How a 180-person B2B SaaS company automated lead routing, follow-ups, and pipeline management - increasing sales velocity by 89% whilst reducing sales ops workload by 42 hours weekly.
TL;DR
Company: DataPulse (analytics platform for e-commerce, Series B, 180 employees, £18M ARR)
Challenge: Sales team overwhelmed by manual pipeline management, leads slipping through cracks, inconsistent follow-up
Solution: Automated lead routing, intelligent follow-up sequences, pipeline health monitoring, and deal progression tracking
DataPulse's sales team (12 AEs, 2 SDRs, 1 Sales Ops Manager) was managing 400+ active opportunities manually. Critical processes broke down as volume increased.
Manual process consumed:
| Activity | Hours/Week | Pain Points |
|---|---|---|
| Lead routing and assignment | 14 hours | Delays, uneven distribution, territory conflicts |
| Follow-up email scheduling | 18 hours | Inconsistent timing, forgotten leads, generic messaging |
| Pipeline hygiene (updating stages) | 22 hours | Outdated data, inaccurate forecasts, missed milestones |
| Deal health monitoring | 12 hours | At-risk deals identified too late, no early warning system |
| Forecast reporting | 8 hours | Manual spreadsheet compilation, error-prone, time-consuming |
| Activity logging (calls, emails, meetings) | 16 hours | Incomplete records, CRM data gaps, difficulty tracking engagement |
| Total | 90 hours | Manual processes slowing entire revenue engine |
"We were haemorrhaging opportunities. Leads would sit unassigned for 6 hours. Follow-ups happened whenever reps remembered. Deals would stall at demo stage for weeks with no intervention. Our forecast accuracy was 62% - essentially worthless." - Marcus Chen, VP Sales, DataPulse (interviewed September 2024)
Additional problems:
DataPulse automated six critical pipeline workflows:
Workflow:
When new lead enters system (form, demo request, sales qualification):
Step 1: AI evaluates lead attributes
- Company size, industry, geography
- Intent signals (pages viewed, content downloaded)
- Product fit score (needs vs capabilities)
- Budget indicators
Step 2: Route to optimal rep
- Territory match (geography, industry vertical)
- Current pipeline load (balanced distribution)
- Rep expertise (product specialization)
- Availability (PTO, capacity constraints)
Step 3: Immediate notification
- Slack alert to assigned rep within 30 seconds
- Email with lead context and recommended approach
- CRM task created with due date (respond within 2 hours)
Step 4: Escalation if unactioned
- If no contact attempt within 2 hours: reminder to rep
- If no contact within 4 hours: escalate to sales manager
- If no contact within 24 hours: reassign to available rep
Time saved: 14 hours weekly
Before vs After:
| Metric | Manual | Automated | Change |
|---|---|---|---|
| Time to first contact | 6.2 hours avg | 47 minutes avg | -87% |
| Lead response SLA compliance | 48% | 94% | +96% |
| Rep workload balance | Uneven (some 2× others) | Within 15% variance | Balanced |
| Territory conflicts | 18% of assignments | <2% | -89% |
Workflow:
When opportunity enters pipeline stage:
Step 1: AI selects appropriate sequence
- Stage-specific templates (discovery, demo, proposal, negotiation)
- Industry-customized messaging
- Personalization using lead data
Step 2: Schedule sequence based on engagement
- If prospect opens email: send next touch in 2 days
- If no open after 3 days: send alternative angle
- If clicked link: prioritize for immediate call
- If replied: pause sequence, notify rep
Step 3: Multi-channel cadence
- Day 1: Personalized email
- Day 3: LinkedIn connection/message
- Day 5: Phone call (auto-logged in CRM)
- Day 7: Video message email
- Day 10: Final value-add email
Step 4: Adaptive timing
- AI learns optimal send times per prospect (time zone, role, engagement patterns)
- Adjusts frequency based on engagement signals
Time saved: 18 hours weekly
Before vs After:
| Metric | Manual | Automated | Change |
|---|---|---|---|
| Follow-up consistency | 62% of leads received planned touches | 96% | +55% |
| Average touches before response | 8.4 | 5.2 | -38% |
| Email open rates | 24% | 38% | +58% |
| Response rates | 12% | 21% | +75% |
Workflow:
Continuous monitoring of all active opportunities:
Step 1: AI assesses deal health signals
- Days in current stage vs historical average
- Engagement level (emails, calls, meetings)
- Stakeholder mapping completeness
- Next step clarity and timeline
Step 2: Flag at-risk deals
- Stalled (no activity 7+ days): Yellow alert
- High risk (missing key milestones): Orange alert
- Critical (likely to close-lost): Red alert
Step 3: Automated intervention
- Yellow: Suggested action sent to rep ("Schedule follow-up call")
- Orange: Manager notified, coaching recommended
- Red: Automated executive outreach sequence initiated
Step 4: Weekly pipeline review automation
- AI generates health report for each rep
- Flags deals needing attention in pipeline review meeting
- Recommends actions (discount approval, executive engagement, etc.)
