01 · Now · current role
Demand generation, led for a living.
Head of Demand Generation
2019 → today
a B2B software-services company, unnamed on purpose
{{ hn1 }}
years in the role
{{ hn2 }}
team, hired and trained by me
~{{ hn3 }}
events worked a year
{{ hn4 }}
meetings booked 2024 to 2026 ↓
{{ heroChipDetail }}
02 · How I run things
Outbound is a relationship business. Mine compounds.
why it compounds
Sends scale linearly. A personal brand compounds with every send, and enterprise buys only from people it knows.
▸
Same craft, two ways to run it: the full comparison
{{ fullHint }}
How most people run outbound
How I run it
{{ vr.label }}
✕{{ vr.t1 }}{{ vr.hl }}{{ vr.t2 }}
✓{{ vr.m1 }}{{ vr.ghl }}{{ vr.m2 }}
Safe automation, proved at scale
32
pages automated
30–50
invites a day, each
7
years, every day
24 active · 7 inactive · one ban, ever
OPEN · confirm publicly safe
The latest close, from record
2022met at a conference
2023follow-ups, timed to the lead
2024–25networking, content, no pitch
2026buying signal spotted
2026deal closed
some leads have been warmed in live campaigns for 5 years and counting
03 · The event circuit
Six years of running the event channel.
Built from nothing in my first month on the job. Now a booked calendar, every event owned end to end:
Hover any event for what it actually returned. Click to pin it.
{{ confStats }}
{{ yr.y }}
{{ yr.count }}
warmer chip = more meetings booked · ● = ongoing · updated 25 Jul 2026
of every 100 replies, how many became meetings
{{ g.pct }}{{ g.y }}
events in 3 years
52
replies
5,782
meetings booked
1,308
Sharper targeting, messaging and strategy every year. meetings capped by rep availability, overflow routed to calls
04 · The build log
The build log
hover or tap any company for more details
▸
in
Missed the team video on the home page?
17 of the 50+ people I hired and trained, each badge a real LinkedIn profile
{{ teamVidHint }}
Every badge links to a real person I hired or trained. The picture is a render, not their actual faces, and only a fraction of the actual headcount.
05 · Cases
Four campaigns, one instinct.
Different companies, different years, same move: find the moment a prospect is already primed, then arrive holding their own data.
01
▸
The prediction that sold itself
fintech · market forecastingclosed well-known mutual funds, banks and pension fundsOPEN · outcome number pending retained as consultant for years after
A fintech that forecast short-term market moves at 56 to 58% accuracy. Real edge, zero outbound. Claiming "58% accurate" in a cold email convinces no one, so the campaign proved it instead:
01
find what they own
public recordsstocks they owntheir biggest bet
every investment firm's stock picks are public record. Start there
02
predict their stock
run the prediction on stock they actually hold
03
give it away
sent freeno strings attached
send it, and wait
04
it comes true
it comes true, and the follow-up writes itself
the free prediction, sent cold
Your fund holds [their actual position] 1. Our model says it moves this direction, by Friday 2. Watch it. No call needed. 3
1 · their money, not a demo
2 · dated before the move
3 · no ask at all
05
the loop
1prediction sent, free
258% of leads watch it come true
3no reply? send the next one
↻every prediction they watch come true lifts the reply rate
1,000 got one
580 watched it come true
336 saw a second
week 0
1,000 firms get a free prediction on stock they hold
week 1
580 watch it come true
week 2
336 watch it twice
a track record, not a pitch · illustrative at 58%, not a guarantee of returns
week 01,000 firms get a free prediction on stock they hold
week 1580 watch it come true
week 2336 watch it twice. A track record, not a pitch
week nno reply yet? the next prediction goes out anyway. Every one they watch come true makes the next message harder to ignore, so the reply rate climbs with each round
illustrative at 58%, not a guarantee of returns
Nothing was claimed that the prospect had not just watched happen. I stayed on as their consultant for several years after.
02
▸
Competitor reviews, turned into pipeline the day they land
B2B software services70% reply ratedozens of projects closedsame-day trigger outreach
The day someone posts a bad review of a competitor is the single best day to reach them. Nobody watches for it, so I built the watcher:
02
the review lands
★★★★★
"missed both release dates, then the team went quiet on us…"
VP of Product · Austin, TX · name and company hidden
03
AI names the reviewer
project details ✓ · industry ✓ · timeline ✓
lead title ✓ · lead location ✓
name and company hidden
→ James C. at Meridian Software
lead title ✓ · lead location ✓
name and company hidden
→ James C. at Meridian Software
04
The message, same day
I saw your review from this morning 1 about [the exact project complaint] 2. We fix exactly that type of breakdown. 3
1 · timing: the day it lands
2 · their words, their pain
3 · one claim, no fluff
70%
reply rate
Old happy reviews get worked too. Someone who outsourced once, and liked it, is the most likely person to outsource again.
