Owlbound is our project. We each run our own ventures and work with a small circle of partners, and this is the system we use to grow them. We are not trying to sell it.
We keep it evolving through research. Every play, signal and rule we test gets written down, and we publish what we learn as guides, field notes, courses and free tools, so the wider GTM community can use it, challenge it and push it further.
We stay open to new conversations with founders and operators in the same space, because that is how the system, and our own understanding of go-to-market, keeps growing.
With care, Hugo & Anis
P.S. If you are working on the same problems, write back. We read every note. Open a conversation
⌥How we structure it
Everything lives in code we can read.
Mission Control, the Allbound Engine and the GTM Brain are three repositories in our own GitHub, not a stack of tools. Code, rules, agents, dashboards and context sit side by side, so every change the method makes is something we can read, review and roll back.
github.com/your-company
One organisation · three repositories · every change a commit
01mission-control/
Mission Control
One place to see the outcomes and run the whole motion.
A 360° GTM dashboard: engagement, campaigns and outcomes in one view
Approvals, so a human signs off on plays before anything sends
Admin panels with advanced analytics
Forecasting from live pipeline data
A custom interface for everything
Call recording analytics
Website
CRM
02allbound-engine/
Allbound Engine
Where the actions run: outbound and inbound as one motion.
Email infrastructure for large volumes, end to end, on isolated IPs
Live updates in dedicated Slack channels
An agent that writes the blog posts
Outbound triggers that pull inbound
The Owlbound platform for outbound
A CLI to launch and run it
SEO and AEO optimisation
Inbound retargeting
collects
03gtm-brain/
GTM Brain
The record every agent reads. It collects what happens on both sides and writes it back into the ICP, the rules and the context.
go-to-market/16 folders · 81 files
syncing context
Strategy
Foundation, positioning, messaging, pricing and channels
Illustration. Three repositories in our GitHub: the brain collects from Mission Control and the Allbound Engine, and keeps every agent's context current.
◇Why we built it
Built for our ventures. Shared in the open.
We kept rebuilding the same go-to-market machine for every venture we started: the same lists, the same routing, the same lessons learned twice. So we wrote it down as one method, ran it across our own projects, and kept what survived contact with real pipelines.
01
Made for our own ventures
Every part of the method exists because one of our projects needed it. Nothing here was designed to be sold, so nothing here is padded to look bigger than it is.
02
Documented in the open
The engine, the playbooks and the signals are written up as research: what we tried, what moved the numbers and what we retired. You can read all of it.
03
Open to compare notes
We are exclusive and focused on our own ventures. We still make time for operators who want to improve the common knowledge with us.
★Field notes
Where the method was tested.
◎The authors
Outbound x Inbound.
The two of us built the method across our own ventures, one from the outbound side and one from the inbound side, working in the same system and the same context layer, from the ICP down to the routing rules and feedback loops.
Outbound expert
GTM Strategist
Hugo Lamy
Outbound & GTM strategist
3× founding GTM, seed → Series B · +$50M pipeline in 4 years
Leads the strategy side of the method across our ventures.
Designs outbound, founder-led, partner and event plays.
Defines ICPs, target accounts, offers and messaging.
One hacks how enterprises buy. The other builds the machine that executes it. Outbound x Inbound, unified in one system, one context layer, one motion.
Claude Workshop Invite #2Astrid · LinkedIn + email
7 replies from 131 contacts
5.3%7 / 131
2nd-Degree ConnectsLance · LinkedIn · Retail ops
5 replies from 127 contacts
3.9%5 / 127
0%12%
Claude Workshop Invite #2, verified
Contacts
131
Enrolled, LinkedIn + email
Human replies
7
Unique people who wrote back
New connections
31
26% of requests accepted
Email opens
49%
55 of 112 tracked sends
Field noteTeambridgeSeries B
Twice the results on a fraction of the ad spend
AV
Arjun Vora
Co-Founder & CEO, Teambridge
“Our team was nowhere close to Anis and Hugo's level… Now we get twice more results with ~$2k on ads spend on top of what Owlbound orchestrates than before with $100k–200k.”
What changed
01Paid acquisition cut from $100k a month to around $10k, for twice the results.
02Quality leads measured with the same definition on the legacy form and the new site.
03Monthly visitors grew from about 8.5k to 18k between April and June.
04Visitor-to-lead rate rose from about 0.15% to over 1.2% over time.
Quality leads: legacy form vs new site
Same quality definition applied to both · complete months, Apr 2025 to Jun 2026
Read from Teambridge's analytics charts. Visitor counts are rounded.
In Arjun's words
Field noteDigital JungleEvents space in San Francisco
Events turned into a growth engine
CA
Cyril Attia
CEO, Digital Jungle
“Owlbound transformed our events from a manual, fragmented process into a scalable growth engine, consistently reaching thousands of attendees and turning every interaction into measurable, six-figure revenue impact.”
What changed
01Event acquisition and follow-up used to depend on repeated manual work.
02$150,000 in additional event-services upsell compared to before.
03300 attendees generated consistently for events.
04Attendee behaviour and partner signals fed one persistent growth loop.
05Automated promotion, reminders, nurture and upsell workflows.
Field noteBrainbaseYC backed
An enterprise motion that sells to the Fortune 50
GE
Gokhan Egri
CEO, Brainbase
What changed
01Approximately $4M in monthly pipeline through an enterprise-focused GTM motion.
02$1M in ARR secured from Fortune 50 customers, including Toyota and Decathlon.
03Lead qualification improved by 40% and conversion rates by 35% with an AI-driven GTM stack.
Field noteTandemHexa backed
Qualified meetings, every single month
CB
Christophe Barre
CEO, Tandem
What changed
01Approximately $500K in pipeline per month, with 10–12 qualified meetings booked monthly.
0230+ deals closed from $40K to $100K ARR, selling from Seed-stage companies through Series C+.
03Reply rates up 54.7%, with 30% engagement from C-suite executives through ICP-led outreach.
Field noteConversion
A full-cycle pipeline, automated end to end
NT
Neil Tewari
CEO, Conversion
What changed
01Scaled the AI product from $0 to $4M ARR in 12 months.
0275%+ conversion rate across approximately 500 demos.
03An automated full-cycle pipeline that scaled prospecting and qualification across B2B and B2C motions.
DataEventsIntentMeetingsFlagship play
Signal-based outbound
A list of high-intent accounts is built across every industry you sell into, and an agent then listens to it without stopping: news, funding, leadership changes, hiring, and engagement with your own content. Intents stack per account. When enough of them land, the opportunity is attributed to whoever owns that industry, the right personas inside the account are pulled, and outreach goes out with the reason for the timing already written.
It also turns public engagement into pipeline. Point it at a topic, a conference or a hashtag; it collects everyone who interacted, enriches and scores them, adapts the messaging to each segment, and hands you a drafted email you send from your own inbox. The contacts land in Owlbound at the same time, ready for a multi-channel sequence.
Part 1 · Account signals
01Build the high-intent list
Accounts across each industry that fit, scored before anything is watched.
02Listen continuously
The agent never stops: news, funding, leadership moves, hiring, content engagement.
03Stack the intents
One signal is noise. An account reading your blog, announcing news and hiring a CXO is a priority.
04Attribute to the owner
The alert names the person who owns that industry, so it lands with someone accountable.
