A vendor portal is any outside website a virtual assistant must sign into to do the work, such as a supplier order system, a client billing portal, or a background check service. These sites often sit outside the company single sign-on and are easy to forget during onboarding. A short checklist prevents the assistant from discovering a missing login after a client has already asked for an update. The checklist should list the portal, the account type, the owner, and the date access was confirmed.
List the portals the role actually uses
Start from the work the assistant has already been assigned rather than from a general list of company tools. Ask which portals appear in the first week of the task calendar and which ones only appear at month end. A portal used once a quarter still needs an owner, but it can wait for a scheduled access window. The checklist should show the timing so nobody rushes an approval that is not needed yet.
Group the portals by the kind of access they grant, such as ordering, billing, records, or scheduling. Grouping reveals portals that hold the same client data and can be reviewed together. It also shows when one portal duplicates another and the role does not need both. A short grouping note is easier to maintain than a long alphabetical list.
Separate named accounts from shared logins
A named account is tied to the assistant and can be disabled when the role ends. A shared login is used by several people and cannot show who did what. The checklist should prefer a named account whenever the portal allows one. A shared login should be an exception with a written reason and a named custodian.
Record the email address and the display name that will appear inside each portal. A portal that shows a vague display name makes the assistant look like an unknown user to the client. The name should match the role and the team convention. The checklist should also note whether the portal sends notices to the assistant or to a shared inbox.
Mark the portals that require a password manager entry and a second factor. Those steps take time and should be scheduled before the first live task. A portal with no second factor should be flagged for the security owner rather than accepted quietly. The checklist should record the date the factor was enrolled.
Record the approval owner for each portal
Every portal needs one person who can approve access and one person who can revoke it. These may be the same person, but the checklist should not assume it. The approval owner should be named by role and by name so the assistant knows who to contact. A general support queue is not a substitute for an owner.
Write the evidence the owner needs before approving, such as a signed scope card or a client request. The evidence should be stored with the request so a later reviewer can see why access was granted. A verbal approval should be confirmed in writing the same day. The checklist should show the date and the channel of the approval.
Record the expected turnaround for each owner so the assistant can plan the first week. A portal that takes five business days should be requested before the start date. A portal that is immediate can be left for day one. The difference should be visible rather than learned through a missed deadline.
Set a secure credential handoff
Do not send a password in the same message as the portal link and the username. A single message turns one mistake into full access. Use the approved password manager or a sealed handoff method the security owner has accepted. The checklist should name the method and the person who confirms receipt.
Require the assistant to change a temporary password at first sign-in and to confirm that the change worked. A temporary credential that stays active is a standing risk. The confirmation should be recorded in the checklist rather than assumed. If the portal does not force a change, the owner should set the rule manually.
Keep the credential record out of shared documents and out of the task board. The checklist should show that access exists without exposing the secret. A note such as password manager entry confirmed is enough. The security owner should be able to audit the entry without reading a password.
Test access before the first live task
Ask the assistant to sign in and complete one harmless action in each portal before real work depends on it. The action might be viewing a past order, opening a draft, or checking a calendar. The test proves the account, the password, and the permission level together. A login that opens the portal but cannot reach the needed screen is still a failure.
Record the result of each test with the date and the person who ran it. A portal that fails should go back to the owner with the exact error message. The checklist should not mark access as complete until the action succeeds. A second test after the fix should be recorded too.
Give the assistant a clear route for an access problem on the first live task. The route should name the owner, the backup, and the expected response time. A problem reported to a general channel may sit while the client waits. The checklist should be linked from that route so the details are easy to find.
Review portal access at day thirty
At the end of the first month, compare the portals in use with the portals on the checklist. Remove access that the role never needed and add portals that were discovered late. The review should also confirm that every account is still named and every second factor is active. A checklist that is never compared with real use becomes a list of guesses.
Ask the assistant which portal caused the most delay and which instructions were unclear. The answer may show that an approval owner changed or that a portal now requires a new step. Update one entry and record the date. A specific repair is easier to test than a full rewrite.
Set the next review date before closing the current one. Access tends to drift when a client adds a tool or a role changes scope. A short quarterly check keeps the checklist close to the work. The owner should be named for the next review, not left to the assistant alone.
Key takeaways
Build the checklist from the work the assistant has been assigned. A general software inventory includes tools the role will never open. It also misses a client portal that appears only in one project. Start from the task calendar and add a portal only when a task requires it.
Prefer a named account for every portal that allows one. Record both the approval owner and the revocation owner beside the entry. A shared login should carry a written reason and a custodian. Test one real action before live work depends on the login.
Review the checklist at day thirty and compare it with actual portal use. Remove access the role never needed and add anything discovered late. Confirm that every second factor is still active. Set the next review date before closing the current one.
Continue building the workflow
Related support: onboarding operations support, contact us. The U.S. Cybersecurity and Infrastructure Security Agency - Require multifactor authentication recommends multifactor authentication for business accounts and administrative access.
Related reading
Build a confidentiality acknowledgement step for a virtual assistant
A practical acknowledgement step that explains the data the role touches, records questions, and keeps proof of the signed commitment easy to find.
Create a sick-day coverage plan for a virtual assistant role
A coverage plan that names the work that cannot wait, prepares a backup, and returns control with a short closing note after the absence.
Set a meeting notes standard a virtual assistant can follow
A note-taking standard that separates decisions from discussion, names the owner and due date, and makes the follow-up reviewable without replaying the call.
