Why most small teams do MFA badly
Most small teams enable MFA the same way they buy software: in a rush, on the wrong accounts first, with no plan for the day someone loses a phone.
Doing it well is not harder. It just needs an hour of inventory, one method picked on purpose and a place to keep recovery codes that is not a sticky note.
The goal of this workflow is simple: at the end of one weekend, every account that could ruin your week is protected, you have a backup admin and you can prove it.
The full flow at a glance
- 1InventoryList every account that touches money, identity, code or customer data.
- 2Rank by blast radiusWhat breaks if this one account is compromised?
- 3Pick a methodTOTP, security key or platform passkey — choose one default.
- 4Enable top-downDomain, email and admin first. Everything else second.
- 5Test recoveryPretend you lost the phone. Can you get back in?
Step 1: Inventory before you enable anything
Open a doc. List every account, not every tool — same vendor often means several accounts (admin, billing, personal).
A useful frame is the four categories of damage:
- Money. Bank, payment processor, accounting, payroll.
- Identity. Domain registrar, email, password manager, SSO provider.
- Code and data. Source control, cloud, customer database, file storage.
- Customer-facing. CRM, help desk, marketing, e-signature, anything your buyers see.
If an account does not fit any of these, it is probably not urgent. Park it.
Step 2: Rank by blast radius
Not every account deserves the same care. Score each one quickly:
- Catastrophic — if compromised, the business stops. Domain registrar, primary email, password manager, bank.
- Severe — if compromised, recovery is days of work. Cloud account, source control, payment processor, customer database.
- Annoying — if compromised, you fix it in an afternoon. Marketing tools, analytics, random SaaS trials.
Catastrophic accounts get protected this weekend. Severe accounts get protected this month. Annoying accounts get cleaned up during your next quarterly SaaS audit.
Step 3: Pick one MFA method and stick to it
Picking three methods at once is how you confuse the team. Pick one default and accept the trade-offs.
- Authenticator apps (TOTP). Free, works almost everywhere, easy to roll out. Good default for most small teams.
- Security keys. Strongest protection against phishing. Worth it for catastrophic accounts and tech-savvy teams.
- Passkeys. The smoothest modern experience when both sides support it. Coverage is growing fast.
- SMS. Better than nothing. Worse than everything above. Avoid as your primary method.
Most small teams should standardize on TOTP for everything, then layer security keys or passkeys on the catastrophic tier. The password manager you picked already supports both — that is part of why it matters.
Stores TOTP codes alongside passwords, supports passkeys and makes shared-account MFA workable for small teams.
Open 1PasswordA dedicated authenticator app with multi-device backup, useful when you want TOTP separate from the password manager.
Open AuthyHardware security keys for phishing-resistant MFA on the highest-value accounts — domain, email, banking.
Open YubicoUse these as authenticator or recovery options. The password manager is usually the calmest place to store TOTP seeds.
Step 4: Enable MFA top-down
Order matters. Protect identity first, because losing identity means losing recovery for everything else.
- Domain registrar. If someone steals this, they own your business name.
- Primary email. The reset link target for almost everything else.
- Password manager. The vault you are about to fill with secrets.
- Payment and banking. Stripe, PayPal, Wise, bank, accounting.
- Cloud and code. AWS, Google Cloud, GitHub, source control.
- Customer-facing tools. CRM, help desk, marketing platform.
- Everything else during the next sweep.
For each one, the rollout looks the same:
- 1Open security settingsFind the MFA or two-factor section.
- 2Enable your chosen methodTOTP first, security key or passkey when supported.
- 3Save recovery codesStore them in the password manager, in a separate note from the login.
- 4Sign out and back inConfirm the new flow works on the device you actually use.
- 5Mark the row doneUpdate the inventory doc. No row should be ambiguous.
Step 5: Test recovery and name a backup admin
This is the step most teams skip. Two things must be true before you call MFA done:
- You can recover access if your phone dies, using only the recovery codes or backup method you saved.
- A second human can recover access if you are unreachable.
If either is fuzzy, you have an outage waiting to happen.
For shared accounts, decide explicitly who is the backup admin, write it down and rotate every six months. For solo founders, give a trusted person sealed recovery instructions — a will-style envelope is unglamorous but works.
What usually goes wrong
- Recovery codes in screenshots. Storing recovery codes as screenshots in cloud photos defeats most of the protection. Use the password manager.
- One person owns everything. If only one teammate can recover anything, MFA is a single point of failure with extra steps.
- SMS as the primary factor. SIM-swap attacks are not rare. Use SMS only when nothing else is supported.
- Shared logins behind personal MFA. Move shared accounts to the password manager’s shared vaults so MFA travels with the role, not the person.
Next steps
- Run the minimum security baseline checklist right after MFA is in place.
- If you have not picked a vault yet, see the best password managers for small teams.
- Use the SaaS due-diligence checklist before signing the next annual deal.