B2B SaaS
ICP: Company size 50–500, revenue above €5M, tech stack includes a tool you integrate with, contact is VP-level or above.
→ Hot-lead call conversion rate typically 3–5× higher than cold outbound.
Use case · CRM + Research + Call agents
AI lead qualification — response time under 5 minutes
An AI lead-qualification agent handles the full inbound cycle without human delay: the Research agent enriches every new form submission with company data, the CRM agent scores the lead against your ICP within seconds, and the Call agent dials hot leads (score above 80) inside 5 minutes. Cold or borderline leads go into nurture sequences instead of rotting in an inbox. Studies consistently show that replying to a hot lead within 5 minutes makes it 9× more likely to convert than replying at 30 minutes — most SMBs average 47 minutes.
Why speed-to-lead matters more than anything
The research on inbound lead response time is unambiguous. Most teams still miss the window.
9×
higher conversion
Leads contacted within 5 minutes convert 9× more often than leads contacted after 30 minutes, per published sales research.
47 min
average response
Average time-to-first-contact across surveyed European SMBs in 2025. Only 7% respond within 5 minutes.
60 %
leads qualified wrong
Share of manually scored leads misclassified as qualified / unqualified in SMBs without structured ICP data.
€1,200
cost per misrouted lead
Estimated revenue impact of a hot lead not contacted within the first hour for a 10-person B2B services firm.
How the agent handles a new lead — 5 steps
Every form submission, webinar signup or CRM import flows through the same loop.
- 01
Receive
A lead arrives — form submission, LinkedIn Lead Gen form, webinar signup, CRM import, business-card scan. The Automation agent fires the qualification flow.
- 02
Enrich
The Research agent fetches company data from public sources: size, revenue band, industry, tech stack, recent news. Fills 15–20 CRM fields in seconds.
- 03
Score
The CRM agent scores the enriched lead against your ICP: hard gates (region, size, industry) + soft signals (tech stack match, hiring intent, buying triggers). Output: 0–100 score with reasons.
- 04
Route
Score > 80 → Call agent dials within 5 minutes. Score 40–80 → Mail agent starts a nurture sequence. Score < 40 → archived with reason logged, not deleted.
- 05
Contact
For hot leads, Call agent introduces itself, qualifies budget and timeline in conversation, books a meeting against your calendar if interested, logs everything to the CRM.
What the agent actually scores against
Your ICP is encoded as a mix of hard gates and soft signals. The agent scores each lead transparently — every score comes with the reasoning.
Hard gates (disqualify if any fail)
- •Company size range (e.g. 5–500 employees)
- •Country or region (matches your sales territory)
- •Industry (SIC / NACE code whitelist)
- •No conflict with existing customer list
- •Minimum revenue threshold (if known)
Soft signals (add to score)
- •Tech stack match (uses tools you integrate with)
- •Recent hiring signals (new roles in relevant function)
- •News / funding events in last 90 days
- •Decision-maker seniority of the contact
- •Engagement history with your brand
Negative signals (reduce score)
- •Competitor domain in email address
- •Prior disqualification in CRM history
- •Generic mail address (gmail, yahoo) for B2B ICP
- •Incomplete form data with no enrichable identity
- •Flagged as low-intent in marketing-automation data
Manual lead qualification vs AI agent
Same inbound volume. Different response times, different conversion outcomes.
| Dimension | Manual qualification | AI agent qualification |
|---|---|---|
| Time to enrichment | 30–90 min per lead (if done) | Under 30 seconds |
| Scoring consistency | Varies per SDR, shifts over time | Deterministic scoring model, logged reasoning |
| First-call latency | Hours to days | Under 5 minutes for score > 80 |
| After-hours leads | Wait until next business day | 24/7 response; scheduled dial at business open |
| CRM data quality | Gaps, typos, inconsistent values | Structured, verified, filled from public sources |
| Nurture sequence start | Manual or delayed | Auto-starts for score 40–80 the same minute |
| Per-lead cost | €12–€25 (SDR time at SMB rates) | Included in €749 per-user subscription |
By industry — what hot leads look like
ICP definition varies by industry. The scoring model is the same; the criteria are configured per customer.
Professional services
ICP: Austrian/German SMB, 10–200 employees, matches the service line (e.g. tax, legal, consulting), contact is owner/CEO.
→ Response within 5 minutes from an AI-voiced call; many bookings same-day.
Real estate
ICP: Buyer in price range, location in agent territory, pre-qualified financing, timeline 0–6 months.
→ Viewing bookings 24/7 without after-hours staffing.
Local trades
ICP: Homeowner in service area, project value above minimum threshold, urgency level, prior quotes from competitors.
→ Emergency calls triaged and dispatched immediately.
Healthcare practices
ICP: Insurance coverage match, appointment type needed, distance from clinic, first-time vs returning patient.
→ Booking friction removed; front-desk time freed.
What does lead qualification cost?
Lead qualification is not billed separately. It is a combination of three agents working together.
Research agent, CRM agent and Call agent are all included in the €749 per-user-per-month DivineMind.AI subscription. There is no "lead qualification add-on" and no per-lead fee.
The only variable cost is outbound voice minutes — the Call agent dials hot leads, and voice minutes are sold as credit packs from €499 (2,000 minutes) to €4,999 (40,000+ minutes). An SMB typically consumes 2,000–8,000 minutes per month across all voice use cases, not just lead qualification.
Compared to dedicated SDR software (Apollo, Cognism, Outreach) layered with a dialer (Aircall, Orum), the line-item cost is often lower and the outcome is tighter — because the CRM agent and Call agent share the same data graph instead of being stitched together via iPaaS.
Frequently asked questions
Is my ICP configuration portable to a different vendor later?+
Yes. The ICP is stored as structured data (hard gates + weighted soft signals) in your tenant. Exportable as JSON at any time. Not locked to DivineMind.AI.
What happens if the agent mis-scores a lead?+
Every score is logged with its reasoning. Sales reps can override the score with one click, and the override trains the scoring model for your tenant (without touching other customers' data or the base model).
Does the Call agent really answer within 5 minutes?+
For score-above-80 leads during configured business hours: yes, typically within 1–3 minutes. Outside business hours, you can configure the agent to either (a) call immediately anyway, (b) schedule for business open, or (c) send an immediate email and call at open.
Can the agent handle lead qualification in multiple languages?+
Yes. The Research agent enriches data regardless of language. The CRM agent scores against your ICP in any language. The Call agent conducts the qualifying call in the lead's language — German, English, French, Italian, Spanish, Greek, plus 25 more.
How does it handle privacy for enrichment?+
Enrichment uses only public-domain data sources (company registries, LinkedIn public profiles, company websites). No scraping of private profiles. No purchase of data from gray-market vendors. The sources are logged per lead and exportable for compliance.
What if a prospect does not want to speak with an AI?+
The Call agent identifies itself as an AI receptionist at the start of every call, per EU AI Act transparency rules. If the prospect prefers a human, the agent offers a same-day callback from a named sales rep and sends a context summary to that rep ahead of the call.
See it qualify your own leads
Bring 10 anonymized past leads to the demo. We set up your ICP scoring model live and show what the agent would have done with each.