Vendor Management Cadence: A Quarterly Review AI Prepares for You

When the COO of a 140-person services firm told me they had 87 active vendor contracts and reviewed exactly zero of them last year, I asked the obvious question. Her answer: "We don't have a person whose job it is to assemble the brief, and without the brief there's nothing to review."

Why do SMB vendors quietly renew without review?

Because the prep work hides the decision. Pulling 12 months of invoices, matching them to actual seat usage, checking the contract terms, scanning the security posture, noting the renewal window — that's three to five hours per vendor for a Finance or Ops person. Multiply by 80 vendors and the answer is always "we'll do it next quarter."

Definition: Vendor cadence — the recurring rhythm (weekly / monthly / quarterly) on which a specific class of vendor decision happens, with a fixed input format and a fixed decision-maker.

The result: contracts renew at last year's price plus the vendor's annual uplift, on terms nobody re-read, with seat counts nobody questioned. The Plan was "review before renewal." The Fact is "auto-renewed." The Gap is invisible because nobody assembled the brief.

What does an AI-prepared vendor brief actually contain?

Four sections. Each section answers one question a decision-maker would ask in the meeting.

Cost section — total spend last 12 months, trend vs prior year, contracted minimum vs actual usage, true cost-per-active-user.

Usage section — seats licensed vs seats active in last 30 days, feature usage breakdown if the vendor exposes it, list of accounts inactive >60 days.

Risk section — contract end date, auto-renew clause, notice period, data-residency / DPA status, last reported security incident (public), single-vendor dependency flags.

Renewal section — recommended action (renew / renegotiate / consolidate / drop), the one number that would change the answer, three follow-up questions for the vendor call.

Definition: Cost-per-active-user — annual contract value divided by users with login activity in the last 30 days. Almost always 1.5-3× higher than the cost-per-seat the vendor quotes.

The quarterly cadence (what actually happens)

Five steps. Roughly 90 minutes total decision time per quarter for the leadership reviewer.

1. Day -7: AI pipeline regenerates the brief for every active vendor — pulls invoices from accounting, seat-usage from each SaaS admin API, contract metadata from the contract repo, security posture from public sources.
2. Day -3: Ops lead reviews the briefs, flags any with missing data, escalates "renews within 60 days" to the top of the stack.
3. Day 0 (the review): Founder/COO walks the stack in priority order. Each vendor gets 3-6 minutes. Decision logged: renew / renegotiate / consolidate / drop.
4. Day +1: Ops sends the renegotiate / drop letters with the standard templates. Calendar reminders set for the renew vendors' next review.
5. Day +30: Spot-check on the renegotiate calls — did the requested change land in the contract?

Copy/paste vendor-review template

This goes in the brief that hits the COO's inbox three days before the quarterly review.

Vendor: [NAME]
Category: [CRM | SaaS-Ops | Infra | Marketing | Other]
Contract end: [DATE]     Notice period: [N days]     Auto-renew: [Y/N]

Cost:
- Annual contract value: [N]
- 12-month actual spend: [N]
- Trend vs prior year: [+/- %]
- Cost-per-active-user (30-day): [N]

Usage:
- Seats licensed / active: [N / N]
- Inactive >60 days: [N seats, listed below]
- Top feature usage: [TEXT]

Risk:
- DPA / data residency: [STATUS]
- Last public security incident: [DATE or "none in 24 months"]
- Single-vendor dependency: [Y/N — describe if Y]

Renewal recommendation: [RENEW | RENEGOTIATE | CONSOLIDATE | DROP]
Rationale (1 sentence): [TEXT]
The one number that would change the answer: [TEXT]
Questions for the vendor call: [3 BULLETS]

The "one number that would change the answer" line is the most important. It forces the AI prep work to surface the actual decision lever instead of a wall of statistics.

Tool tip (AIAdvisoryBoard.me): The quarterly vendor review is a textbook Plan → Fact → Gap workflow. Plan: renew at fair terms, consolidate the long tail, drop dead seats. Fact: which vendors are above contracted spend, which have <30% seat activity, which auto-renew in the next 60 days. Gap: the AI-prepared brief makes the deltas obvious in 90 seconds per vendor instead of three hours of pulling reports. The 7-day diagnostic on https://aiadvisoryboard.me/?lang=en shows the same pattern across every operational workflow in the company.

