Lead Ops · Routing · Scoring · SLA

Hot leads to humans, under 30 seconds.

Speed-to-lead is the single strongest predictor of close rate. Most teams take 18 hours to reply to an inbound. We build a routing layer that enriches, scores, and lands the hot ones in the right AE's Slack DM before the lead has closed the browser tab.

✓ Enrich · score · route in < 30s✓ Round-robin · territory · expertise✓ SLA tracking · escalation · recovery✓ Dedupe + attribution built in
Best fit

Teams with a real product or internal tool to ship.

  • New products that need design, build, and launch in one motion.
  • Internal tools or CRMs replacing spreadsheet-and-Slack chaos.
  • Founders who want one team owning the outcome, not just tickets.
Not a fit

Probably not us if this is just hourly-staff shopping.

  • Cheapest-rate procurement.
  • Dev-only staff augmentation on greenfield product work.
  • Trying to polish a broken process without changing the workflow.
On the first call

You get the likely stack, shape, and next step.

  • What to build first and what can wait.
  • How we’d structure the stack and delivery.
  • A straight answer on fit, timeline, and budget shape.
What we build

A lead never falls through the cracks again.

The routing layer sits between your intake channels (forms, chat, ads, calls) and your sales team. It makes the first 30 seconds of every lead's life go well.

01

Real-time routing rules

Round-robin, territory, vertical expertise, account ownership, language — routed instantly. Out-of-office and PTO respected automatically. No "who owns this?" Slack threads.

02

AI-powered scoring

Every lead scored 0–100 on intent + fit against your ICP. Hot leads flagged in Slack with a suggested opening line. Warm leads get a nurture sequence. Cold leads get logged and forgotten without a human touching them.

03

Enrichment + dedupe

Every lead enriched with company size, revenue, role, tech stack (via Clearbit / Apollo / Clay) and deduped against existing contacts before it hits sales. No duplicate accounts, no repeat-intro awkwardness.

04

SLA tracking + recovery

Every lead has an SLA (e.g. 5 min for hot, 2 hours for warm). Breaches trigger escalation: re-assigned to the next AE, then to the manager, then to a Slack war-room. No dropped balls.

How we work

Live in 10–14 days. Tuned weekly after.

Routing is only useful when it's fast, trusted, and adapts. We build the v1, watch it for a month, and adjust the rules until the sales team stops second-guessing it.

01

Intake audit

Day 1–3. We map every way a lead currently enters — forms, chat, ads, inbound calls, events, referrals. Each one gets a route, a score threshold, and an SLA.

02

Scoring model

Day 3–7. We pull 500+ historical leads from your CRM labeled as won / lost / junk, and train a scoring model (AI prompt + fit rules) that matches your actual close patterns — not a generic template.

03

Routing logic

Week 2. Round-robin / territory / vertical / language rules encoded in n8n. Out-of-office, PTO, and capacity rules layered on top.

04

Slack + CRM wiring

Week 2. Hot leads → AE's Slack DM with context. Warm leads → nurture sequence in Klaviyo / HubSpot. Cold leads → CRM for archive. Everything logged.

05

Shadow + launch

Week 2. Runs in shadow mode for 3–5 days (logs decisions, doesn't act), then cut over to live. SLA dashboard goes up in Slack.

06

Tune monthly

Ongoing. Monthly review of what the scoring got right vs. what sales overturned. Rules updated, model retrained, SLAs recalibrated.

Our stack

Tools we ship with — every routing build.

Routing engine
n8nTemporalPostgresRedis queues
Scoring
ClaudeGPT-5Fine-tuned classifierStructured outputs
Enrichment
ClearbitApolloClayLinkedIn Sales NavFullContact
CRM · notify
HubSpotSalesforcePipedriveSlack DMsSMS fallback
SLA · observability
Slack SLA dashboardGrafanaPagerDutyEscalation rules
Pricing

Scoped per build. No per-lead fees.

We don't charge per lead routed. We build the system, deploy it on your infra, and hand you the source. Enrichment APIs (Clearbit / Apollo) billed at pass-through.

Core routing · 1 channel · round-robin from $7k
Routing + AI scoring · 2–3 channels from $15k
Full lead-ops layer · all channels · SLA from $28k
Monthly retainer · rules + model tuning from $1.8k/mo
Every build includes Slack-based SLA dashboard, escalation rules, dedupe against your existing CRM, and a monthly accuracy review. Enrichment API costs pass-through (~$0.10–$0.30 per lead depending on provider).
Questions worth asking

The short version.

Does this replace our CRM?

No — it feeds your CRM. We write enriched, scored, routed leads into HubSpot / Salesforce / Pipedrive as deals with all the context attached. The CRM stays the source of truth; routing is the layer in front of it.

What about duplicate leads?

Dedupe is part of the build. Every incoming lead is matched against your CRM by email, phone, and company-domain before anything else. Duplicates get logged + linked to the existing contact instead of creating a new deal. No more embarrassing re-intros.

How accurate is the scoring?

On our typical client, the scoring model agrees with the sales team's gut on ~85% of leads after the first tuning cycle, climbing to ~92% after 30 days of feedback. We show the eval dashboard on every monthly review. The remaining disagreements are exactly where we tune.

Can you route by language or region?

Yes — any routable attribute. Language (detected from lead content), region, territory, company size, vertical, product interest — whatever your team carves up sales by. We've routed leads across 11 languages and 8 territories for a single client.

What happens when an AE is out?

Out-of-office and PTO rules are baked in. Leads that would route to an away AE skip to the next round-robin slot, and the original AE gets a summary when they're back. If nobody's available (e.g. weekend), leads go into a warm-queue with an auto-reply and wait for Monday.

Can we see what the scoring agent is thinking?

Every lead in CRM has a note attached: score + top 3 reasons + confidence. AEs can see exactly why a lead was flagged hot or cold, and flag disagreements for the next tuning cycle. No black box.

Inbound is leaking?

Let's plug the holes.

30 minutes. We walk through your current intake → routing → response flow, identify the biggest leak, and quote the fix. Usually recoverable in the first month of the rebuild.