Why old outbound stopped working
Every B2B buyer's inbox now filters aggressively — Google's Postmaster tools flag any domain sending >5,000 cold emails/day. LinkedIn deprecated bulk messaging. The sequence-blast playbook (Outreach, Salesloft, Apollo at scale) has 0.3% reply rates today. The math doesn't work at any team size.
What replaced it
1. Sourcing from closed-won lookalikes
Instead of buying a 10,000-account TAM list, start with your 40 best customers. Extract firmographic (industry, size, geography), technographic (what's in their stack), and behavioral (how they buy) fingerprints. Find the 400 accounts that match. Reply rates 5-10× higher.
2. Intent triggers, not calendar batches
Don't send Monday at 9am. Send when the account just raised, hired a new VP of Eng, deprecated a competitor, or hit your pricing page for the third time. Reply rates on intent-triggered outbound run 15-25%.
3. Multi-channel from the start
Email → LinkedIn view → LinkedIn message → warm intro request → phone. AI orchestrates the sequence in the buyer's timezone, backs off on any engagement, and hands to a human the moment there's a reply.
4. Written in the founder's voice
"Template that looks personalized" converts worse than a plain-text three-sentence email from the founder. AI writes in your voice, not in "AI SDR" voice.
Benchmarks
- Reply rate on intent-triggered: 12-20%.
- Reply rate on lookalike, non-intent: 6-10%.
- Meeting rate on replies: 30-40%.
- Meetings per 1,000 accounts touched: 25-40.
How CoLive runs outbound
Mira mines closed-won lookalikes from your Salesforce, layers Clay + intent signals, drafts sequences in the founder's voice, and hands to a human the moment a reply lands. Every meeting drops into Salesforce with intent tags and closed-won-match score attached.