Somewhere in almost every system switch, someone says, “We’ll just migrate everything.” It sounds like the safe choice. It’s also usually the most expensive way to be safe, and it’s the heart of the data migration vs. data archiving question.
Migration and archiving both deal with data from a system you’re leaving behind, so they get treated as the same job. They aren’t. Migration puts data to work in your new system. Archiving keeps data safe and findable without it living there. Most projects need some of each, and the real work is deciding which records go where.
This guide gives you a simple way to make that call, a worksheet to bring to your first planning meeting, and the mistakes we see most often. It’s written for pharmacy owners, IT leads, and operations teams, but the logic holds for health systems too.
The short version
- Migrate the data your team needs on screen in the new system to do their jobs.
- Archive the data you have to keep and be able to find, but rarely open.
- Retire the data that’s past every retention requirement, with compliance signing off first.
- Most projects use all three. Very few should use only one.
Data Migration vs. Data Archiving: What Each One Does
Data migration moves records from the old system into the new one and converts them to fit. That means mapping fields, cleaning duplicates, fixing formats, and testing the results. It’s detailed work, and the effort grows with every record type you include. An allergy that technically transfers but lands in a free-text notes field hasn’t been migrated well. It’s just been moved.
Data archiving pulls records out of the old system and stores them in a separate, secure, searchable repository. The records stay readable and retrievable. They just aren’t part of day-to-day work in the new system. Because the data isn’t being reshaped to fit a new structure, archiving generally takes less time and effort than migrating the same volume.
The three-bucket sort
Here’s the method we’d suggest for working through data migration vs. data archiving. Take every record type in your old system and put it in one of three buckets.
- Bucket 1: Live. Staff will need this on screen during a normal week. Think current patient profiles, allergies, active prescriptions, recent fill history, current insurance. These get migrated.
- Bucket 2: Reference. You have to keep it, and you have to be able to find it, but nobody opens it often. Think older fill history, signature logs, scanned documents, closed accounts. These get archived.
- Bucket 3: Retire. The data is past every retention requirement, has no legal hold on it, and has no ongoing use. After compliance or legal signs off, it can be securely disposed of according to your policy. Plenty of organizations skip this bucket and end up paying to store and protect old records indefinitely. It’s worth at least asking the question.
If you can’t decide between Live and Reference for a record type, ask a more specific question: Will someone need this during an ordinary workflow, or only when something unusual happens? A refill question at the counter is ordinary. An audit request from three years back is unusual.
What usually lands where
This is a typical starting point, not a rule. Your workflows, state requirements, and new system’s limits will move some of these.
| Record type | Usual starting point | Why |
| Patient demographics | Migrate | Needed for every transaction |
| Allergies and conditions | Migrate | Safety information must be visible at the point of care |
| Active prescriptions | Migrate | Needed for refills on day one |
| Recent fill history | Migrate | Used for profiling and refill questions |
| Older fill history | Archive | Rarely opened, but must stay retrievable |
| Insurance and billing history | Depends | Current coverage migrates, older claims data often archives |
| Prescriber records | Migrate active, archive inactive | Active prescribers are used daily |
| Signature logs | Archive | Mostly needed for audits and disputes |
| Scanned documents and images | Archive | Many new systems can’t accept them cleanly |
| Notes and free text | Depends | Recent notes may be worth migrating, older ones usually archive |
| Controlled substance records | Check requirements first | Federal and state rules apply |
| Closed or inactive accounts | Archive | Keep, but no daily use |
For health systems, the same logic applies. Current problem lists, medications, allergies, and recent encounters tend to belong in the new EHR. Older encounters, scanned charts, and data from legacy ancillary systems tend to be archive candidates, especially when a system has grown through acquisitions and is carrying several old platforms.
Six questions that settle most debates
When your team disagrees on a record type, these usually break the tie.
| Question | If yes, lean toward |
| Does staff need this during a normal week? | Migrate |
| Does the new system have a clean place to store it? | Migrate (if no, archive) |
| Is the data clean and consistent? | Migrate (if messy, consider archiving) |
| Is it mainly needed for audits, legal requests, or patient record requests? | Archive |
| Is it older than your realistic lookback window? | Archive |
| Is it past retention with no hold? | Retire, after sign-off |
The “is it clean” question deserves a note. Messy legacy data costs more to migrate, because someone has to clean it before it can be mapped. Archiving is more forgiving. That can be a real advantage, but it also means any quality problems come along with the data, so it helps to know what you’re storing.
A worked example
Here’s a hypothetical to make data migration vs. data archiving concrete.