Time saved: 12 hours weekly
Before vs After:
| Metric | Manual | Automated | Change |
|---|---|---|---|
| Average deal cycle (lead to close) | 38 days | 20 days | -47% |
| Deals identified as at-risk | 23% (too late) | 71% (early warning) | +209% |
| At-risk deal save rate | 14% | 34% | +143% |
| Pipeline stage accuracy | 68% | 91% | +34% |
Workflow:
Automatic capture of all sales activities:
Step 1: Email integration
- All emails to/from prospects auto-logged in CRM
- Sentiment analysis flags concerns or urgency
- Key topics extracted (pricing, timeline, competitors)
Step 2: Calendar sync
- Meetings automatically logged with attendees
- AI generates meeting summary from transcript
- Action items extracted and created as tasks
Step 3: Call logging
- Phone calls auto-logged (via phone system integration)
- Duration, outcome recorded
- AI transcribes and summarizes
Step 4: Engagement scoring
- All activities contribute to engagement score
- Score decay over time if no recent activity
- Low engagement triggers intervention
Time saved: 16 hours weekly
Before vs After:
| Metric | Manual | Automated | Change |
|---|---|---|---|
| Activity logging compliance | 58% | 97% | +67% |
| CRM data completeness | 64% | 93% | +45% |
| Time spent on data entry | 16 hours/week | 2 hours/week | -88% |
Workflow:
Automated forecast generation:
Step 1: AI analyzes historical patterns
- Win rates by stage, rep, industry, deal size
- Seasonal trends
- Conversion rates across pipeline stages
Step 2: Weighted pipeline calculation
- Each deal weighted by: stage probability × AI confidence score × rep track record
- Identifies "sandbagging" (deals more likely to close than rep indicates)
- Flags "over-optimism" (deals less likely than rep forecasts)
Step 3: Real-time forecast dashboard
- Updated hourly based on pipeline changes
- Scenario modeling (best case, likely, worst case)
- Trend analysis (forecast vs actual over time)
Step 4: Automated reporting
- Weekly forecast email to leadership
- Monthly forecast accuracy review
- Rep-specific forecast coaching recommendations
Time saved: 8 hours weekly
Before vs After:
| Metric | Manual | Automated | Change |
|---|---|---|---|
| Forecast accuracy (±10%) | 62% | 87% | +40% |
| Time to generate forecast | 6 hours | 15 minutes | -96% |
| Forecast update frequency | Weekly | Real-time | ∞ |
| Leadership confidence in forecast | 5.2/10 | 8.6/10 | +65% |
Week 1: Process audit
Week 2-3: Build and integrate
Week 4: Test and launch
Tools used:
Investment:
| Metric | Before | After | Change |
|---|---|---|---|
| Average deal cycle | 38 days | 20 days | -47% |
| Sales velocity (opportunities per month) | 84 | 159 | +89% |
| Win rate | 18% | 26% | +44% |
| Sales ops workload | 90 hours/week | 48 hours/week | -47% |
| Forecast accuracy | 62% | 87% | +40% |
| Pipeline health visibility | Reactive | Proactive | N/A |
| Lead response time | 6.2 hours | 47 minutes | -87% |
Financial impact:
"The automation fundamentally changed how we sell. Reps focus on conversations, not admin. Leads get instant response. Deals don't stall silently. Our forecast went from 'educated guess' to 'reliable projection.' The velocity increase let us hit our annual target in 8 months." - Marcus Chen, VP Sales
What worked well:
Challenges faced:
Advice for similar implementations:
Ready to automate your sales pipeline? Athenic connects to Salesforce, HubSpot, and Pipedrive to automate lead routing, follow-ups, pipeline health monitoring, and forecasting. Explore sales automation →
Related reading: