A contract approval workflow for a team of 10 to 30

Why 10 to 30 people is exactly where this gets painful

Below about 10 people, one person (usually the owner) reviews and signs everything, and that works fine because there's no real coordination problem. Past 30 or so, most businesses have hired someone specifically for operations or finance who owns this properly. The 10 to 30 range is the uncomfortable middle: enough people are generating contracts, sales reps sending client agreements, operations staff signing vendor terms, that the owner can't personally review everything anymore, but there's no dedicated person whose job this is yet. Without a defined workflow, what actually happens is either everything still funnels to one overloaded person, or contracts start going out with no review at all, both of which cause real problems eventually. The overloaded-owner version shows up as deals sitting for days waiting on a signature that was never really the bottleneck, it was just the queue, and the owner usually doesn't realize how long things are actually sitting until a client or vendor complains about the delay directly. The no-review version shows up later and worse, a contract with terms nobody with authority actually looked at, discovered only when something in it causes a problem.

Roles and dollar limits

Define three tiers based on financial exposure and risk, not department:

• Self-approve: standard contracts under a set dollar threshold (commonly $1,000 to $5,000 depending on your business, pick a number that matches your actual risk tolerance) using a pre-approved template with no changes to the core terms. The person sending it can send it, no second signature required.
• One-step approval: anything above that threshold, or any non-standard terms even under it (a custom liability clause, a payment term that's not your usual 30 days), needs sign-off from one named approver, usually a department lead or the owner, before it goes out.
• Two-step approval: anything involving real legal or financial exposure, a lease, a multi-year commitment, anything over a higher threshold like $25,000, or anything your standard template doesn't cover, needs both a department lead and the owner or a lawyer before it goes out. Keep this tier genuinely rare. If most contracts are landing here, your thresholds are set too low.

Write these thresholds down somewhere everyone can see them, not just in the owner's head. My signing authority guide covers the related question of who's legally allowed to actually sign once something's approved, which is a distinct question from who approves it internally.

The actual steps, end to end

1. Draft from an approved template whenever possible. A contract built from scratch every time is both slower and more likely to contain a mistake.
2. Self-check against the dollar and risk tiers above. The person drafting it decides, using the written thresholds, whether this needs zero, one, or two approvals.
3. Route for approval if required, to a named person, with a deadline (48 hours is reasonable for most contracts, faster for anything time-sensitive, with an explicit way to flag genuine urgency).
4. Approver reviews and either approves, sends back with changes, or escalates to the next tier if they spot something outside their own authority to decide alone. A contract review checklist keeps that read quick and consistent.
5. Send for signature once approved, through whatever e-signature tool you use.
6. File the signed copy and log the approval trail (who approved, when, at what tier) alongside it, which matters later if anyone ever needs to reconstruct why a specific deal was approved.

A worked example: a 22-person agency's actual thresholds

Picture an agency with 22 staff, a handful of account managers who send client agreements and a small ops team handling vendor contracts. Their self-approve threshold sits at $3,000, which covers most standard retainer renewals using an unchanged template. Anything from $3,000 to $25,000, the bulk of their new client contracts, needs one approval from the agency's operations lead, turned around within 48 hours as a standing rule. Anything over $25,000, or any contract with custom liability or indemnification language, goes to both the operations lead and the owner. In practice, roughly 60 percent of their contracts never leave the self-approve tier, 35 percent hit the one-step tier, and the two-step tier gets used maybe once a month. That ratio, most things self-approved, a meaningful chunk at one step, the top tier genuinely rare, is a reasonable target for a business this size. If your own numbers look inverted, with most contracts landing in the top tier, that's usually a sign the thresholds themselves need resetting, not that your business is unusually risky. The same tiered limits work for spending, as in this purchase order approval process.

Tools that handle this without extra software

At this team size, you rarely need a dedicated contract approval platform. A shared project management tool you already use, Asana, Trello, or similar, handles the routing and visibility step fine: one card or task per contract, moving through stages (Draft, Pending Approval, Approved, Sent, Signed), with the approver tagged directly on the card. Most e-signature tools also support a built-in approval step before a document goes out for signature, which keeps the whole thing, draft to approval to signature, inside one tool instead of three. Writing the steps up as a short contract SOP helps new staff follow the same route. The deciding factor isn't which specific tool, it's whether everyone involved can see, at a glance, where any given contract currently sits without asking someone directly.

Where this workflow actually breaks down

Three patterns repeat across businesses at this size. First, thresholds get set once and never revisited, so a $1,000 self-approve limit set three years ago is now too low for current deal sizes, and people start routing around the process because it's slower than it needs to be. Second, "approval" happens verbally, a quick hallway conversation or a Slack thumbs-up, with no record, which is fine until someone disputes later whether a specific term was actually approved or just assumed. Third, the two-step tier becomes the default for anything slightly unusual because nobody wants to be the person who approved something that later went wrong, which defeats the entire point of having tiers at all. The fix for all three is the same: revisit the thresholds roughly once a year, require every approval to leave a written trail even if it's just a one-line approval inside your project tool, and genuinely trust the one-step tier to do its job for most non-standard-but-not-risky contracts.

How I'd approach this

When I look at an approval workflow for a business this size, I'm mainly checking whether the thresholds match actual risk rather than nervousness, and whether the process adds real friction to the deals that need it while staying out of the way of the ones that don't. A workflow where 80 percent of contracts need two signatures is usually solving an anxiety problem, not a risk problem. Where this becomes a genuine software conversation is when the approval routing needs to connect automatically to other systems, a contract that's approved should update a CRM deal stage or trigger a specific onboarding task without someone manually doing it, rather than living entirely inside a shared project board. That's a real integration, and it's where a built process starts to save meaningful time rather than just looking tidier on paper.

I'd also push back gently on any version of this workflow built primarily to protect the owner from ever being surprised, rather than to actually catch risk. Those two goals sound similar but produce very different thresholds. A process built around "nothing goes out without me seeing it" doesn't scale past the owner's own bandwidth, by definition, and it's usually the real reason a 10 to 30 person business still has every contract bottlenecked through one inbox months after hiring people specifically meant to take that load off. If delegation is the actual goal, the thresholds have to be generous enough to mean something, not just generous enough to look reasonable on paper.