Total Cost of Ownership for Email Tools (Small Business)

The total cost of ownership of email tools for a small business is the number that decides whether the choice you make in month one still looks smart in month thirty. The mailbox invoice is the visible half. The other half - security add-ons, an assistant or client layer if you buy one, the admin time nobody bills for, and the migration you will do twice in three years - sits below the line until someone finally totals it.

This is a decision guide, not a warning list. It gives you the spreadsheet structure - five line items, one formula each, and the questions to ask so the numbers you drop in come from the vendor's own page rather than a sales deck. A worked example for a ten-person business at the end shows the arithmetic on one screen.

We build AI Emaily, so it shows up in the verdict section as one of the options being weighed. It is not the right answer for every small business - that is stated plainly, with the reader it is not right for named up front.

The short answer#

A small business email stack has five cost buckets, not one. Any TCO model that shows a single per-seat number is doing the vendor's marketing rather than your budget.

For a business of five to twenty-five people, the three-year total lands between roughly two and two-and-a-half times the mailbox provider's annual list price. The multiplier is not magic - it is what happens when you add a client or assistant layer, a couple of security add-ons that arrived as bundle upgrades or standalone tools, the admin time your founder or ops lead absorbs, and the migration cost that stays at zero until the quarter you switch providers.

The multiplier is for the tender-review meeting, so nobody quotes only the sticker. The five-line breakdown below is what stops the choice from surprising you in month four.

Criteria that actually matter for a small business#

Enterprise TCO frameworks weight identity plumbing, provisioning automation and audit tooling. A small business rarely runs those, so weighting them at all misses what is actually spending your money.

Five criteria carry the model for a small business. Weight them for your situation before you fill in numbers - moving a weight after the totals arrive is how a TCO becomes a justification for what the founder already picked.

• Mailbox provider is the anchor. Google Workspace, Microsoft 365, Zoho, Fastmail or a standalone IMAP host - this is the line that scales linearly with seats and the one your other costs inherit from. A cheaper provider that forces a paid add-on for every other line is not cheaper.
• Client or assistant layer is optional and often duplicated. If your provider bundles an AI assistant (Gemini for Workspace, Copilot for 365) and you also buy a third-party AI email tool, you are paying twice. Sometimes worth it, always worth naming.
• Security add-ons are the line small businesses miss most. Backup, retention, phishing defence beyond the provider default, DMARC monitoring, a shared inbox tool if you have support mail. Almost never zero, almost never in one budget line.
• Admin time is real money even when it is unpaid. A founder debugging DKIM on a Sunday is priced at their loaded rate, not zero. Ops leads answering "why can't I see this shared calendar" are on the same clock. Under-costing this line is what makes the provider's cheapest plan look attractive.
• Migration in and out. Migration in is done at signup, once. Migration out is planned at signup, done later. Both cost real hours; the second is the one every model forgets because it stays at zero for two years and then lands in a quarter you did not schedule.

The five-line TCO model#

Same shape whether you land on a suite, a standalone client, or a provider plus assistant combination. What changes is where the honest number comes from.

Worked example: ten-person business over three years#

Made-up unit costs, real formula. A ten-person services business. Two shortlists to compare - the arithmetic is what the choice turns on, not the sticker.

Loaded rates used below: founder time $110/hour, ops admin time $80/hour, engineering or contractor $130/hour. Adjust to your own compensation before running the model.

Two things to notice. First, the multiplier for Path B is higher because line 2 is a whole separate invoice - but line 4 only rises by one hour a month, not four. If the standalone assistant genuinely saves the founder or ops lead five hours a week, the extra $2,880 on line 4 comes back on the founder's time in about six weeks. The comparison is not sticker versus sticker; it is total versus total including the hours you buy back.

Second, the security line is roughly the same shape either way. Suite-bundled backup and phishing are rare; almost every small business buys these separately regardless of which client sits on top. If you are pricing a suite and finding line 3 near zero, you have almost certainly not looked for it yet.

Red flags that break a small business TCO model#

A TCO stops being predictable when one of the inputs is not knowable at signing. For a small business specifically, watch for these before the invoice, not after.