05Fetch the right personas
Inside the account, the agent finds the seats worth approaching, not just the one who triggered it.
06Outreach from there
Enriched, scored and handed over with the reason for the timing already written.
Part 2 · Social signals
01Point the agent at a signal
A topic, a conference, a hashtag or a post: anything people are publicly engaging with.
02It builds and enriches the list
Everyone who reacted, commented or posted is collected, enriched and matched to a person.
03It scores lead and intent
Each contact rated on both, with the reason for the fit written out in plain language.
04The alert lands in Slack
One message per contact in a dedicated channel, with a copy/paste LinkedIn DM ready to go.
05Publish the piece they need
A tailored, research-backed article written to the same intent, under your byline, so the first touch leads with value, not a meeting request.
06Open the email in Gmail
Pre-drafted and pointing at the article, colleagues already in copy. It sends from that person's own account.
07Leads land in Owlbound
Pushed to the platform and, when you want, straight into a multi-channel sequence.
KPI we watch
Stacked-signal accounts converted
Accounts firing three or more intents should convert far above the rest. If they don't, the signal set or the weighting is wrong.
KPI we watch
12 coffee chats from 100 outreaches
LinkedIn and email combined, all booked before the conference opened. Value-first outreach is why the rate holds, the article does the asking.
Why it matters
Stacked signals separate a priority from a ping.
Every alert has a named owner, so nothing is admired and dropped.
The list is built from public intent, not bought.
Leading with a useful piece beats leading with an ask.
The account is worked at several seats, not one.
One account, one alert: the company profile, the signals that fired with their sources, the key contacts enriched with email, phone and LinkedIn, and a written case for why it matters right now, ending in a HIGH score.Every qualified signal posts into its own Slack channel, one message per contact, with Lead and Intent scored, the reason for the fit written out, and a copy/paste LinkedIn DM already sized for the channel. Nothing auto-sends.Open the thread and the email is written and waiting. “Open in Gmail compose” loads it in your own inbox with colleagues in copy, send as-is or polish it first.The piece the email points at: published ahead of the outreach, written to the same intent, under your own byline. It is why the first touch gives something away instead of asking for time.Original research and data inside the article, not a rehash. It positions the team as the people who already did the thinking on the problem the conference is about.The article closes with the two people who will actually be in the room, so a reader can go straight from the piece to a conversation.The agent collects everyone who engaged with the signal, enriches them, scores them against the ICP and writes the opening line per contact.The draft opens in Gmail with your colleagues already in copy. Send it as written, or polish it first, it goes out from that person's own account.The same contacts land in Owlbound and can be pushed into a multi-channel sequence across email and LinkedIn.
A single signal is a coincidence. A stack is a programme, and the alert says which one, to whom, and what to do about it. Never auto-send: every DM and every email waits for a human to press send.
InboundIntentMeetingsFlagship play
Website-intent follow-up
Every lead the site produces is collected, qualified and activated without mixing its provenance, across more than 400 sources. Each record keeps where it came from and when, so attribution is real rather than guessed. High-intent actions, a booking-page visit above all, are routed automatically into the right Slack channel or straight to the person who owns it, with the follow-up written from what they actually looked at.
01Capture every source
Forms, signups, newsletter, industry pages, demo requests, imported lists and deanonymised visits: 400+ in one database.
02Keep provenance intact
Each lead carries where it came from, so nothing gets blended into an untraceable pile.
03Watch the high-intent actions
A booking-page visit or a demo form is treated differently to a blog read.
04Attribute the source
Captured time, page and source stored on the record, so the channel can be judged on meetings.
05Route automatically
Straight into the right Slack channel, or a notification to the member who owns it.
06Follow up in context
The outreach references what they actually looked at, not the fact that they visited.
KPI we watch
Alert-to-first-touch time
Plus meetings sourced per channel, which only means anything because provenance is kept on every record.
Why it matters
Four hundred sources, one database, no blending.
Provenance survives, so attribution is real.
No hot lead waits for someone to notice it.
One contact database behind the play: every lead with its source, the moment it was captured and its status, demo forms, Get Started signups, the newsletter, industry pages and deanonymised visits, kept separate rather than merged.
The Teambridge example turned a booking-page interaction into a meeting, the alert fired before anyone thought to check the dashboard.
OutboundMeetings
LinkedIn outbound for named accounts
Targeted lists built in Sales Navigator, personalised outreach, and structured follow-up against a named account list rather than a broad search.
01Build the named list
Sales Navigator, filtered to the account list, not a broad title search.
02Map the stakeholders
Several seats per account, so one no doesn't end it.
03Connect before you write
The connection lands before the first message does.
04Personalise per account
Research per account, not a merge field.
05Follow up on a structure
Fixed intervals and a defined end state for every thread.
KPI we watch
Connection acceptance
33.4% at Artefact USA, running alongside a 5.6% email reply rate. Then reply rate, then meetings.
Why it matters
Named accounts get research, not templates.
Acceptance is measured before reply rate.
Multi-threading beats single contacts.
Current core use case for Artefact US.
InboundIntent
Account deanonymisation & IP resolution
Surface the likely companies behind web activity, then enrich and prioritise the relevant contacts inside those accounts, for account-based follow-up where form fills are limited.
01Resolve the traffic
Anonymous sessions matched to a likely company.
02Confirm the account
Checked against the tiered list before anyone acts.
03Find the right seats
Enrich the contacts who actually own the problem.
04Prioritise
Fit plus observed intent decides the order.
05Route to the owner
With the pages they read attached.
KPI we watch
Identified accounts per week
And the share that convert into a first conversation.
Why it matters
Works where form fills are rare.
Turns traffic into an account list.
Feeds ABM without buying data.
ContentReplies
Positioning & competitive narrative
Messaging, competitive positioning and the category narrative, written early, so the market hears your framing before a better-funded competitor claims it as the default answer.
01Name the question
The one the category is actually arguing about.
02Map the competitors
What each one claims, and where the gap is.
03Write the angles
Positioning per ICP and per seat, not one deck for everyone.
04Publish first
The narrative belongs to whoever writes it first.
05Test in replies
The market's words beat the whiteboard's.
KPI we watch
Message-to-reply lift
Which angle produces positive replies, measured per ICP rather than per campaign.
Why it matters
The narrative is written by whoever writes first.
Angles are tested in replies, not workshops.
Positioning feeds content, outbound and ads.
DataIntent
Job Postings Monitor
Track relevant job postings across my list.
01Watch the boards
Postings across every account on the list, not just the big names.
02Match to the list
Only roles that mean something for your product.
03Alert, don't pitch
Tracking only, hiring is a signal, not a reason to sell.
KPI we watch
Relevant signals surfaced per week
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OutboundReplies
Best Customer Lookalikes
Find lookalikes of best customers and outbound with social proof.
01Profile the best
What your strongest accounts actually have in common.
02Find the lookalikes
Companies matching that profile, scored on fit.
03Lead with proof
The similar customer is the first line of the email.
KPI we watch
Reply rate against a generic list
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
InfraDeliverability
Private cold-email infrastructure
Higher-volume outbound run through dedicated email infrastructure, controlled inboxes, DNS setup, sending limits and message variation. Infrastructure plus operating standards, not a sequencing tool.
01Isolate the domains
Outbound can never damage the primary domain.
