Skip to content
Vanthis

ResourcesOutbound

Signal-Based Outbound

Spot buying signals on your mapped accounts, pick the right person and reason, then queue each to a sequence or a rep.

Use in Vanthis

21 steps2 decisions

ClaudeKnowledge baseClay×2ClayClayClayClaudeHubSpotSlackHubSpotClayClaudeKnowledge base
Shown with

This playbook watches your mapped accounts for the events that come before a purchase, such as a champion changing jobs, a new leader or a funding round. Each fresh signal on a fitting account gets one person, a short reason to reach out and a place in a sequence or a rep's day. Once a month the team checks which signals turned into meetings and tunes the rules.

Who does what

  • GTM Lead owns the signal rules, the capacity per rep and the monthly review, and signs off every new version of the rules.
  • Signal Builder the GTM engineer or RevOps operator who builds the monitors, the scoring, the contact lookups and the sequence enrollment.
  • Rep works the accounts routed to a person and approves the angle for dream and tier 1 accounts.
  • Reviewer samples the day's queue and settles the edge cases of the first decision.
  • Privacy Lead checks website tracking consent and the channel rules of each market.
  • Agent Claude or ChatGPT, studies past deals, drafts each angle and writes the monthly report.

What you need

  • The published account map with its tiers and exclusions, from ICP & TAM Mapping.
  • Clay for the monitors, the signal table and the scoring.
  • A CRM, HubSpot or Salesforce, with your won and lost deals.
  • A sending tool with one sequence per signal type: HubSpot Sequences, Smartlead, Instantly or lemlist.
  • A team chat for rep alerts, Slack or Microsoft Teams, and an agent for the study, the angles and the report.

Some answers are yours to set before the first run:

  • How many signal accounts each rep can work per day and tier.
  • The monthly budget for monitors, visit identification and enrichment.
  • The markets in scope, and whether website tracking runs behind your consent banner there.
  • The freshness window and the cooldown of each signal type.
  • Which tiers and bands go to a rep.

1Choose the signals

  • Start a signal watch
  • Find signals behind past wins
  • Signal rules
  • Set up signal monitors

The watch starts from the published account map, so every signal lands on an account that already fits. Before anything runs, the GTM Lead records the map version, how many signal accounts each rep can work per day and tier, a monthly budget cap and the markets in scope.

The agent then studies your won and lost deals. For each won deal it lists the observable events in the 90 days before the opportunity opened, compares them with the losses and keeps the three to five signal types that came clearly more often before wins. Each type gets:

  • A weight: how much the type counts in the score.
  • A freshness window: how many days the signal stays worth acting on. Sources disagree, such as 30 to 90 days for a champion's job change, so the window names its source and the monthly review tunes it.
  • Eligible tiers and a cooldown: which tiers it may act on, and how long an account rests after it was worked.

With few won deals the weights stay a hypothesis. The GTM Lead signs off the signal rules, which every later step reads.

The Signal Builder turns each type into a monitor in Clay. Past buyers, users and open or lost deal contacts go into a job-change monitor by LinkedIn URL, and hire, promotion, job-posting, news and funding monitors watch only the map's accounts. Website visits count at company level only, and in the EU the tracking script loads through the consent banner unless the Privacy Lead documents an exemption. Each monitor runs on a few accounts first, checked by hand, and gets a spending ceiling before it goes live.

2Catch and rank

  • Collect new signals daily
  • Signal table
  • Score each signal account
  • Act on this signal now?
  • Back on watch

Every day the monitors write new signals into one signal table, matched to a mapped account by domain. Signals from outside the map, noise such as visits from internet providers or your own staff, and signals on customers, open deals or objected contacts are marked instead of worked. Each row keeps when the event happened, when it was detected and its proof.

Clay then scores every account with fresh signals, so a rep can read why it was picked:

  • Weight times tier times recency: a strong signal on a weak fit cannot outrank a fitting account, and signals from the last week count far more than older ones.
  • Stacking: two signal types, or several people at one account, within about two weeks earn a bonus.
  • A plain reason: one line such as "new VP Sales plus two pricing visits this week".
  • A band by capacity: act today up to the day's rep capacity, then act this week, and the rest stays on watch.