Good vs bad vendor brief

Bad: "We spent $84,000 with Vendor X last year, up 6%. Renewal is in March."

Good: "We spent $84,000 with Vendor X last year on a 50-seat contract — but only 31 seats have logged in within 30 days, so cost-per-active-user is $2,710 vs the $1,680 the vendor quotes. Notice period is 60 days, auto-renew kicks in March 14. Recommended: renegotiate to 35 seats and target $58k. The one number that would change the answer: if the 19 inactive users are dormant by policy (parental leave, contractors paused), the right move is reassignment, not downsize."

The second version contains a decision. The first version contains data.

Manager scan (2-minute digest example)

• Plan: quarterly review of all vendors >$5k/year; auto-renew windows flagged 60 days ahead
• Fact: 87 active vendors, 12 with renewals in the next 90 days, 6 above contracted spend, 11 with <30% seat activity
• Gap: 3 auto-renewed in the last quarter without review — root cause: brief missing for those 3 because of missing accounting data
• Plan: consolidate the 4 overlapping CRM/marketing tools by Q2
• Fact: consolidation analysis was started, paused at vendor-comparison step
• Gap: no owner — needs reassignment to Head of Marketing
• Plan: zero inactive seats older than 90 days
• Fact: 47 inactive seats across 9 tools
• Gap: automatic deactivation rule exists for 3 tools, manual for the other 6 — schedule a Friday sweep

Micro-case (what changes after 7-14 days)

A 110-person engineering services firm had 64 active vendors and a "vendor review" line item that had slipped four quarters in a row. After standing up the AI-prepared brief pipeline, the first quarterly review took 110 minutes and produced 9 renegotiations, 3 consolidations, and 4 drops. The 4 drops alone covered the cost of the pipeline. By the second quarter the cadence had compressed to 80 minutes because half the briefs read "no material change, recommend renew" and the COO trusted the prep enough to skim those. By the end of the second quarter the company had recovered roughly $48,000 in annualized spend and — more importantly — knew which vendors were structurally critical vs structurally drifting.

Note on this case: This example is illustrative — based on typical patterns we observe with companies of 30-500 employees, not a single named client. Specific numbers are rounded approximations of common ranges, not guarantees.

Tool tip (AIAdvisoryBoard.me): Vendor management is one of the workflows that breaks first when a company scales from 50 to 200 people — the founder used to know every contract, then suddenly there are 80 of them and nobody knows which ones matter. The Plan → Fact → Gap daily digest extends naturally to quarterly vendor review: the system already knows what was supposed to happen, what actually happened, and where the gap is. Walk the diagnostic on https://aiadvisoryboard.me/?lang=en and the vendor-review pattern is the same shape as every other operational cadence in the company.

FAQ

Should the AI make the renewal decision? No. The AI assembles cost, usage, risk, renewal sections and writes a recommendation. The human signs. Vendor renewals carry legal and relationship weight that doesn't belong in an automated decision. The value is killing the prep cost, not the decision.

What if vendor APIs don't expose seat usage? Most major SaaS tools do via admin API. For the ones that don't, two options: (a) export the user list monthly via the admin portal and feed that into the brief generator, or (b) use SSO / IdP login data as a proxy for activity. The proxy is good enough for "is this user using the tool" in 90% of cases.

Quarterly feels too slow — shouldn't this be monthly? For the brief generation, monthly is fine — it's automated. For the leadership review, quarterly is usually right; vendor terms don't move fast enough to justify a monthly leadership meeting unless you have an unusual concentration of short-term contracts.

Where does this fit in if we don't have a procurement function? Most 50-200-employee SMBs don't. The review owner is usually the COO or Head of Finance, depending on whether the brief is more about ops fit or pure spend. Pick one — splitting it across both creates the "nobody assembled the brief" problem from the start.

Conclusion

Vendor management dies at the brief, not the decision. The decision takes 3-6 minutes per vendor when the brief is right. The brief takes 3-5 hours per vendor when assembled by hand — which is why it doesn't happen.

Automate the brief. Keep the cadence quarterly. Keep the human signing.

If you want a system that surfaces the Plan → Fact → Gap automatically — every day, across the company, including the vendor stack — see how the 7-day diagnostic works at https://aiadvisoryboard.me/?lang=en.