02Configure DNS
SPF, DKIM, DMARC, hosting and an isolated IP.
03Warm from day one
Started before the copy even exists.
04Cap and vary
Sending limits per inbox, message variation to avoid pattern detection.
05Rotate and retire
A burned identity is replaced, not rescued.
KPI we watch
Inbox placement
Plus bounce rate, complaint rate and days to first send. Deliverability is the tax on every other number here.
Why it matters
Outbound can never burn the brand domain.
Infra fully ready by day 16.
Volume scales without deliverability collapsing.
OutboundRepliesMeetings
Sequencing & persona campaigns
Select → Tier → Persona → Sequence → Measure → Dispose. One list becomes several campaigns: the angle and the ask change with the seat, and each seat gets its own scoreboard.
01Select from tiers
Accounts come from the scored list, never a bought file.
02Split by seat
Executive, marketing, AE / sales, RevOps.
03One angle per seat
Same offer, different proof, different ask.
04Measure replies
Unique replies, not sends.
05Follow the positives
Shift volume to seats that convert, not seats that reply.
KPI we watch
Positive rate per persona
Reply rate alone will send you toward the loudest seat. We rank seats by positives.
Why it matters
Each seat gets its own proof.
Volume follows positives, not noise.
No lead sits in limbo.
Live pilot, first 12 days, Executive 887 contacted, 1.5%; Marketing 590, 4.4%; AE/Sales 273, 3.7%. Marketing replies three times more and carries almost all the opt-outs.
PaidReach
LinkedIn engagement boost
Your internal team automatically likes and comments on each LinkedIn post, opt-in, per person, so it clears the early-engagement threshold that decides reach. Paid amplification then runs behind the posts that already performed.
01Opt the team in
Participation is optional and chosen per person, never assumed.
02Fire on publish
Likes and comments land inside the first-hour window that decides reach.
03Vary the comments
Real sentences from real accounts, never the same line twice.
04Watch organic reach
Only posts that genuinely earn attention move to the next step.
05Push ads behind winners
Paid amplification on proven posts, so spend follows evidence.
KPI we watch
Cost per engaged impression
Paid performance on boosted posts against cold posts. Organic lift is the leading indicator; ad efficiency is the lagging one.
Why it matters
The first hour decides a post's reach.
Ad spend follows proof, not hope.
Your team's network becomes distribution.
Participation is opt-in per team member, and comments are written rather than generated, the play boosts reach, it does not fake a conversation.
OutboundMeetings
AI-assisted “last mile” outreach
Research, enrichment, scoring and email drafting are automated; a human stays in control of review and sending. Sales receives a ready-to-send message with business context and a proposed meeting time.
01Auto-research
The account is read before a human opens it.
02Enrich and score
Fit and intent attached to the record automatically.
03Draft with context
A message that names the business reason, not a template.
04Propose a time
A concrete slot, so the reply is a yes or a reschedule.
05Human reviews and sends
The rep edits and presses send. Never the machine.
KPI we watch
Trigger → ready-to-send draft
Plus the share of drafts sent without edits, the honest measure of draft quality.
Why it matters
Reps stop doing research by hand.
Context arrives with the draft.
A human still owns the send.
OutboundReplies
Nurture existing high-value contacts
Focus on the 5,000–6,000 qualified contacts who already have interaction history, rather than continually rebuilding large cold lists. Engagement history decides who to re-engage, and when.
01Consolidate the base
One list of everyone who has ever engaged, deduplicated.
02Read the history
Opens, replies, meetings, site visits, event attendance.
03Segment by recency
Recent and deep engagement first; dormant contacts later.
04Re-engage with a reason
A new proof point, not a check-in.
05Suppress and recycle
Live deals suppressed; no-replies returned to the pool.
KPI we watch
Re-engagement rate
On known contacts, measured against the cold baseline. It should be several times higher.
Why it matters
Warm beats big, every time.
Existing history is free enrichment.
Less list-buying, more compounding.
OutboundMeetings
Reply intelligence & triage
Every reply is read by a human, classified into one of four buckets and turned into a next action the same day. An “interested” tag from a tool is not a yes.
01Pull daily
Every reply, every day, not weekly.
02Classify
Real positive, relevant-not-a-yes, hard no, do-not-contact.
03Book same day
Real positives go straight to a calendar invite.
04Mine the misses
Wrong-person replies become internal introductions.
05Feed DNC back
So the same bad list is never bought twice.
KPI we watch
Real positives
And the accuracy of the tool's own tag. A dashboard that inflates intent is worse than no dashboard.
Why it matters
Real positives get time the same day.
Wrong-person replies become introductions.
The dashboard never inflates intent.
In the live pilot the sequencer tagged four replies as interested. Two were real buying intent. We reported two.
InfraReplies
CLI-launched campaigns at scale
Campaigns built and launched from the command line rather than clicked together in an interface, hundreds of narrow, intent-specific sequences created in the time a team would spend assembling three. It runs on Email Bison, a private sequencer that costs $599/month on its own and is included at no cost with Owlbound.
01Slice by intent
One campaign per signal, persona and pain, not one campaign per quarter.
02Template once, in code
Structure, variables and guardrails defined once, then reused across every slice.
03Launch from the CLI
Campaigns created, populated and started in bulk from the command line, not clicked together in a UI.
04Route to the right identity
Each campaign assigned to the inbox that suits its volume and audience.
05Measure per slice
Every campaign reports on its own, so narrow winners are visible instead of averaged away.
KPI we watch
Reply rate per intent slice
Plus live campaigns per sending identity. Hundreds of narrow campaigns only beat one broad one if each is measured on its own.
Why it matters
A UI caps you at what a human can hand-build.
Narrow intent beats broad volume on reply rate.
The sequencer is $599/month standalone, included at no cost.
Runs on Email Bison, $599/month standalone, included free with Owlbound.
OutboundReplies
New Hire Cake Email
Send an email with a cake picture to any new hires in a job title I care about.
01Watch the hires
New starters in the job titles you care about, across the account list.
02Send the cake
Congratulations first. No pitch in the first message.
03Follow the warmth
New starters have budget, mandate and no loyalty to the incumbent.
KPI we watch
Reply rate on new-starter sends
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OutboundReplies
Product Launch Outreach
Send a LinkedIn message when a target company launches a product or a new feature.
01Track the launches
Across every account on the target list.
02Read what shipped
The message references the actual feature, not the press release.
03Message on the day
Launch week is the only week this works.
KPI we watch
Reply rate within 48 hours of a launch
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OutboundMeetings
New Office Cookies
Send cookies with an email when a target company opens a new office.
01Watch for new offices
Expansion is a budget signal and a goodwill moment at once.
02Send cookies with the email
The gift arrives, then the note.
03Visit if it's your city
An office opening is a reason to be in the building.
KPI we watch
Meetings per gift sent
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OutboundLeads
My Post Engagers
Enrich likers and commenters on my post, filter for ICP, email them.
01Pull the engagers
Everyone who liked or commented, enriched automatically.
02Filter for ICP
Most engagement is peers; only some of it is buyers.
03Email with context
The post they engaged with is the reason to write.
KPI we watch
Reply rate from post engagers
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OpsReach
Viral Post Slack Alert
Send a Slack alert on viral posts so the team can pile into the comments.
01Watch the feed
Posts from your market, your buyers and your competitors.
