A B2B SaaS nurture sequence should stop as soon as the contact’s current state no longer matches the job of the sequence. If someone requests a demo, opens an opportunity, starts a trial, becomes a customer, is disqualified, opts out, or simply ages beyond the original signal, the next scheduled email is not “nurture.” It is a message sent with stale commercial context.
The fix is not a longer calendar or more personalized copy. Build nurture as a routing system. Define why a person enters, which state the program is meant to change, what messages are allowed, which fact forces exit, and where the record goes next.
At Stackmatix, the growth agency I co-founded, marketing work is judged by whether it produces worthwhile customers, not by how many automated touches run. That commercial perspective informs this playbook. The examples and numbers below are hypothetical editorial models, not client results or benchmarks.
A calendar is not a nurture strategy
“Send six emails over 21 days” describes timing. It does not describe the buyer, the decision, or the condition that makes email four irrelevant.
A useful B2B nurture motion should be able to answer:
- What verified event made this person eligible?
- Which person and account conditions must remain true?
- What unresolved decision is the content helping with?
- Which new fact changes ownership or message?
- When does eligibility expire without progress?
- Which system records the transition?
If those answers are absent, the sequence will drift toward generic education and activity reporting. Opens and clicks may show that a message was delivered or interacted with. They do not prove that the contact was still eligible for the program or moved toward a commercial decision.
Separate nurture from active sales follow-up
Marketing nurture and sales follow-up can support the same buyer journey, but they should not compete for attention.
| Current state | Primary owner | Useful job | What to suppress |
|---|---|---|---|
| Eligible prospect, no active buying process | Marketing | Help the person understand the problem, fit, and evaluation path | Generic sales pressure unsupported by intent |
| Demo requested or accepted opportunity | Sales | Resolve buyer questions and advance an agreed decision | Broad automated nurture that ignores the live process |
| Trial or product evaluation | Product, lifecycle, or coordinated revenue owner | Help the user reach and evaluate real value | Content that acts as if no evaluation exists |
| Customer | Customer team or defined expansion owner | Support adoption, retention, or a relevant next use case | First-purchase acquisition messages |
| Closed-lost or “not now” | Named owner based on reason | Re-enter only when time or evidence changes | Immediate recycling into the same sequence |
| Disqualified or prohibited | None until the controlling state changes | Preserve the reason and respect the restriction | All promotional re-entry from weaker signals |
The sales follow-up playbook shows how to keep active opportunity messages tied to buyer decisions and agreed next actions. Nurture should exit when that sales-owned process begins, unless the two teams have deliberately coordinated a specific account message.
Use the STATE nurture contract
Write five fields before building the workflow: Source, Target, Actions, Transitions, and Evidence.
Source: why is this person here?
Name the qualifying event, timestamp, person, account, and evidence quality.
Examples include:
- downloaded a specific implementation guide;
- attended a technical webinar and answered a fit question;
- viewed a pricing or integration path while belonging to an eligible account;
- was referred by a partner for a supported use case;
- returned after an approved cooling period with a new qualifying event.
“Known contact” or “marketing qualified” is too broad unless the underlying rule is available and current. The startup ICP framework helps define the account, use case, constraints, and buying context that must accompany the event.
Store the source as a durable record. If the campaign cannot explain the original signal after three weeks, the sequence cannot know whether that signal is still meaningful.
Target: what state should change?
Choose one primary transition. Examples:
- eligible reader to accepted evaluation request;
- trial evaluator to completed activation milestone;
- dormant but qualified account to renewed buying conversation;
- active customer to accepted expansion review.
Do not use “engaged” as the final state. Engagement is evidence that may support a decision, not the commercial decision itself.
The target should also identify the next owner and system state. A demo request that creates a form submission but no routed sales record is not a completed transition.
Actions: what is the sequence allowed to do?
Define the message job and the channels that fit the current state. A sequence might:
- explain the problem and relevant product mechanism;
- help a reader compare evaluation paths;
- answer implementation, security, or integration questions;
- provide a bounded case example with evidence limitations;
- invite the person to choose a next step or indicate “not now.”
Specify which claims, offers, destinations, and senders are approved. Do not let automation invent a stronger commercial promise than the landing page, product, or sales team can support.
Frequency caps and quiet hours can protect the experience, but they do not replace state logic. Adobe’s current business-rule documentation, for example, describes channel rules that can limit message frequency and set quiet hours. Those controls determine how often a communication can be sent, not whether the recipient still belongs in the journey. Adobe: business rules.
Transitions: what forces exit or rerouting?
List positive, negative, and time-based exits.
Positive exits
- requests or accepts a demo;
- starts the relevant trial or evaluation;
- reaches the defined product milestone;
- becomes a customer;
- accepts an expansion review.
Negative exits
- opts out or enters a controlling suppression state;
- is disqualified for account, use case, geography, role, or product fit;
- asks not to receive the sequence;
- enters an active opportunity owned by sales;
- opens a customer issue that makes the message inappropriate.
