By David,
Signature software looks like the easiest purchase in IT: a small per-user price, a monthly bill, done. Which is exactly why it deserves a real TCO pass - small per-user numbers are where cost structure hides best, because nobody audits a rounding error. Multiply by headcount and years, and the model you chose matters more than the sticker you compared.
Here are the six costs the sticker doesn't show. We won't quote competitor prices - they change and tier - but the structures are stable, and structures are what you can evaluate.
1. The seat gap
Per-seat licensing bills the users you license; value accrues on the users who carry a managed signature. The gap between those numbers - service accounts swept in by group-based licensing, seasonal staff, the acquisition you haven't onboarded, the pilot that stalled at 60% rollout - is pure spend. In most real tenants the gap is double-digit percent, invisible because the invoice never itemizes it.
The metered alternative prices the other side of the gap: Atlas Ink bills $1 per active user per month, always - a user counts only in a calendar month where a managed signature was actually applied to their mailbox. Pilot with fifty people, pay for fifty; the unenrolled thousand cost nothing until they're actually covered.
2. The commitment
The good per-seat rate is usually the annual rate, paid up front, sized to today's headcount - which converts a tool decision into a forecast. Downsize, divest, or decide the tool isn't working in month four, and months five through twelve are sunk. Metered billing has no contract-year: it follows usage monthly, billed via Stripe as prepaid units, scales down as freely as up, and stopping costs nothing but stopping.
3. The tier trap
SaaS signature pricing typically comes in editions, and the feature you actually need has a way of living one tier up - scheduling in the mid tier, analytics in the top one, with the whole campaign engine priced into seats that will never run a campaign. We've written before about buying a marketing platform to solve a hygiene problem; the TCO version is simpler: capability you don't exercise is spend, not value. Atlas Ink's answer is structural - one metered rate, no editions, and no campaign engine to pay for at all, by design.
4. The review that renews
A hosted signature platform holds your directory copy and - in stamping architectures - transits your outbound mail. Assessing that properly is real work: questionnaires, DPA negotiation, data-flow mapping, and an annual renewal of all of it, forever. Cost it at loaded security-engineer rates and the "small" tool grows a four-figure annual shadow. An in-subscription deployment shrinks the exercise to reviewing a resource group your controls already cover - the data-residency piece shows how short the flow list gets.
5. The infrastructure line (ours, not theirs)
Honest ledger, both directions: Atlas Ink's model includes a cost most SaaS pricing hides inside the subscription - you pay Azure directly for the small infrastructure the deployment consumes (App Service, database) at Azure's rates, on your invoice, visible in your cost views. It's modest, it's transparent, and it buys the ownership the previous section priced. Factor it in; we'd rather you model it than discover it.
6. The mail-path premium
Stamping architectures put a vendor service in your outbound delivery path. Most months that costs nothing. The tail risk isn't nothing: a stamping-service incident is a mail-flow incident, and mail-flow incidents are the most expensive hours in IT. However you weight that, weight it consciously - it's a cost dimension per-user pricing never mentions, and the architecture choice that avoids it entirely is on the table.
The model comparison, in one table
| Cost dimension | Per-seat SaaS | Metered, customer-owned |
|---|---|---|
| Unit | Licensed seat, use it or not | Active user with an applied signature |
| Commitment | Annual for the good rate | None - monthly usage |
| Editions | Feature-gated tiers | One rate |
| Security review | Vendor environment + data flows, annually | Your own resource group |
| Infrastructure | Hidden in subscription | Small, explicit, on your Azure bill |
| Mail path | In it (stamping architectures) | Untouched by default |
| Exit | Offboarding + sunk commitment | Stop pushing; meter stops |
Run your own numbers
The worksheet is four lines: (1) licensed seats × sticker × 12 for the subscription you're comparing; (2) actually covered users × $1 × 12; (3) add the small Azure infrastructure line to ours; (4) add your honest annual review-hours cost to theirs. Then read the pricing page for the standalone-vs-add-on details, or the buying guide if the shortlist is still forming.
Small numbers, multiplied by everyone, forever - that's signature TCO. Price the model, and the sticker takes care of itself.