03
▸
Outbound that opened with the prospect's own load time
e-commerce · Shopify performancesubscriptions 2x in 3 monthsenterprise clients signedcontracted engagement, company unnamed
A company making Shopify stores up to 2x faster had spent years buying inbound with minimal results. The fix: stop describing speed, and open every message with the prospect's own number.
02
measure their speed
PageSpeed: seconds their store takes to load
03
keep only the slow
6.2s ✓2.4s ✕5.8s ✓1.9s ✕
the send list: measurably slow stores only
04
open with their number
"your store loads in 6.2s…"
the message starts with their own load time, not a pitch
the first line of every message
Your store loads in 6.2 seconds 1, [Competitor]'s in 2.9 2. Here's what a second costs in revenue. 3
1 · their own reading, not a claim
2 · measured against their rivals
3 · third-party research ↗, not our word
2x
subscriptions
in 3 months
in 3 months
04
▸
One acquisition filter beat a year of cold calls
healthcare · data maskingsteady clients from one campaignthe startup itself got acquired
Data masking matters at exactly one moment: when patient data is about to be shown to outsiders. The campaign waited for that moment:
02
the filter
holds patient data ✓
just got acquired ✓
buyer integrating patient data ✓
just got acquired ✓
buyer integrating patient data ✓
all three, or no message
03
the moment
integration begins: outside contractors are about to see patient data
HIPAAPIPEDA
the one week masking stops being abstract
04the message, sent the week of the deal
Congrats on the acquisition by [Buyer] 1. As integration starts, outside contractors will be near your patient data 2. Masking keeps that clean without slowing the handover. 3
1 · the trigger, days old
2 · their exact situation, not a persona
3 · one claim, tied to that week
before
✕hundreds of cold calls a day
1 sale
in a full year
and that one came from the senior rep's personal network
after
✓one campaign, ongoing
steady clients
every quarter
the attention it drew got the startup itself acquired
the control: no trigger, no message. A company got contacted only when an acquisition had just put outside contractors near its patient data.
06 · SalesOps
The sales drop, explained by the data.
First assignment as SalesOps: sales had dropped and nobody could say why.
I join as SalesOps
01
the mystery
sales down for months, every explanation a guess
02
combine what syncs
LinkedIn outreachemail outreachscheduled callsscheduled in-person meetingscall recordingsCRM
no new tracking, just the data already flowing
03
bottlenecks surface
2business days
avg. first answer to a lead
the biggest issue surfaced, not the only one
04
the restructuring
team rebuilt around the read: half the size, twice the speed
the scoreboard it left behindbefore → after
{{ sr.k }}
{{ sr.d }}
{{ sr.b }}
{{ sr.v }}
the principle: reps sell, the system types
messages, 
auto-synced via APIstatuses, 
propertiesanalyzed and filled by AI
propertiesanalyzed and filled by AIrecordings, summariesprocessed by a custom-built recorder
calls, meetingspulled from calendars
business cardscustom app scans, fills LinkedIn + CRM
in-person notesthe only thing reps fill by hand
everything a system can hear, it records. Reps type only what no system was in the room for
OPEN · confirm drop/halving/recovery safe anonymized · year anchor · a real median reply-time number
07 · The working toolkit
The working toolkit
{{ grp.g }}
hover any tool for status and what it does. Click to pin
{{ toolName }}
{{ toolSite }} ↗
{{ toolGroup }} · {{ toolDesc }}
{{ toolStatus }}
color = current stack · dimmed = not in use anymore
the AI and engineering stack is a different list, it lives on /automationOPEN · ?-marked tools need confirm-or-cut · per-tool ratings pending
the AI and engineering stack is a different list, it lives on /automation
08 · Off hours
When the laptop closes.
riding
Yamaha MT-07
my favorite gym commute
playing
poker tournaments
won the bad beat jackpot at Wind Creek, Bethlehem
building & racing
FPV drones
crashed one on a cloudy day at the Duga radar, Chernobyl. filmed ↗
based in
Allentown, PA
90 min from NYC · 60 min from PHL
22
● all green
self-hosted services on the home rack. The dishwasher is not one of them. Yet.
check out my automations →
09 · Standing invitation
If you got this far, just write.
I read everything sent here. Mozzi reads nothing. Between us, it's covered.
hello@meetvitaliy.com
The other side of me: the systems I build → /automation