Time exits
- the original signal expires;
- the sequence window ends;
- the person is unresponsive and no stronger event renews eligibility;
- the account enters an approved cooling period.
The exit should name the destination. “Unenrolled” describes automation state. The business still needs a next classification such as sales-owned, trial-owned, customer, recycled with a review date, disqualified, suppressed, or complete with no state change.
Evidence: how will the team judge the motion?
Measure whether the routing and state change worked.
Useful evidence includes:
- eligible records entering by source and account;
- state-valid messages actually sent;
- exits by reason and time;
- accepted evaluations, trials, opportunities, or other named target transitions;
- incorrect messages sent after a state change;
- median delay between the controlling event and exit;
- sales or customer-owner corrections;
- mature commercial outcomes by entry cohort;
- people who reach the target state without the nurture motion.
Do not treat every later opportunity as nurture-sourced. Use the marketing-sourced pipeline rules to preserve an explicit source window and avoid assigning credit merely because a contact received an email.
Put suppression above inclusion
A person can satisfy several audience rules at once. The system needs a deterministic precedence order.
Use this starting order:
- consent, preference, policy, or prohibited-state suppression;
- disqualified or unsupported fit;
- current customer, active opportunity, or active evaluation state;
- campaign-specific exclusion or cooling period;
- approved entry rule;
- default holdout when identity or evidence is incomplete.
This prevents a new page view from overriding a stronger fact such as an open deal or active customer relationship.
HubSpot’s current workflow documentation says contact-based workflows can use suppression segments, goals, and settings to remove contacts who should no longer proceed. It also explains that a contact who enters a suppression segment while enrolled can be removed immediately from remaining actions. HubSpot: set unenrollment triggers.
Customer.io similarly documents exit conditions based on conversion criteria or no longer matching campaign filters. Customer.io: campaign exit conditions.
Those are implementation capabilities. The marketing team still has to define the business states, source fields, refresh timing, and conflict order.
Work through a hypothetical 1,000-contact program
Assume a fictional B2B SaaS company wants to nurture people who downloaded a technical evaluation guide. It begins with 1,000 contact records.
The numbers are hypothetical and are not recommended rates.
Entry review
| Starting state | Contacts | Routing decision |
|---|---|---|
| Existing customers | 100 | Exit acquisition nurture; route only to an eligible customer motion |
| Active opportunities | 80 | Exit; sales owns the current decision |
| Incomplete account or use-case evidence | 70 | Hold out until fit is resolved |
| Eligible prospects with no active buying process | 750 | Enter the 30-day nurture program |
| Total | 1,000 | 750 enter; 250 exit or hold |
Already, the entry rule is doing commercial work. Sending all 1,000 the same sequence would collide with 180 known customer or sales-owned relationships and would guess at the fit of 70 unresolved records.
State changes during the 30-day window
| Outcome among 750 entrants | Contacts | Exit or next route |
|---|---|---|
| Requested an evaluation | 45 | Exit to sales-owned evaluation |
| Started a relevant trial | 30 | Exit to trial activation path |
| Replied “not now” with a usable review date | 60 | Exit to reason-coded cooling state |
| Lost eligibility, opted out, or entered another suppression | 90 | Exit immediately with reason |
| Completed the window without target transition | 525 | Exit as completed; require a new event for re-entry |
| Total | 750 | Every entrant reaches a defined end state |
The target-transition rate is 75 of 750, or 10%, if the program’s target includes accepted evaluation requests and relevant trial starts. That arithmetic does not prove the nurture caused those transitions. Some people would have progressed without the sequence.
The more important operating checks are:
- Were the 180 customer and opportunity records suppressed before the first send?
- Did the 45 evaluation requests exit before the next scheduled email?
- Did all 90 suppressions stop the workflow within the defined service level?
- Did the 525 completed records remain out until a genuinely new qualifying event occurred?
- Did sales accept the evaluation handoffs, or did the entry rule create noise?
Run a holdout or another credible incremental design when the population and commercial stakes justify it. At minimum, compare similarly mature cohorts and preserve the state history needed to distinguish routing quality from assigned credit.
Treat “not now” as a state, not an invitation to restart
A buyer who says “not this quarter” has provided more useful evidence than a silent email recipient. Record:
- the stated reason;
- the source and date;
- the earliest sensible review date;
- the event that could reopen the conversation;
- the owner of the next review;
- any messages that remain appropriate during the cooling period.
Do not drop the person immediately into a generic long-term sequence. Re-entry should require the review date, a new qualifying event, or an owner decision.
The same rule applies to closed-lost opportunities. A loss for timing, missing capability, budget, no decision, or poor fit should not produce the same nurture path. Preserve the reason and only promise what the product and offer can now support.
Coordinate customer and expansion journeys separately
An active customer should not remain in acquisition nurture merely because the original contact still matches the entry filter. