Your outbound is spraying a list nobody scored
A single trigger firing is not intent. Rank every account against your last 20 wins first, so scarce human time lands on the ones most likely to buy.
The 90-second version
The loud move this week was pointing an agent at every trigger feed teams could buy. More volume, faster, on a list nobody scored. That is not pipeline, it is a faster spray.
Jordan Crawford put the provocation plainly: run your triggers through the filter of your own closed-won deals, and almost none survive.
A single trigger is noise. A stack of signals that preceded your last deals is intent. Only the second earns a send.
Florin Tatulea reports signal-based plays converting several times better than cold outbound, but only when the signals are scored, not chased raw.
The European read: the trigger that predicts a deal is not the same in every market, so a scored list per country is how a lean team makes scarce human time land where volume bounces.
The thread: the edge is not how many signals you can buy or how fast an agent acts on them. It is whether you score them against your own win pattern before a human reaches out. A scored stack compounds. A raw feed just gets louder.
Read time: ~7 min.
This week’s number 1 move
Score last month’s triggers against your recent wins, and rank the accounts, before anyone sends a thing.
For 2 years the pitch was to add AI to outbound. This week the market raced to point an agent at every trigger feed it could buy. Jordan Crawford framed the counter-move on the signal-based selling discourse: most triggers are noise, so run them through the filter of your own closed-won deals and almost none survive. Florin Tatulea’s number lands on the same point. Signal-based plays convert several times better than cold, but only the scored ones do. Put the two together. The gap is not how many signals you can buy or how fast an agent can act on them. It is whether you score them against your own win pattern before a human reaches out. An agent can weight a live trigger feed against your last 20 wins in minutes, which no rep can do by hand. So build the scoring step, then let your people work the top of the ranked list.
My take: Here is the part most people skip. As the leader I do not just buy the signal tool. I set the standard for what a real signal is worth, because I own the number and I build the layer that carries it. You cannot delegate a number, and you cannot delegate the judgment underneath it either. The agent scores; the rep decides and reaches out, because personalized at scale is still not the same as personal. Monday morning, I would score last month’s triggers against my recent wins and rank the accounts before a single sequence goes out. The tool is not the edge. The standard you set for what earns a human touch is.
Do first: Pull the signals present in the weeks before your last 10 closed-won deals. Circle the ones that show up again and again. That short list is the start of your win pattern.
The rule: A single trigger is noise. A scored stack is intent. Score the signal before anyone reaches out.
Run it at your stage:
Founder-led, no agent yet: score your last 10 to 20 wins in a sheet by hand, rank this week’s accounts against them, and work the top. No agent, no new tool, no new spend.
At scale: point your AI-SDR layer at the scored base, let RevOps own the monthly retrain, and report the reply-rate lift to the board as the ROI on the tools you already pay for.
Sources: Jordan Crawford, “Signal vs insight based selling and the future of AI in GTM” (July 25) · Florin Tatulea, “How to use intent-based selling for sales development success” (30MPC, 2026)
Hi, it is Koen Stam and welcome to GTMcraft OS: The European GTM Operator. This newsletter is built from 100,000+ GTM signals collected from 100+ operators and founders, combined with 13+ years of my own lessons and failures from the trenches. I write at the intersection of go-to-market practice and AI-powered systems for founders and GTM operators scaling their next 2M, 5M or 10M ARR.
100,000+ GTM relevant signals from LinkedIn, Newsletters and Podcasts indexed. Translated into 100+ GTM plays, skills and training for you to implement today.
13+ years over 3 GTM operator jobs across 3 GTM motions (SMB, MM, ENT). Scaling from 2-10M+ ARR multiple times. Same recipe. Different motions.
Now all part of the GTMcraft Operator System.
Problem 1: Your outbound agent sprays a list nobody scored, so it scales the wrong accounts faster
You bought a signal tool and wired it into your outbound agent. Now the agent fires on every trigger the feed produces. A hire here, a funding round there, a page visit somewhere else, and it reaches out to all of them at machine speed. The problem is that most of those triggers never preceded a deal, so the agent does not fix a weak list, it scales it. Your reps chase alerts that mean nothing, and the account with real intent gets the same generic first touch as the rest. Todd Busler’s warning this week is blunt: the market is hot on signal-based outbound, and most of the firing is noise, so score before you act. The move on your pipeline motion is to put a gate in front of the agent. Before it sends, weight each trigger by how often that signal appeared before your recent wins, rank the accounts, set a threshold, and let the agent draft only for the top. Everything below the line goes to nurture, not to a rep.
My take: An agent pointed at an unscored list is the most expensive spray you can run, because it burns machine speed and rep hours at the same time. Personalized at scale is not the same as personal, and a generic first touch fired at 500 accounts is neither. Point a rep at an unscored feed and the week goes to a funding alert that meant nothing, while the account with 3 stacked signals sits untouched. That trade is why I do not let the agent decide who is worth a human. I set the gate, the agent ranks against it, and my people work the top. On a lean team that gate is the difference between outbound that compounds and a faster way to get filtered.
The European read: In smaller markets your account list is finite and word travels, so a wide unscored blast burns names you cannot get back and reads even flatter in a second language. A gated agent that only touches the scored top is how a lean team stays precise across countries instead of spraying every market at once.
Do first: Open your trigger feed from last month and count how many of those accounts actually became opportunities. That ratio is how much noise your agent is scaling.
Do this week: Write down the 6 or 7 signals that preceded your last 20 wins, give each a weight by how often it showed up, and set the threshold an account must clear before the agent drafts a touch.
Do this month: Put the gate in front of the agent so it scores and ranks the feed and drafts outreach only for accounts above the line, and route the rest to nurture.
The rule: If your agent acts on every trigger, it is a faster spray. Gate on the score, then send.
You know it worked when: Your reps work a short ranked list instead of a long raw feed, and reply rate on the touched accounts climbs.
The play: 5 steps to set the trigger-gated outbound sequence your agent runs (2026 series). Puts a scoring gate in front of the agent so it only drafts for accounts above your threshold, and the rest go to nurture. Best for 2-5M teams running signal-based outbound across markets.
Sources: Todd Busler, “The market is hot on signal-based outbound” (2026) · Florin Tatulea, “How to use intent-based selling for sales development success” (30MPC, 2026)
Problem 2: You treat one trigger as intent, so a stacked account and a random alert look the same
A single trigger tells you almost nothing. A company hired a role. A company raised money. On its own, each fires for hundreds of accounts that will never buy. Yet most setups treat every trigger as equal, so a random alert and an account with 3 stacked signals land in the same queue. Your best-fit account gets buried, and your rep spends the day on whichever alert popped up first. Kyle Poyar’s read on the intent revival is that automated signal plays convert several times better than cold, but the lift comes from the stack, not from any one trigger. The move is to score the base your agent runs on, not the trigger. Weight the signals, add a bonus when 2 or more fire on the same account, add a bonus for freshness, and down-weight signals that also preceded your losses. Now a stacked, recent account rises to the top and the lone alert sinks. The agent does the weighting across the whole feed in minutes. You decide the threshold and who gets a human touch.
My take: Winning in 2026 is not about more outreach. It is about mastering ICP, demand, and timing, and a stack of signals is timing you can actually see. One trigger is a coincidence dressed up as intent. I do not let reps guess which alert matters, because guessing is how the loud trigger beats the right one. I score the base: which signals, stacked how, actually preceded a buy. That base is the judgment I refuse to hand to a tool blind. The agent applies it at speed across the feed, but the definition of a real signal is mine, and it is the part that decides where my scarce selling time lands.
The European read: The stack that predicts a deal is not the same in every market. A hiring surge that signals intent in DACH can be noise in France, where the buying pattern is different. Import one signal base across your countries and you scale the noise faster. Score the base per market, so the ranked list reflects how each country actually buys, not how your home market does.
Do first: Take one account with a single trigger and one with 3 stacked signals. Ask your rep which they would call first, then check whether your tooling ranks them that way.
Do this week: Add 2 rules to your scoring base, a bonus for stacked signals on one account and a bonus for recent over stale, so the agent stops treating every trigger as equal.
Do this month: Feed your closed-lost deals back in and down-weight any signal that preceded losses as often as wins, so the base reflects what actually buys.
The rule: One trigger is a coincidence. A recent stack is a reason to call.
You know it worked when: Your top-ranked accounts are the stacked, recent ones, and your reps stop working alerts in the order they arrived.
The play: 5 steps to fix the base before you point an agent at it (2026 series). Turns your win pattern into the scored base the agent ranks on, so a stacked account beats a lone alert every time. Best for 2-5M and 5-10M teams tightening targeting across markets.
Sources: Kyle Poyar, “An outbound playbook for 2025” (2025) · Florin Tatulea, “How to use intent-based selling for sales development success” (30MPC, 2026)
Problem 3: Your scoring model never learns, so it keeps ranking signals that stopped predicting wins
Say you build the scoring base once. You weight the signals, rank the accounts, and ship it. 6 months on, the market has moved, and 2 of your top-weighted signals no longer precede deals. But the model never learned, so it keeps sending your reps to accounts that fit last year’s pattern. A scoring agent that never retrains decays into the same noise it was meant to cut. The edge in signal-based selling is not the base you build once. It is the loop that keeps it honest. The move on your pipeline motion is to schedule the retrain. Every month, feed the agent your new closed-won and closed-lost deals. Raise the weight on signals that keep predicting wins, cut the ones that now precede losses, and re-score the live feed against the updated pattern. Review the model quarterly, so the standard your team runs on stays tied to what is actually closing, not to what closed a year ago.
My take: The model stopped being the moat a while ago. The loop around it is. A scoring base you set once and never revisit is just yesterday’s judgment on autopilot, and yesterday’s judgment quietly stops matching the market. I treat the retrain like a forecast review: a standing ritual with one named owner, not a project someone gets to when they can. The agent does the re-scoring in minutes, but the decision to trust it again this quarter is mine, because I own the number it feeds. Kept honest, the loop is what makes a scored list keep beating a raw feed instead of slowly becoming one.
The European read: Each market moves at its own pace, so a signal that predicts in one country can decay first in another. A single global retrain hides that. Run the loop per market, so the ranked list in each country tracks how that market is buying now, which is also how you catch a new motion opening before your competitors do.
Do first: Look at your top 3 weighted signals and check whether each still appeared before your last 5 wins. If not, the model is already stale.
Do this week: Put a monthly retrain on the calendar with one named owner, and wire your closed-won and closed-lost data as the input.
Do this month: Run the first retrain, adjust the weights, re-rank the feed, and compare the new top accounts against the old list.
The rule: A scoring model you never retrain becomes the noise you built it to cut.
You know it worked when: Your signal weights shift each quarter to match what is closing now, and the ranked list keeps beating a raw feed.
The play: 5 steps to truth-test your pipeline before an agent forecasts on it (2026 series). Builds the recurring honesty check that keeps your scored feed tied to what is actually closing, so the agent never ranks on a stale pattern. Best for 2-5M and 5-10M teams running an agent on their pipeline.
Sources: Todd Busler, “The market is hot on signal-based outbound” (2026) · Jordan Crawford, “Signal vs insight based selling and the future of AI in GTM” (July 25)
Save this. 3 AI moves from this week’s Claude Signal, each on a real pipeline motion, with my operator take and a move you can run today.
Send it to 1 founder or GTM operator scaling toward 10M in Europe who reads everything and ships none of it.
Steal this move: Build a signal-scoring agent under your outbound motion in 4 weeks
The number 1 move is to score the signal before anyone reaches out, so here is the hero play in full. It moves your outbound off a raw trigger feed and onto a scored, ranked list your reps work top-down.