02Alert on velocity
Only posts climbing fast are worth interrupting the team for.
03Pile into the comments
Early comments on a climbing post buy real reach.
KPI we watch
Comments landed inside the first hour
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OutboundReplies
Competitor Post One-Liner
Send a one-line comparison to everyone engaging with a competitor's post.
01Watch their posts
Engagement on a competitor's content is a buying signal.
02Filter the engagers
Customers, prospects and tyre-kickers all look the same at first.
03Send one honest line
One sentence of genuine difference. Never a teardown.
KPI we watch
Reply rate on comparison sends
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OutboundLeads
Competitor Post to Email
Enrich a competitor's LinkedIn post, filter for ICP, email them.
01Enrich the engagers
Names and emails behind the likes.
02Filter for ICP
Only the ones you could actually serve better.
03Email the comparison
Specific, fair and short.
KPI we watch
Meetings sourced from competitor engagement
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OutboundIntent
Email Openers to LinkedIn
Send a LinkedIn connection request to anyone who opened my email 3+ times.
01Watch for repeat opens
Three opens is interest, however quiet.
02Connect on LinkedIn
A different channel, same person, warmer moment.
03Let the switch work
Acceptance rates from repeat openers beat cold connects by a distance.
KPI we watch
Acceptance rate from repeat openers
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OutboundIntent
Email Clickers to LinkedIn
Send a LinkedIn connection request to anyone who clicked my email link.
01Catch the click
The strongest passive signal an email produces.
02Connect the same day
Within hours, while the page is still in their head.
03Reference what they clicked
Not the email, the thing they went to read.
KPI we watch
Acceptance and reply rate after a click
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OutboundReach
Autolike the List
Auto-like LinkedIn and X posts for everyone on a list.
01Load the list
The accounts and people you are about to work.
02Like their posts
Quiet, consistent presence in their notifications.
03Warm before you write
The name is familiar before the first message arrives.
KPI we watch
Acceptance lift against cold connections
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
EventsMeetings
Coffee in My City
List ICP people in my city and ask for coffee.
01List who's nearby
ICP-fit people within travelling distance.
02Ask for coffee
Not a demo, not a call. Coffee.
03Keep it small
One ask, one place, one half hour.
KPI we watch
Coffees held per month
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OutboundReplies
Alumni Angle
Email alumni from my college who fit my ICP, referencing the school.
01Filter by school
ICP fit first, alumni status second.
02Reference it honestly
One line, early, without making it the whole pitch.
03Keep the ask small
Alumni open the door; the product still has to walk through it.
KPI we watch
Reply rate on alumni sends
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
InboundLeads
Inbound-to-outbound retargeting
Inbound engagement becomes outbound action: engaged accounts are identified, context is added, and the right follow-up is routed through email or LinkedIn. The allbound motion, connecting both sides.
01Capture engagement
Site, content, events and webinars, all into one record.
02Identify the account
Resolved and matched against the tiered list.
03Add the context
What they read, when, how often.
04Choose the channel
Email or LinkedIn, whichever that seat actually answers.
05Route and measure
Attributed back to the inbound source that started it.
KPI we watch
Meetings from inbound-triggered outbound
Reported separately from cold outbound, so the loop can be judged on its own.
Why it matters
Outbound gets a list nobody bought.
Content performance becomes a pipeline signal.
One loop instead of two teams.
InboundMeetings
Intent routing & CTA workflows
Intent scored and routed to a named owner with an SLA, and behaviour-triggered sequences fired by what someone actually clicked, not by a list upload on a Monday.
01Score intent
Pages viewed, return frequency, firmographic fit.
02Assign an owner
Territory, tier, existing relationship.
03Set the SLA
On first substantive contact, not first logged activity.
04Instrument the CTAs
Each behaviour mapped to one sequence and one wait.
05Suppress and hand off
Protect live deals; pass to a human at the threshold.
KPI we watch
Signal → substantive contact
The one number the inbound engine lives or dies on. Real response, not logged activity.
Why it matters
Rules are written down, not remembered.
Sequences match behaviour, not a calendar.
Breaches escalate automatically.
ContentTraffic
SEO & GEO content engine
Content produced for search engines and for answer engines: researched, drafted, published and measured on a fixed cadence from your own admin panel.
01Audit first
Technical audit and a website fix list before writing anything.
02Map the questions
Built from reply data, call recordings and support tickets.
03Publish on cadence
Research, draft and ship from your own admin panel.
04Structure for GEO
So AI answers can cite you, not just rank you.
05Compound authority
Backlinks and internal structure, over months.
KPI we watch
Sessions → identified → pipeline
Plus ranked queries and AI-answer citations. Traffic that never becomes a name is not a result.
Why it matters
Fixes land before content does.
Your buyers' own words drive the map.
Authority compounds instead of resetting.
ContentTraffic
Backlinks & domain authority
Backlink acquisition and authority work run alongside the content engine, so published pages can actually rank rather than sitting on a domain nobody trusts yet.
01Audit the profile
What links exist today, and what they are actually worth.
02Pick the targets
Publications, partners and communities your buyers already read.
03Earn the placements
Through partner work, guest content and original data.
04Fix the internals
Structure and interlinking before chasing external links.
05Track monthly
This one moves slowly; weekly reporting on it is theatre.
KPI we watch
Referring domains per quarter
And their effect on ranked queries. Authority is a lagging number, reported as one.
Why it matters
Content cannot rank on a domain nobody trusts.
Partner work and SEO reinforce each other.
Authority compounds instead of resetting.
Media costs for placements are passed through at cost.
PaidTrafficMeetings
Competitor keyword capture
Paid capture of buyers searching for named competitors, so the people furthest down the funnel find you at the moment they are comparing options.
01Map what they own
Which terms each competitor ranks and bids on today.
02Pick the winnable ones
High intent, reasonable cost, honest comparison.
03Build a real page
A genuine comparison, not a bait page, it has to survive scrutiny.
04Bid and measure
Spend judged on meetings, never on clicks.
05Feed it back
Winning terms become targets for the organic engine.
KPI we watch
Cost per meeting from paid search
Compared against the outbound baseline. If it is worse, the budget moves.
Why it matters
Comparison shoppers are closest to buying.
The page has to be honest to convert.
Paid finds the terms organic should own.
Media spend is passed through at cost.
ContentLead magnet
Whitepapers & agentic demo
Automated whitepaper production with gated downloads and a nurture trigger, plus an AI demo embedded on the site that adapts to each visitor's use case and logs every question it is asked.
01Pick the question
The one your buyer cannot answer internally.
02Draft and edit
The engine drafts, a human decides.
03Gate and nurture
Short form, sequence fired on download.
04Build the demo
The three or four use cases that actually close.
05Feed it back
The question log drives messaging and the content queue.
KPI we watch
Fit rate & demo → meeting
Downloads are vanity until you check them against the ICP.
Why it matters
Outbound gets an offer, not a calendar link.
Traffic turns into named contacts.
Questions become next quarter's messaging.
InboundLeads
Website Visitors to LinkedIn
Filter site visitors for ICP and enroll them in a LinkedIn campaign.
01Identify the visitor
Resolved to a person, not just a company.
02Filter for ICP
Only the ones worth a rep's attention.
03Enroll on LinkedIn
A connection often lands where a cold email would not.
KPI we watch
Identified visitors enrolled per week
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
InboundLeads
Product Signups to Email
Filter product signups for ICP and enroll them in an email campaign.
01Catch the signup
Every self-serve signup, enriched on arrival.
02Filter for ICP
Separate the hobbyist from the buyer.
03Enroll in the right sequence
Product context decides which one.
KPI we watch
Signup → meeting rate
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
ContentTraffic
Influencer Post to Blog
Send a blog post to everyone engaging with an influencer's post.
01Pick the post
An influencer post on the problem you solve.
02Match the blog
Your best piece on that exact question.
03Send the read
A useful link, not a pitch. The pitch comes later.
KPI we watch
Click-through to blog, then meetings
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
EventsMeetings
Conference, during
The days are for meetings and for material. Both are countable on the day.
01Block the days
One to three days purely for meetings.
02One booking link
Shared across the whole team.
03Reserve a quiet room
A café or private dining room nearby.
04Twelve a day
Half-hour meetings, with real breaks between them.
05Bring a gift and listen
A thoughtful small gift, a personal note, and their problem rather than your deck.
06Film five customers
A casual on-camera Q&A with strong-fit customers.
KPI we watch
12 meetings / day
And five customer stories filmed. Those conversations become material you can use long after everyone flies home.
Why it matters
The private room beats the booth.
Listening produces better notes than pitching.
The footage pays for the next event.
ContentReach
Conference, four weeks after
This is the phase everyone skips, and it is the one that makes the next event cheap.
01Edit
Send the footage to a freelance editor.
02Cut
Fifteen to twenty short, captioned clips.
03Publish
A clip every few days on LinkedIn and X.
04Reopen
The strongest clips start the next conference early.
KPI we watch
Cost per meeting
The whole playbook runs at a few thousand dollars. Plus clips published and meetings sourced from clips.
Why it matters
The footage keeps working for a month.
Clips open the next conference early.
Each event lowers the cost of the next.
EventsLeads
Event invitation & post-event follow-up
Targeted contacts invited with high personalisation, then attendees, no-shows and engaged prospects nurtured afterwards on separate paths.
01Pick the invite list
Tiered accounts with a real reason to attend.
02Personalise the invitation
Written to the person, not to the segment.
03Confirm and remind
Registration is not attendance; the reminder sequence matters.
04Split the follow-up
Attendees, no-shows and engaged prospects get different paths.
05Recycle the engaged
Non-converters return to nurture with the event as context.
KPI we watch
Attendee → meeting rate
Plus attendance against registrations, and pipeline per event.
Why it matters
Invitations are personal, not blasted.
No-shows are a segment, not a loss.
Every event leaves a list behind.
Particularly relevant to Artefact's event-led motion.
EventsLead magnetPipeline
Webinar engine
The whole webinar lifecycle as one play: topic generation, invite motion, the event itself, the clips, and the nurture that follows.
01Generate topics
From reply data and search demand, not from opinion.
02Run the invite motion
Email and LinkedIn as one sequence.
03Run the event
Recorded in full.
04Cut the clips
Straight into the content engine.
05Nurture both paths
Registrants and no-shows are not the same audience.
KPI we watch
Attendee → meeting rate
Plus pipeline per event. Credit consumption is per event, so cadence scales with appetite.
Why it matters
Topics come from evidence, not opinion.
One event feeds weeks of content.
Cadence scales without headcount.
+$1.3M pipeline generated consistently across 135 clients and 300 attendees.
PartnersRepliesReachFlagship play
Influencer & partner outreach
An agent goes and finds the creators, operators and community owners your buyers already follow, against criteria you set, and keeps a live database of them, scored, prioritised and explained. It then opens the conversation about co-marketing, sponsorship or a partnership. The names it surfaces are ones no bought list would have given you, and the reply rates are unlike cold outbound.
01Set the criteria
Audience size, vertical, whether they run a course or community, and who their followers actually are.
02The agent goes looking
It searches continuously rather than working from a list someone exported once.
03It builds the database
Name, LinkedIn, email, organisation, follower count, source and a written reason for the fit.
04It scores and prioritises
A confidence level and a P0–P2 priority on every row, so the best are worked first.
05It keeps the database current
Contact status and reply status update as the campaign runs, nothing goes stale.
06Then it opens the conversation
A short, specific message about a real collaboration. A human approves before it sends.
KPI we watch
17 replies from 95 contacts
An 18% reply rate on a live campaign, several times what cold outbound returns, because the ask is a collaboration rather than a meeting.
Why it matters
The agent finds people no list would have surfaced.
Every row carries the reason it is there.
Reach compounds: one partner lends you their whole audience.
The database the agent builds and maintains itself. Each creator arrives scored and prioritised, with follower count, channel, tags and a written “why fit”, and contact and reply status update as the campaign runs.Then it opens the conversation. This is a LinkedIn Top Voice with 341,000 followers replying within minutes: “Well, you have my attention. What are you thinking?”One campaign, 95 contacts enrolled: 17 replies at 18%, 27 connection requests accepted, and the queue surfacing exactly which contacts need a human next.
Audience size is no barrier: creators with hundreds of thousands of followers reply, because the message is about their audience and their work, not about your product.
PartnersMeetings
Ex-VP & warm introduction mapping
Former operators from inside your target companies identified and mapped into warm paths, so enterprise accounts are reached through someone trusted rather than through a cold sequence.
01List the closed doors
The enterprise targets cold outreach will not open.
02Find the alumni
Former VPs and operators who have since moved on.
03Qualify the relationship
Who still has standing internally, and who only has a title.
04Make the ask specific
One named person, one clear reason, one sentence to forward.
05Track the paths
Which introductions convert, and who keeps giving them.
KPI we watch
Warm-path coverage of Tier 1
What share of your top accounts can be reached through someone who already knows them.
Why it matters
Enterprise doors open from the inside.
One introduction beats fifty cold sends.
Paths are mapped before they are needed.
PartnersReach
Stakeholder networking
Every member of the founding and GTM team systematically connected to the people who matter in the market, so the network exists before anyone needs to use it.
01Split the map
Each team member owns a slice of the network, with no overlap.
02Connect deliberately
Target seats and accounts, not connection volume.
03Warm with context
Engage with their work before connecting, not after.
04Keep it alive
Light touches on a cadence, rather than one burst.
05Draw on it when it counts
Launches, events and introductions pull from a network that already exists.
KPI we watch
Relevant connections per team member
And how many introductions the network produces once it is actually needed.
Why it matters
Relationships take weeks to warm.
Five networks beat one.
Built before you need them, not during.
EventsPipeline
Premier event space, San Francisco
An exclusive partner space in San Francisco for executive dinners, roundtables and launches, the meeting a booth will never get you.
01Pick the cluster
The accounts worth putting in one room together.
02Invite twelve to twenty
Each with a real reason to attend.
03Run a roundtable
Not a pitch. The guests do most of the talking.
04Film the stories
While everyone is already in the room.
05Feed the next cycle
Clips into content, attendees into nurture.
KPI we watch
Pipeline per event
And attendee → meeting rate. A dinner that produces no second conversation was a dinner.
Why it matters
Pairs directly with the conference motion.
A room beats a booth.
The footage compounds.
EventsCoverage
Conference Keyword Tracking
Build a list of everyone posting about a conference by keyword. Tracking only, no outreach.
01Pick the keyword
The hashtag and the phrases people actually use.
02Collect the posters
Everyone talking about it, going back years.
03Track only
No outreach here, this play builds the list the others use.
KPI we watch
Attendees identified before the event
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
EventsMeetings
Conference Keyword Emails
Track the keyword, then enrich attendees, filter for ICP, and email them before the event starts.
01Track the keyword
The same collection as the tracking play.
02Enrich and filter
Names, emails, roles, then ICP fit.
03Email before it starts
The week before is worth more than the week of.
KPI we watch
Meetings booked before day one
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
EventsMeetings
Conference Attendees Coffee
Enrich an attendee list, filter for ICP, and ask for coffee.
01Enrich the list
Published attendee lists, sponsors, speakers.
02Filter for ICP
Twenty right people beat two hundred names.
03Ask for coffee
A half hour near the venue, not a booth visit.
KPI we watch
Coffees accepted per hundred asks
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
EventsPipeline
Conference Follow-Ups
Follow-up emails to the people you actually spoke with at a conference.
01Log who you met
On the day, while you still remember the detail.
02Write within 48 hours
After that you are one of forty identical emails.
03Reference the conversation
The specific thing they said, not the event.
KPI we watch
Follow-up → second meeting rate
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
EventsMeetings
Webinar Attendee Follow-Up
Follow-up email to webinar attendees.
01Split the audience
Attendees and no-shows are not the same people.
02Send with a reason
The recording plus the answer to a question actually asked.
03Route the engaged
Anyone who watches to the end goes to a rep.
KPI we watch
Attendee → meeting rate
A quick play: it runs inside the wider motion rather than replacing it.
Runs on
OpsForecast
360° GTM dashboard
Engagement, lead activity, campaign signals and outcomes brought into one operational view, so teams can see where attention is needed instead of reconciling fragmented reports across tools.
01Pick the questions
The handful of operational questions the view has to answer.
02Pipe every source in
Campaigns, site, CRM, product and support.
03Build one view
White-labelled, and readable by someone who didn't build it.
04Run the weekly off it
The decision meeting reads from that one screen.
05Retire the side reports
If a shadow spreadsheet survives, the view failed.
KPI we watch
Decision latency
Hours from a result appearing to a decision being made. Auto-reporting exists so nobody waits for an update.
Why it matters
No shadow spreadsheets.
Attention goes where it is needed.
One number, defined once.
Live on the platform: 13,385 contacts, 63 active enrollments, 8% reply rate, 61% completion, 11 identities.
OpsResponse time
Slack-based GTM orchestration
Enriched alerts, lead scores, context and drafted messages pushed into Slack or the team's own workflow, so the motion is actionable without anyone monitoring multiple systems.
01Decide what deserves an alert
Everything else stays in the dashboard.
02Enrich before sending
The alert arrives complete, not as a ping.
03Attach the draft
A ready message, not a task to write one.
04Act in the thread
Approve, edit or dismiss without leaving Slack.
05Log back
The action is written to the CRM automatically.
KPI we watch
Alerts actioned within SLA
An alert nobody acts on is noise, and gets removed.
Why it matters
Nobody has to watch five dashboards.
The action is one click from the alert.
Context arrives with the notification.
OpsContext
GTM Brain, persistent knowledge layer
Interconnected company, ICP, messaging and playbook knowledge maintained in one place, so AI workflows retain context over time instead of losing it between tools.
01Write the company truth
Product, proof, positioning, once, properly.
02Connect ICP and messaging
So a change in one updates the other.
03Store the playbooks
Every play in this library, versioned.
04Feed every workflow
Agents read from the brain rather than a prompt.
05Update from outcomes
What worked writes back into it.
KPI we watch
Play quality over time
Reply and positive rates as the system learns. The line should go up without new headcount.
Why it matters
Context stops dying between tools.
Every agent starts informed.
Plays improve instead of resetting.
DataForecast
CRM & data model
Objects, relationships and required fields defined once, then synced both ways, so every tool and every agent reads the same truth.
01Define the objects
Relationships and required fields, before any automation.
02Name the record
Identity resolution and a system of record per object.
03Sync both ways
Fully customisable Attio integration.
04Write stage rules
No entry and exit criteria means it's a label, not a stage.
05Enforce on entry
Validation at creation, not clean-up later.
KPI we watch
Field completeness & duplicate rate
Plus the share of stages with written entry and exit rules. Target: all of them.
Why it matters
Every tool and agent reads the same truth.
A weak model makes everything downstream unreliable.
Week one, not week nine.
DataForecast
Attribution & pipeline truth
Source truth first, then first and last touch, influence and multi-touch. Built for decision confidence, not for model elegance.
01Fix source truth
At creation, not retroactively.
02Scope each model
Agree what it is allowed to answer, and what it isn't.
03Compare on one set
Channels judged on the same opportunity set.
04Show leading and lagging
Side by side, in the same view.
05Ask the decision question
What investment decision does this inform?
KPI we watch
Forecast accuracy within tolerance
Can leadership plan from these numbers without keeping a shadow spreadsheet?
Why it matters
Channels compared on one opportunity set.
Models are judged by decisions, not elegance.
Leadership plans without a shadow spreadsheet.
$4.3M inbound pipeline of 241 leads, 80 demos and 553 opportunities only existed as a number once source truth was fixed.
OpsCost per meeting
The Owlbound Platform
The connective tissue between every tool in the stack, shaped to each motion rather than to a template.
01Store the signals
LinkedIn data and real-time social signals in one place.
02Coordinate the channels
Email, LinkedIn and signals fired as one action.
03Connect the stack
Clay and the sequencer through a CLI integration; Attio both ways.
04Publish and track
Admin panel for SEO publishing, lead tracking, CRM connections.
05Close the loop
Every output improves the next one.
KPI we watch
Credit spend per meeting
Plus cost per identity, and the share of GTM actions coordinated in one place rather than five.
Why it matters
Replaces per-identity tooling at ~$499/mo each.
Saves around 50% of Clay enrichment credits.
One source of truth for your team and your agents.
Standalone value $1,490/month, included at $0 with the service.
DataIntent
AI transformation signals
Target accounts monitored for signals around GenAI adoption, data modernisation, AI governance and AI hiring, with outreach shaped to whichever transformation priority the signal actually reveals.
01Define the signal set
GenAI launches, data platform moves, governance roles, AI hiring.
02Monitor continuously
Across news, job posts, filings and executive posts, not a quarterly sweep.
03Read the priority
The signal tells you which problem they are solving this year.
04Match the angle
Outreach references that priority, never a generic AI pitch.
05Score the cluster
Several signals on one account move it to the top of the list.
KPI we watch
Reply rate on transformation-signal outreach
Against the generic baseline. Accounts showing stacked signals should convert best, if they don't, the signal set is wrong.
Why it matters
The signal says which problem to lead with.
Generic AI outreach is indistinguishable now.
Stacked signals mark a real programme, not a pilot.
DataPipeline
Existing-client expansion signals
Current clients monitored for new initiatives, leadership changes, hiring and engagement with your content, with cross-sell opportunities surfaced to account teams alongside a suggested angle.
01Point the engine inward
The same signal monitoring, aimed at clients rather than prospects.
02Spot the adjacency
A new initiative sitting next to a capability you already deliver.
03Flag leadership change
A new leader is a renewal risk and an opening at the same time.
04Surface with an angle
The account team receives the signal and a suggested opening, not an alert.
05Track what converts
Expansion sourced from signals, reported on its own.
KPI we watch
Expansion pipeline from signals
Reported separately from new logos, so account teams can see what the monitoring is actually worth.
Why it matters
The easiest pipeline is already a client.
A leadership change cuts both ways, see it early.
Account teams get an angle, not just a notification.
OutboundMeetings
Named-account executive outreach
Account maps built across the C-suite, data, AI, marketing and business leadership, then coordinated outreach to several seats at once, built around a sector-specific problem you can actually solve.
01Map the account
C-suite, data, AI, marketing and the business owner of the problem.
02Pick the problem
Sector-specific and provable, not a capability list.
03Write per seat
A CFO and a Chief Data Officer do not share a reason to reply.
04Coordinate the timing
Several seats inside the same fortnight, so it reads as a campaign.
05Consolidate the replies
One account view, whoever answers first.
KPI we watch
Accounts with a live conversation
Multi-threaded accounts rather than individual replies, one reply inside a named account is not coverage.
Why it matters
One contact is not an account.
Executives reply to problems, not capabilities.
Coordinated timing reads as intent, not spam.
ContentLead magnetLeads
Industry signal intelligence → content
Recurring signals scraped and aggregated by industry, hiring trends, executive posts, technology adoption, announcements, regulatory shifts and the themes coming up in prospect conversations, turned into a whitepaper, benchmark or point of view and distributed back to the segment it came from.
What repeats across the sector, not what happened once.
03Publish the point of view
A benchmark or report the sector cannot get anywhere else.
04Distribute to the segment
Email and LinkedIn, to the accounts the data came from.
05Read the engagement
Who downloads and who shares becomes the follow-up list.
KPI we watch
Asset engagement → opportunities
Downloads and shares matter only insofar as they name accounts worth a conversation.
Why it matters
The data you already collect is the asset.
A sector benchmark opens doors a pitch cannot.
Engagement hands you the follow-up list.
This is the play that turns signal collection into a reason for the market to come to you.
ContentLeads
Thought-leadership distribution by vertical
Existing reports, webinars and case studies matched to the right sector and buyer profile, each with its own targeted distribution and follow-up sequence, so work already published keeps producing pipeline.
01Inventory the assets
Everything already published, tagged by sector and by seat.
02Match to the segment
One asset, one vertical, one buyer profile.
03Build the distribution
Email and LinkedIn sequences per asset, never one blast.
04Follow up on engagement
The asset opens; the follow-up carries the ask.
05Retire and refresh
Assets that stop converting get updated or dropped.
KPI we watch
Pipeline per published asset
Most firms measure downloads. The number that matters is meetings per asset.
Why it matters
Published work is the most underused asset.
One asset per vertical beats one for everyone.
Assets earn their keep or get retired.
EventsMeetings
Event attendee → qualified meeting
Priority accounts targeted ahead of the events that matter, your own, plus Google, Databricks and partner events, with personalised invitations before and follow-up shaped by who actually turned up.
01Pick the accounts
Priority accounts attending, speaking or sponsoring, not the whole list.
02Invite personally
One reason per account, written for the seat you want in the room.
03Confirm and remind
Registration is not attendance; the reminder does real work.
04Follow the behaviour
Attended, no-show and content-engaged each get a different follow-up.
05Qualify to a meeting
The ask is a working session, not a recap call.
KPI we watch
Attendee → qualified meeting
Measured per event, against the full cost of attending it.
Why it matters
An event is a list, not just a day.
The invitation is where the meeting is won.
Attendance decides which follow-up they get.
EventsPipeline
Post-event nurture
Attendees, no-shows and content engagers separated after every event, each receiving a relevant recap, case study or next step instead of the same generic newsletter.
Recap for attendees, recording for no-shows, proof for the curious.
03Offer a next step
One concrete thing to do, never “let us know”.
04Route the warm
Anyone engaging twice goes straight to the account team.
05Recycle the rest
Back into nurture, with the event as context.
KPI we watch
Post-event engagement → pipeline
Plus what share of registrants are still engaging thirty days later.
Why it matters
No-shows are a segment, not a loss.
A generic newsletter wastes the whole event.
Second engagement is the handover signal.
PartnersPipeline
Partner ecosystem activation
Accounts relevant to joint plays with Google Cloud, Databricks, Salesforce, Snowflake and other partners identified and worked together, co-marketing, events and co-selling around one shared account narrative.
01Overlap the lists
Your target accounts against each partner's customer base.
02Pick the joint plays
Where both sides genuinely gain, partner by partner.
03Agree the narrative
One story both teams tell, written down before anyone calls.
04Run it together
Co-marketing, events and co-selling on a shared calendar.
05Report per partner
Sourced pipeline attributed, so the relationship is judged on evidence.
KPI we watch
Partner-sourced pipeline
Per partner, per quarter. Relationships without attribution quietly stop being worked.
Why it matters
Partner customers arrive pre-qualified.
A shared narrative prevents two conflicting pitches.
Attribution keeps the relationship honest.
PartnersMeetings
Industry advisor & warm-introduction network
Former VPs, CxOs and senior operators from leading companies in each target industry, now consultants, advisors or independent, identified, mapped to priority accounts, and engaged for introductions, expert input or referral partnerships.
01Find the operators
Former VPs and CxOs now advising, consulting or independent.
02Map the relationships
Their former employers against your priority accounts.
03Qualify the standing
Who still holds real influence, and who holds only a former title.
04Choose the engagement
Warm introduction, expert conversation, referral or co-marketing.
05Maintain the network
Kept warm continuously, not contacted only when you need something.
KPI we watch
Introductions made per quarter
And the share of priority accounts reachable through someone already in the network.
Why it matters
Advisors open doors employees cannot.
Their expertise is also content.
A network only works if it is maintained.
OpsIntentMeetings
Strategic-account ABM
Website activity, LinkedIn signals, CRM history, event engagement and partner context combined at account level, so accounts showing several high-intent signals rise to the top with a recommended next action attached.
01Unify at account level
Every signal attached to the account, not scattered across contacts.
02Weight the signals
Some carry real intent; most carry none.
03Rank the accounts
Several signals together beat one loud one.
04Recommend the action
The next best step, named, per account.
05Review weekly
The account list moves, and the priority moves with it.
KPI we watch
Multi-signal accounts converted
Accounts showing three or more signals should convert far above the rest. If they don't, the weighting is wrong.
Why it matters
Signals mean more together than apart.
Sales get a ranked list, not a firehose.
Priority is reviewed, not set once a quarter.
OpsCoverage
Global-to-local campaign localisation
Campaign templates and GTM infrastructure built once globally, then adapted per region, targeting, proof points, content and messaging localised, while reporting and learnings stay shared.
01Build the template once
Structure, sequence logic and infrastructure, centrally.
02Localise the substance
Targeting, proof points and messaging per region.
03Keep definitions identical
Same KPIs everywhere, or nothing is comparable.
04Share what works
A winning angle in one region is offered to all the others.
05Report in one place
Regional detail underneath, global roll-up on top.
KPI we watch
Time to launch a new region
Plus like-for-like performance across regions, which only works if the definitions match.