• Storage caps in the base tier. A provider quotes at the entry tier while your team already sends attachments that will blow the cap in year two. Then line 1 upgrades mid-contract and every seat moves with it.
• Bundle upgrades that raise all seats. Moving one seat to a higher tier for one feature (SSO, retention, shared drive quota) usually means moving every seat - line 1 rises by the delta times ten, not times one.
• Free migration in, paid migration out. Import tools bundled at signup, export routed through a support queue or a paid professional-services engagement. That is line 5's second half growing at the moment you have least leverage.
• Assistant metering that clips in a heavy month. If the AI layer meters messages, drafts, credits or tokens, model your heaviest month across three years - not an average. A quiet-month average is a discount that does not exist.
• Security add-on gaps that only surface after a phishing test. A provider default that stops obvious spam but not a targeted display-name spoof is a line 3 you have not budgeted for yet. Better to add it deliberately than reactively.
• Admin absorbed by a founder as an unbilled favour. Line 4 exists whether or not anyone invoices it. Founders who "don't count their time" are the same ones who eventually hire an ops manager because email admin ate a quarter.

What we would pick, honestly#

We build AI Emaily, so this is the vendor's answer sheet rather than a review. The point is that every line above is checkable against our pricing page and our behaviour in a trial - and we are honestly not the right answer for every small business.

Where we are the right pick on TCO. A five- to twenty-five-person business across Gmail, Outlook or standard IMAP that wants approve-before-send by default, an audit log, undo, no training on user mail, and a per-seat bill that does not surprise in a heavy month. AI Emaily is per seat with a monthly AI credit allowance and a 7-day free trial (card required, $0 if cancelled before day 7) - there is no permanent free tier, and no per-seat trial for teams. Bring-your-own-model is supported for AI usage, which stops line 1's overage in its tracks. Current numbers on /pricing.

Where we cost more than the sticker page implies. Line 4 is per-user today rather than org-wide - single sign-on, automated provisioning and an org-wide autonomy policy are on the roadmap rather than shipped. For a team of five to twenty-five that is fine; if you already administer identity in Okta or Entra, we do not yet plug in the way a suite-native tool does. Line 5's out-half is a rebuild of rules and Context by hand today because no other client imports our JSON export - honest, real, and a real reason to prefer a suite-native assistant if identity plumbing is your top-weighted criterion.

Which reader should stop reading and buy the other thing. If your business is small but rolling out inside a mature identity platform, or if you accept the AI assistant your suite already bundles as good enough, then adding a second assistant tool is line 2 you do not need. Google Workspace with Gemini, or Microsoft 365 with Copilot, will win your TCO comparison on lines 2 and 4 because you are already paying to administer the platform. Check the current per-seat number on each vendor's own pricing page - workspace.google.com/pricing and microsoft.com/en-us/microsoft-365/business - and price the bundled AI at the tier delta rather than at zero.

Where we are the better answer on TCO. Suite-mixed small businesses (some seats on Gmail, some on Outlook, a couple on IMAP for a client project), professional services firms whose partners will not accept an assistant that auto-sends without approval, and founder-led teams who want the multiplier down because their own hours are the biggest variable in the model. Approve-before-send, per-seat pricing, no training on mail, and a self-serve export mean line 1 does not surprise you and line 5's exit half stays short. We build AI Emaily; both halves of that comparison are ours to defend.

Handing the model to finance#

One page, five lines, the arithmetic showing. Keep the assumptions column beside the numbers - seat count, loaded hourly rates, heavy-month scenario, migration hours - because that is what finance will interrogate first.

Then read the total against the sticker. If the multiplier lands under 1.5 for a small business, a line is missing - usually 3 (security add-ons) or 4 (admin time absorbed by the founder). If it lands above 2.5, one specific line is doing the work, and naming it - metered overage in a heavy month, a bundled tier upgrade to unlock one add-on, a migration-out estimate that assumes a rebuild rather than an import - is what makes the TCO defensible.

The number is not the deliverable. The formula column is, because it is what stops the decision being reopened in six months by someone who was not there when you picked.