared drive path has to be re-plumbed |
The middle of that table is the honest answer to the question in the title. The tables, forms and reports can be the same. The code behind them cannot be carried across, only recreated, and that rebuild is where the money and the weeks go, as on moving an Access database to Power Apps and moving SQL Server to Dataverse. Which of those two your system is decides the band above.
How I prove it is the same
- Write down the old system first. Every screen, report and button, with a screenshot of each, before anything is built. This is the spec, and it is the document I hold myself to.
- Pick five real records. Two ordinary, one awkward, one old, one that only the boss understands. They must look identical in both systems.
- Reconcile the reports. Run the month-end report in both and compare the totals. A difference of one cent is a bug, not rounding, until shown otherwise.
- Run a real day in parallel. The two people who use the old system most enter the same work in both for a day. Anything that takes them noticeably longer is fixed before go-live.
- Sign off screen by screen. A tick against the phase-one list, by the person who uses that screen. Not by me.
Whether to test on made-up records first is covered in building on fake data before the real records. The step-by-step for each source is on the migration hub, and the 38-point SQL migration checklist is the working document behind the sign-off.
What to do with "while you're in there"
Every owner has a list, and most of it is good. I keep it in a parked list with a rough size next to each item, and nothing on it is built until phase one is signed off. This is what protects the fixed price: an improvement asked for mid-build is a second project, and pricing it separately means you can say yes or no to it without touching the first. When phase one has been live for a couple of weeks, the list is shorter, because some of it turns out to have been the old system's quirks rather than real needs.
When a like-for-like copy is the wrong plan
- The old workflow is the problem. If staff work around the database with spreadsheets and paper, copying it faithfully copies the workarounds. Redesign first, and use the workflow review.
- Nobody can say what the old system does. If the forms and macros are undocumented, the inventory comes first and I plan for the top of the band.
- The old platform is what is failing. Sometimes the honest answer is a packaged product, not a rebuild; when Power Platform is the wrong answer covers it.
Costs beyond the build: Dataverse means Power Apps Premium at NZ$32.40 per user per month, so NZ$324 a month for ten people; SharePoint lists cost nothing on top of Microsoft 365. Build bands are my fixed prices for firms under 20 staff, from the services page. Licence price is Microsoft's NZ list, September 2026.
Common questions
Can Power Apps look and work exactly like my old Access or desktop database?
Close, screen for screen: the same tables, fields, field order, lookups and report totals. It cannot be pixel-identical or run the old macros and code as they are. Those are recreated as Power Automate flows, business rules and formulas, so the result is the same, reached differently.
How much does a like-for-like move cost?
For a firm under 20 staff, $4,000 to $10,000 into SharePoint lists or $8,000 to $18,000 into Dataverse, fixed price, with data migration, validation and cut-over included. Dataverse adds NZ$32.40 per user per month for Power Apps Premium; SharePoint lists add nothing on top of Microsoft 365.
Why not improve the workflow while moving it?
Because the improvements become a second project inside the first one, and that is what blows the schedule. Copy first, sign it off, then choose improvements from a priced list. Some of the wish list disappears once the old system's quirks are gone.
How do I know the new app matches the old one?
Compare five real records, run the month-end reports in both and match the totals, then have the two heaviest users enter one real day of work in both systems. Each screen is signed off by the person who uses it.
Want the same app, just newer?
Tell me what the old system is and who uses it. One free 30-minute call and you get a band, a schedule and the list of things that cannot be copied.