Why it matters
Local teams own their market.
Templates stop each region rebuilding.
Shared definitions make comparison possible.
OutboundLeads
Page followers → connections
The followers of your LinkedIn page, and of every competitor and adjacent page, scraped continuously, filtered for ICP fit, and connected with a reason. Following a page in your category is a declared interest.
01Watch the pages
Your own page plus every competitor and adjacent player.
02Pull new followers
Continuously, so the list stays live rather than a one-off export.
03Enrich and filter
Role, company, seniority and ICP fit before anyone is touched.
04Connect with a reason
Reference the category, never the fact that you scraped them.
05Route the accepts
Accepted connections enter the matching sequence, not a blast.
KPI we watch
Acceptance rate on page followers
Against cold connections. Followers of a competitor's page should accept at a noticeably higher rate.
Why it matters
Following a page is a declared interest.
Competitor followers are in-market by definition.
The list refreshes itself every week.
Volume is capped per identity per day, this play compounds slowly on purpose.
OutboundIntent
Post engagers → connections
A standing watchlist of posts, yours, competitors', partners' and industry voices', with likers and commenters pulled continuously, filtered for ICP and connected while the post is still fresh.
01Build the watchlist
Accounts and hashtags whose posts your buyers actually engage with.
02Pull engagers continuously
As posts happen, rather than in a weekly sweep.
03Filter for ICP
Most engagement is peers, job seekers and noise.
04Connect while it's warm
Within days, referencing the post they engaged with.
05Feed the sequence
Accepted connections move into the play that matches the post.
KPI we watch
Accepted connections per week from post engagement
And how many turn into conversations, acceptance without a reply is just a bigger network.
Why it matters
Engagement is public intent.
Freshness decides whether the reference lands.
One watchlist feeds several plays.
InboundIntent
Profile visitors → connections
The people who viewed your team's LinkedIn profiles, captured continuously, filtered for ICP and followed up, the LinkedIn equivalent of deanonymising website traffic.
01Capture the viewers
Across every team member's profile, not just the founder's.
02Enrich and filter
Role, company and ICP fit, before anyone is contacted.
03Judge the intent
A repeat viewer, or one who visits just after a post, is worth more.
04Reach out lightly
A connection or a short note, never “I saw you viewed my profile”.
05Log to the account
The visit joins everything else that account has done.
KPI we watch
Profile visitors identified → conversations
The closest LinkedIn equivalent to website-visitor deanonymisation, and measured the same way.
Why it matters
A profile visit is intent nobody logs.
Every team member's profile is a sensor.
It joins the account's other signals.
Tone matters here: the follow-up never mentions the visit.
DataMeetings
ICP & tiering engine
“Who we sell to” becomes a scored, tiered list your systems act on every morning, the operational definition of where you actually win, not where you wish you won.
01Learn from closed deals
Interview closed-won and closed-lost; extract the real pattern.
02Write the filters
Firmographic, structural and timing criteria a machine can apply.
03Score the market
Map TAM and score every account against them.
04Tier the effort
Tier 1 / 2 / 3, each with a different level of effort.
05Re-score each cycle
From what actually closed, not what was hoped for.
KPI we watch
Meeting rate by tier
Tier 1 should out-convert Tier 3 by a wide margin. If it doesn't, the ICP is aspirational and every number downstream is noise.
Why it matters
Tier 1 gets effort Tier 3 never earns.
Every message has a reason to exist.
Delivered in weeks 1–2, before any copy.
At Artefact USA this scoring fed 39 active sequences rather than one undifferentiated list.
DataCoverage
TAM mapping & 100-point scoring
The total addressable market mapped and clustered by pain point and signal, then scored out of 100 on fit and timing, so targeting runs off a growing database rather than a static list.
01Pull the universe
Every company that could plausibly buy, before any filtering.
02Cluster by pain
Grouped by the problem they have, not by industry code.
03Attach the signals
Which triggers matter for which cluster.
04Score out of 100
Fit and timing combined into one number per account.
05Map the committee
Who signs, who blocks, who uses: named per tier.
06Keep it growing
New accounts enter and are scored automatically.
KPI we watch
Scored accounts in the database
And the share of closed-won that came from the top scoring band.
Why it matters
Targeting runs off a database, not a list.
Clusters explain why a message lands.
New accounts are scored on arrival.
OutboundRepliesMeetingsFlagship play
Cold email campaigns
Cold email run on our own private sequencer, isolated IP, dedicated domains, inboxes warmed before volume, so deliverability is something we control rather than hope for. Campaigns are split by seat, each carries one angle, and every reply is read and classified by a human before anything is called a win.
01Isolate the sending
Our own private sequencer on an isolated IP, with dedicated domains and inboxes warmed before any volume, so messages land in the inbox, not in promotions.
02Split by seat
Executive, marketing and growth, AE and sales, RevOps, website visitors: one campaign each, one scoreboard each.
03Write to one angle
A single idea per campaign, tested in the market rather than argued about in a workshop.
04Measure unique replies
Replies divided by contacts. Sends exceed contacts because of follow-ups, so sends are never the number we report.
05Read every reply
Four buckets: real positive, relevant-but-not-a-yes, hard no, do-not-contact.
06Report what is true
Including the replies the tool scored wrong, and the clusters that never launched.
KPI we watch
6.3% reply rate
Every reply in thread, across 7,123 contacts and 21,555 sends in nine campaigns. Unique repliers alone come to 3.2%.
Why it matters
Isolated sending keeps the brand domain out of the blast radius.
Reply rate by seat tells you where the volume should go.
A tagged “interested” is not a meeting until someone asks for time.
Across nine campaigns
6.3%
reply rate
452 replies from 7,123 contacts and 21,555 sends. Counting unique repliers only, 3.2%.
Campaign by campaign
7,123leads contacted
21,555emails sent
227unique replies · 3.2%
452replies in thread · 6.3%
CampaignContactedSentRepliesRate
Website Visitors [AUTO]246694166.5%
S.I.A4211,209194.5%
Security17148974.1%
Construction12936053.9%
LTC Penalties (v2)1,8376,216703.8%
Janitorial1,5835,880473.0%
MIX1,2843,657302.3%
Webinars Jun 168592,457202.3%
LAST CALL · Webinars593593132.2%
Every campaign cleared 2% on unique repliers, and the strongest ran at 6.5%, the one built from website visitors rather than a bought list.
Nine campaigns in the sequencer, each reporting contacted, sent and replies on its own, so a weak angle is visible immediately rather than averaged away.
ClusterContactedSentRepliesRate
Executive8871,488131.5%
Marketing / Growth590887264.4%
AE / Sales273286103.7%
Twelve days live. Marketing replies three times more than Executive and carries almost all of the opt-out volume, the copy is landing, the list is noisy. Executive is furthest through its sequence and replies least, which is normal for that seat. Two clusters never launched: that is the gap, not more volume into the ones that did.
The sequencer we run it on: multi-step campaigns with spintax variation per send, and a preview of the exact email a given lead will receive before anything goes out.
Reported as it stood on day 12, including the two tagged replies that were a personalisation correction and a one-word question, and the two clusters still unlaunched. Partner and prospect names withheld.
Brainbase
Tandem
Owlbound
A project by Hugo Lamy & Anis Nanai
