Skip down to the research record

Bitwarden import limits: 3 checks before a team load

Quota Forecast Editorial Team · Published · 10 min read

Check folder relationships, collection relationships, and text field growth before importing shared logins into a Bitwarden organization.

A team vault load is a single import of shared logins into a Bitwarden organization. Three limits decide whether it goes through cleanly: the item-folder count, the item-collection count, and the size of text fields once Bitwarden encrypts them during import. Each limit is easy to miss, and each one can be checked before the load rather than after it fails.

The figures here come from Bitwarden's own import help page. If your team uses another tool, check that tool's documentation instead of reusing these numbers.

To keep things concrete, we follow one case through every section: a shared CRM login that a sales team uses daily. The login sits in three collections, the East sales collection, the West sales collection, and a renewals collection, because three groups need it. Its notes field holds the account owner, the recovery contact, and the setup steps the team has built up over several years. Your records will look different, but the same questions apply to each one.

Why a rep change raises the stakes

A routine migration leaves time to find problems. A rep change does not. The export often arrives late, the owner is buried in handoff work, and the new owner has not yet been told which logins they will need on day one. Every hour spent on a failed import is an hour when the team may have no working login for a renewal call.

That is the reason to read the limits before the load. They matter most when records are widely shared or kept in several places, and checking them takes little more than the export file, a spreadsheet, and a list of who owns each login.

Ownership comes first. Each login in the team vault needs an owner who can answer questions after the rep has gone. For the CRM login, that is the sales manager who approves access, not the rep who happened to create it. Someone has to be answerable for each record, and that person should be named before the export is taken.

Start with the destination

Before the limits matter, decide where the records are going. Bitwarden's import help page says the article "covers importing to your own vault or to a collection that you have access to." It adds: "If you're administering an organization, you may need to import to an organization instead."

For a team load, the organization path is usually the one you want. A login the team depends on should belong to the company, not to whoever ran the import. If the records land in one rep's personal vault, they depend on that person's account, which is exactly the situation a team load is meant to end.

Our CRM login belongs to the sales organization, so the destination is the shared organization. The outgoing rep's personal copies need a separate decision, which the folder section covers. Settle the destination first, because every count you take next should describe where the records will actually go.

One more step belongs here. Ask the outgoing owner to stop editing the logins once the export is taken. Record any later changes separately so they can be checked against the imported records.

Limit one: item-folder relationships

The first limit is 7,000 item-folder relationships. One item in one folder is one relationship. The same item in three folders is three relationships, even though the password is a single secret.

That gap is where most surprises come from. Reps often keep personal copies of shared logins in folders such as "Q3 accounts" or "Renewals watch," so one login turns up in several places at once. Each placement adds to the total, and a personal folder rarely feels like a placement until someone counts it.

In our example, the CRM login sits in one team folder in the old export. The outgoing rep also filed copies in two personal folders. That makes three placements, so the login alone counts as three item-folder relationships. Repeat that pattern across the team and the total climbs faster than the record count suggests. A row count is the wrong thing to check.

To measure it, open the export and count folder assignments for each record. If folders appear in one column, a record with several folder names has several placements. If the export uses a separate row for each folder, count the rows that share a login name. If you cannot tell how your export represents folders, load a small batch first and see how the total comes out before you trust the rest.

Write down the total for each rep. Then decide which personal copies are actually needed. Some may hold the only copy of a login the rep still uses, so delete nothing until the rep confirms. A cleanup that looks tidy can cut off access on the day of the handoff, and that is the one day nobody can afford it.

The count also says something about habits. A high number of personal placements is a reason to ask whether the team has been working around the shared vault. Use the load as a chance to ask what the rep still uses. A login the team never receives is one it cannot support after the handoff.

Limit two: item-collection relationships

The second limit concerns collection assignments. The import page lists 80,000 item-collection relationships and applies the same counting rule: "a single item in three collections is counted as three item-collection relationships."

The count grows with sharing, not with the number of login records. A login shared with one group adds one relationship. A login shared with several groups adds one for each group. A team that puts every login in an all-sales collection and also in a regional collection can push its count up without adding a single new login.

The CRM login follows that pattern. The East and West groups both work the account, so that sharing is real and should stay. Ask the renewals group lead whether its placement is still needed.