An independent pharmacy has fourteen years of data in its current system and is moving to a new dispensing platform. Before kickoff, the owner asks the pharmacists and technicians to jot down every time they go looking for older patient history over one week. The pattern is clear: nearly all lookups reach back about two years, mostly for refill questions and insurance rejections. Almost nothing older gets opened except during the occasional audit.
So the plan looks like this:
- Migrate: patient profiles, allergies, active prescriptions, and the last two years of fills.
- Archive: everything older, plus signature logs and scanned documents.
- Review for retirement: records past the pharmacy’s retention requirements, flagged for compliance to review.
The two-year window is just this pharmacy’s answer. Yours may be different. What matters is that it came from asking the people at the counter instead of guessing. The result is a smaller migration, fewer fields to validate, less testing, and a better shot at hitting the go-live date.
The mistakes we see most
- Migrating everything “to be safe.” Every extra record type adds mapping, validation, and testing. It also adds risk to your go-live date. Safe and complete aren’t the same thing.
- Drawing the line without asking the people who use the data. If the cutoff gets set in a meeting that didn’t include pharmacists, technicians, or clinicians, the first week after go-live turns into a hunt for records that are sitting in the archive. A short conversation up front prevents most of that.
- Keeping the old system running “just in case.” This is probably the most common one. Licenses, hosting, and support contracts keep renewing for a system nobody wants to use, because nobody has confirmed the data is stored safely somewhere else. A tested archive is what lets you turn it off.
- Archiving without testing retrieval. An archive only counts if someone can actually find a record in it. Before you decommission anything, pick a handful of real records, including a scanned document and an older fill, and retrieve them. Do it with the people who will do it for real, not only the project team.
- Treating it as an IT-only decision. What stays live, what gets archived, and what gets retired involves operations, compliance, and clinical or pharmacy leadership. If it’s left to whoever happens to be on the project, it tends to default to “everything” or “nothing.”
A quick word on retention and access
Retention is the reason archiving exists, so it’s worth knowing where the rules come from.
HIPAA requires covered entities to keep certain compliance documentation for six years, but it doesn’t set how long medical or pharmacy records themselves must be kept. That mostly falls to state law and, for pharmacies, state boards of pharmacy. Controlled substance records are covered by DEA requirements, which set a federal minimum, and some states require longer. Payer and PBM contracts can add requirements of their own.
Access matters just as much as storage. Under HIPAA, patients have a right to access their records, and organizations generally have 30 days to respond. If your archive takes a week of digging to produce a record, that’s a problem waiting to happen. When you evaluate archive options, ask how long a typical retrieval takes and who can do it.
Because these rules differ by state and record type, confirm yours with your compliance officer or legal counsel before you lock in an archive policy.
Frequently asked questions
Is archiving cheaper than migrating?
Usually, for the same amount of data. Archiving keeps records in a searchable format without reshaping them for a new system, so there is less mapping and testing. Your final cost still depends on data volume, data quality, and how many source systems you have.
Can I do both in one project?
Yes, and most organizations do. In practice, data migration vs. data archiving isn't an either-or choice. Current, working records get migrated into the new system. Older reference records go to an archive so they stay retrievable without being converted. Planning both together keeps them on one timeline and avoids gaps where a record type falls through the cracks.
Can we shut down the old system once data is archived?
Usually, yes, once the archive covers every record type you are required to keep and you have tested retrieval with real examples. Check documents and images too, not only structured data, and confirm there are no legal holds or open audits before anything is decommissioned.
How long do we have to keep records?
It depends on your state and the record type. HIPAA sets a six-year retention period for its own compliance documentation, but not for medical records. State law and boards of pharmacy usually set those timelines, and federal rules like DEA requirements apply to controlled substances. Confirm the specifics with compliance or legal counsel.
Download the Data Migration vs. Data Archiving Planning Worksheet
Want to try the sort with your own team? Download our two-page planning worksheet. It covers the three buckets (migrate, archive, retire), a record-by-record sheet for twelve common record types with room to add your own, and prompts for your lookback window, legal holds, and retrieval test. There’s also a “not sure” option for records you need to check with compliance. Print it, fill it in together, and bring it to your first planning meeting.
Download the worksheet today.
Where to go from here
Once you’ve sorted your records into live, reference, and retire, the data migration vs. data archiving question gets much easier to answer, and the rest of the project gets easier to plan. You’ll know how much data needs mapping, what the archive has to hold, and what you can finally stop paying to keep around.
If you’d like a second set of eyes on the sort, the InfoWerks team can walk through your systems with you and help set the line. Reach out, and we’ll start with what you have.
This article was reviewed by Jeff Heilsberg, Chief Revenue Officer at Infowerks Data Services, with more than two decades of experience helping healthcare organizations simplify complex data ecosystems.