Play: 5 steps to turn outbound trigger noise into a signal-scoring agent
Fixes: An outbound agent firing on every trigger, so it scales a bad list faster and your reps chase noise.
Best for: 2-5M and 5-10M, any pipeline motion, and the European reality of a finite account list where a wide unscored blast burns names you cannot get back.
The 5 steps (Define / Build / Validate / Operate, over 4 weeks):
Week 1, Define: pull the signals present before your last 20 to 30 wins. Name the recurring, stacked ones. That is your win pattern.
Week 2, Build: weight each signal by how often it preceded a buy. Add a stack bonus and a recency bonus. Down-weight signals that preceded losses.
Week 3, Validate: have the agent score and rank the live feed. Set a threshold, take the top as the week’s list, and compare it to your gut list.
Week 4, Operate: feed each account’s driving signal into the sequence, keep a human on the send call, and schedule a monthly retrain.
Standing: review the model quarterly and re-score per market, so the ranked list tracks how each country buys now.
Lean version (founder-led, no agent yet): do steps 1 and 2 in a spreadsheet against your last 20 wins, rank by hand, and work the top 20 this week. Add the agent in phase 2, once the pattern is proven and the manual version is already earning replies.
Template: A 1-5 signal-scoring rubric. Rows: win pattern pulled from real closed-won, signals weighted not equal, stack bonus set, recency bonus set, losses down-weighted, monthly retrain owned. Score your setup 1-5. Anything under 3 is this week’s work.
Paste this into your AI:
Here is my closed-won and closed-lost export and my live trigger feed: [paste]. Find the signals that showed up in the weeks before my wins, weight each by how often it preceded a buy, add a bonus for stacked and recent signals, and down-weight any that preceded losses as often as wins. Then score and rank my live feed against that pattern and show me the top 20 accounts, with the driving signal for each.
Also on the radar
Jordan Crawford, “Signal vs insight based selling and the future of AI in GTM” (July 25). Run your triggers through the filter of your own closed-won deals and almost none survive. Before you buy another signal source, test the ones you have against your last 20 wins.
Florin Tatulea, “How to use intent-based selling for sales development success” (30MPC, 2026). Signal-based plays convert several times better than cold, driven by stacked signals, not any one trigger. Pick the 2 signals that stack most on your wins and start there.
Kyle Poyar, “An outbound playbook for 2025” (2025). Most accounts are not in-market at any moment, so scoring is the case, not reacting to every alert. Score so scarce human time lands on the few that are ready.
Why this matters now
For 2 years the outbound advantage was access to more signals. This week the market showed the next line. Point an agent at a raw trigger feed and you scale a bad list faster. Point it at a scored list and scarce human time lands on the accounts most likely to buy. The edge is not the number of signals you can buy or the speed of the agent. It is whether you score them against your own win pattern before anyone reaches out, and whether you keep that pattern honest as the market moves. A scored stack compounds. A raw feed just gets louder. This builds on last week, where the move was to own the layer that runs your revenue instead of renting it. This week is the upstream beat: own the standard for what a real signal is worth, so the right accounts reach a human at all.
What is your number 1 takeaway from this week’s signal: which of your outbound triggers actually predicts a win, and which are you paying to send to?
Reply or send me a DM.
Go deeper with GTMcraft OS
Scale on plays and skills, not prompts.
You do not need another newsletter, podcast or LinkedIn post.
You need the system under it.
Read the plays. Every single play is free and live in GTMcraft OS on Notion. Click through from any edition.
Build your own. GTMcraft OS members connect the full libraries, 100+ plays plus 100,000+ LinkedIn, newsletter and podcast signals, straight into their own AI or Claude context, so their team builds on the whole corpus instead of one-off prompts.
The single plays are free to read. Wiring the whole corpus into your own operating system, and keeping it current, is what the membership gives you.
Master your human GTM skills AI can’t replace inside GTMcraft,
Koen
Your (human) GTM Agent
PS. Co-written by Wispr + 3 GTMcraft Skills + Claude Opus 5; edited & approved by Koen