Collection membership also works as an access review. Keeper's enterprise page, for example, says it provides administrators with granular role-based enforcement policies, delegated administration and detailed visibility. Whatever tool you use, the question is the same: who can see this login, and who decided that? Ask each group lead whether the login belongs with their team. That conversation takes longer than the import itself, so start it early. It also catches logins that should never have been shared at all, which a folder-by-folder cleanup would miss.

Total the lines and compare the result with the figure on the import page, leaving room for changes that happen before the load. Treat the limit as a ceiling to measure against. The useful habit is counting before the load starts, so the number is known rather than discovered.

If a count sits close to the limit, the cleaner fix is to remove sharing nobody uses. Splitting one team's logins across two loads adds coordination during a handoff, when coordination is already the hardest part.

Limit three: field size during import

The third limit concerns field size after encryption. The import page says: "Bitwarden's encryption process expands text 30-50% during import, which may push your field(s) beyond the character limit." The length you see in the file is not the length the vault will see once the import runs.

The same import page names the field that causes the most trouble. Notes, it says, is "the most common offender," and a Notes field "can increase from 8,000 to over 10,000 characters, exceeding the limit and triggering the error." Notes collect text over time, so start there when you check field lengths.

The CRM notes hold a recovery contact and years of setup steps, so this is exactly the kind of field the limit can catch.

Check the current import page for the character limit before you measure anything against it.

To measure, sort the export by the length of each text field, starting with notes. Any field that sits close to the limit once you allow for growth of up to 50% needs a change. You can shorten it, or move long background into a document the team keeps elsewhere and leave in the vault only what the person needs to sign in, plus the one contact who can help. Plan for the top of the range, not the middle, because the top is the case that fails.

Test with the longest records early. The first batch should include the fields most likely to fail, not a random sample of easy ones. Keep the batch small enough to check by hand and large enough to include the awkward cases. A handful of records that covers folders, collections, and a long notes field teaches more than a long run of short logins.

Field length also shapes what you keep. A notes field full of old passwords or copied secrets is a risk in any vault. If the team is moving those logins anyway, decide now what belongs in the notes at all.

Use this comparison to plan the checks:

CheckWhat to countAction before the load
Item-folder relationshipsFolder assignments for each recordConfirm which personal copies are needed
Item-collection relationshipsCollection assignments for each recordReview membership with group leads
Field sizeLength of each text field, Notes firstShorten or move long background before the import

Put the three limits in order

Here is the order we would use. It keeps personal copies, team sharing, and long notes from colliding on load day.

First, confirm the destination. If the records belong to the company, load to the organization, and write the destination into the load plan so nobody has to guess later.

Second, count placements. Count folder and collection placements for each login, then write both totals down. Ask the rep to confirm personal copies before anything is removed, and ask each group lead to confirm collection membership.

Third, measure the text fields, especially notes, and allow for encryption growth. Fix or shorten any field that sits close to the limit.

Fourth, run a small test batch. Include the CRM login, the longest notes, and a login with several collections. Check each one in the vault against the export before loading the rest.

Fifth, load the remaining records and verify. Compare a sample of records in the vault with the export for folder placement, collection membership, and notes text. Keep the export file somewhere restricted while you work, because it holds login details, and delete it once the load is verified and the owner has confirmed.

For related record preparation, see sales data import checks.

Walk the CRM login through all three checks

For the illustrative CRM login, turn the counts and field measurements into decisions before the test batch.

Each check points to a different action:

CheckIllustrative CRM loginNext action
Item-folder countIllustrative: one team folder and two personal copiesTalk with the rep about the personal copies
Item-collection countIllustrative: East, West, and renewals collectionsAsk the group leads whether all three placements are still needed
Field sizeIllustrative: long notes with recovery and setup detailsShorten the notes or move the background details to a document kept elsewhere

Notice what the example does not need. It does not need a special tool, a new policy, or a second vault. It needs the export, a list of who owns the login, and a decision about where the long notes go. That is a short job for one login, and the same steps repeat across the rest of the team's records. None of it requires an import to fail first.

After the load

The load is not finished when the import reports success. The team needs to know where each login now lives and who can open it. Send a short note to the group leads listing the collections they have access to, and ask them to confirm within a set window. Anyone who does not confirm should be followed up with directly, since an unconfirmed group is the group most likely to discover a missing login during its first renewal call.

Then record the final counts beside the counts you wrote down before the load. If the numbers differ, you have a specific place to look. If they match, the next handoff starts from a known state, which is the main reason to do this work in writing at all.

Sources