The first decision asks whether to act now. The account must be in act today or act this week, its strongest signal inside its window and the account out of its cooldown. A single weak signal, such as one topic surge or one funding round, waits for a second one, and an account without a permitted channel in its market stays on watch and is flagged. Every account keeps its decision, the reason and the date it may be worked again.

3Pick person and reason

  • Pick the right person
  • Write the signal angle

For each account to act on, the Signal Builder picks the person the signal points to: the champion who moved, the new leader who was hired, or the person who engaged. Company signals such as funding or visits go to the map's persona whose problem the signal touches. CRM contacts and Lead Sourcing records come first, and a Lead Sourcing run is asked for only for who is missing. The person must still work at the account, and the email and phone eligibility from Lead Sourcing come along. Dream and tier 1 accounts may get a second contact.

The agent writes the angle, the one reason to reach out now:

  • It names what the company is doing now, such as a hire, funding or a stack change, and what that usually means for the person's goals.
  • For a champion's job change it recalls what worked at their last company and asks about their first months.
  • It never mentions a visit or a page someone read.
  • It checks the proof before the angle is approved, and the rep approves it for dream and tier 1 accounts.

The angle goes into a field the sequence or the rep reads. The sequence copy itself belongs to the outbound campaign.

4Queue the outreach

  • Route to sequence or rep?
  • Add to a sequence
  • Alert the rep
  • Reach out by hand
  • Review the queue and log
  • Signal accounts queued

The second decision routes each account:

  • Rep: dream accounts, champion and past-buyer signals, and tier 1 accounts that act today. Accounts in markets that do not allow cold email for this contact, such as Germany or Austria without consent, go to a rep for the permitted channels.
  • Sequence: tier 2 and tier 3, and tier 1 accounts that act this week, when the email channel is eligible and verified.

A dream account over the rep's daily capacity moves to the next day and never into a sequence.

On the sequence route the Signal Builder enrolls the contact into the sequence built for its signal type, with the angle passed as a field, within the sender's daily limit and only when the contact is in no other sequence. On the rep route the account owner gets a same-day CRM task and a chat alert with the person, the signal, its proof and the drafted opener, due before the signal's window ends. The rep checks the proof, makes the opener their own and reaches out by hand on an allowed channel. LinkedIn messages stay manual.

Every day the Reviewer samples each route for a fresh signal, the right person, a true angle and the right route. Every signal is logged with its decision, route and cooldown, including the ones kept on watch, and objections go to the suppression list Lead Sourcing uses.

5Review what converts

  • Monthly signal review
  • Report which signals convert
  • Tune weights and windows
  • Signals tuned

Once a month the agent links the logged signals to replies, meetings and opportunities within the team's attribution window and reports:

  • Queued accounts, positive replies, meetings and opportunities per signal type and tier.
  • Each type's meeting rate against the same tier's sequences without a signal.
  • The median signal age at queueing, the share of noise per monitor and the spend per queued account against the cap.

It proposes at most three rule changes, each with the accounts behind it, and changes nothing itself. The GTM Lead decides each one: more weight for types whose accounts led to meetings, less for the rest, and a type with no meetings over several reviews stops. The next version of the signal rules is dated and says what changed and why.

Compliance checks

The playbook carries checks for the EU, Germany and Austria. Four examples:

  • Under the EDPB's reading of Art. 5(3) ePrivacy and § 25 TDDDG, a visitor-tracking script in the EU usually needs consent, so it loads through the banner.
  • Website signals stay at company level, and an angle never mentions a visit.
  • A signal is no consent to contact. Cold email to contacts in Germany and Austria follows § 7 UWG and § 174 TKG 2021, which is why those accounts go to a rep.
  • LinkedIn's user agreement bans automation, so LinkedIn steps stay manual.

They are checks for whoever owns privacy on your team and no legal advice.

Run it in Vanthis

Pick your CRM, sending tool, team chat and agent above the Canvas, and the steps and guides on this page follow your choice. In Vanthis the playbook becomes your own Motion with that stack. Your agent, Claude or ChatGPT, asks which tools your team uses, then works through it step by step, and the GTM Lead signs off every rule change. To run it somewhere else, download it as Markdown or as an agent skill.

Use in Vanthis