AI for sales operations at a 30-person SaaS: from forecasting to follow-up

At a 30-person SaaS, sales ops is usually one or two people supporting 6-10 AEs, 2-3 SDRs, and a sales leader. The math is brutal: every hour of sales ops time touches roughly 12 people downstream, but the team is too small to invest in big infrastructure projects. AI changes the use equation more here than at almost any other size company.

What's actually broken at 30 people

Three patterns show up at most B2B SaaS companies in this size range:

1. The forecast is a vibes-based estimate. AEs forecast their deals in CRM but the data is incomplete. The sales leader does a "gut check" pass every week. The number that goes to the CFO has 30-40% error bands.

2. AE productivity varies wildly. Top AEs do 3-4x the volume of the bottom AEs, and nobody fully knows why. The sales leader has theories. The data doesn't yet prove which theory.

3. Lead routing is brittle. Marketing hands leads to sales via a janky Zapier workflow. Edge cases break it twice a month. Sales ops spends 5+ hours/week on repair work.

These three are what AI fixes first.

The 5-workflow stack

1. Meeting Notes → CRM ($4,995)

The single highest-use automation in sales ops at this stage. AI joins every customer call (or processes the recording), writes structured summary, extracts action items with owners and due dates, pushes them into HubSpot/Salesforce automatically.

For a 30-person SaaS: 8 AEs × 12 calls/week × 8 minutes of post-call admin = 12.8 hours/week recovered AE time. Plus, the deal data in CRM finally reflects what's actually happening on calls.

Knock-on: forecast accuracy lifts 5-10 points. CS handoffs work 3x better. Pipeline reviews stop being "tell me the latest" because the latest is already in the system.

2. CRM Hygiene Sprint ($2,495)

If your CRM has 6,000 contacts and 2,500 of them are stale, every other workflow is built on bad data. One-time, 7 days. Cleans dupes, fills missing fields, archives unreachables.

3. Custom Lead Routing Agent (~$6-8K)

Replaces the brittle Zapier workflow. AI reads incoming leads, applies routing logic (round-robin, ICP-fit-based, named-account), updates CRM, notifies the assigned AE. Handles the edge cases the Zapier flow couldn't.

Specifically valuable when you have: - Multiple sales segments (SMB, MM, enterprise) routing differently - Geographic distribution (not all leads are inbound from the US) - Named-account assignments that don't fit a simple round-robin

4. AE Coaching Insights (~$10-15K custom)

This is the "why are top AEs 3x better" question, finally answerable. AI reads transcripts across all AEs over 30+ days, identifies the language patterns, talk-time ratios, question depth, objection-handling moves that correlate with closed-won. Outputs a per-AE coaching brief.

The sales leader gets specific data instead of vibes. Coaching gets targeted. Bottom-quartile AEs typically lift 10-20% within 60 days of focused coaching on the AI-identified patterns.

5. Proposal Drafter ($3,995)

Already covered in adjacent posts. AEs draft 80% complete proposals in 30 minutes. Win rates lift on faster turnaround.

The honest math at 30 people

Investment: ~$25-30K total (productized + custom mix)

Year-one returns at typical $10M ARR: - AE time recovered: ~$300K of selling capacity - Forecast accuracy: hard to dollarize but board-relevant - Win rate lift from coaching + proposals: $300-600K of incremental closed-won - CRM data quality compound: another $100-200K of indirect value

Total: ~$700K-1.1M of impact on a $30K investment. 25-35x in year one.

What sales ops should NOT automate yet at 30 people

Three things often pitched but usually wrong at this stage:

1. AI sales engagement (auto-personalized cold email). This is an SDR tool, not a sales ops tool. And at 30 people the SDR motion is usually too immature to benefit from AI personalization at scale. Stay manual until SDR efficiency stabilizes.

2. Predictive deal scoring. Tempting at this stage but usually trained on too few closed deals to be reliable. Wait until you have 200+ closed-won and 200+ closed-lost in the same product.

3. Auto-generation of sales decks. Generic AI-drafted decks lose to thoughtful human-built ones every time at this stage. Custom decks per major opportunity stay human work.

What sales ops SHOULD do beyond the 5-workflow stack

Three things that aren't AI but compound with AI:

• Document the actual sales process in a way the AI workflows can reference. Without this, every AI build is starting from scratch.
• Define the ICP precisely so lead routing has clear logic. Vague ICP = unreliable routing regardless of AI quality.
• Set up a single source of truth for pricing. Proposal drafting works only as well as the pricing matrix it pulls from.

These three are the prerequisite work. AI accelerates everything once they're in place.

The 90-day rollout plan

Days 1-7: CRM Hygiene Sprint. Foundation work. Days 8-21: Meeting Notes → CRM build + rollout. Live for all AEs by Day 21. Days 22-45: Custom lead routing agent. Replaces Zapier. Days 46-75: Proposal Drafter rollout. Days 76-90: AE Coaching Insights initial deployment + first round of coaching.

Most sales ops leads at 30-person SaaS hit a 30x ROI by Day 90 if they execute in this order.

Where to start

A 30-minute audit specific to your current stack and sales motion outputs a ranked build plan with prices. Most sales ops leads leave with a 90-day plan that lifts forecast accuracy, win rates, and AE productivity inside one quarter — at a cost that's a small fraction of one AE's fully-loaded comp.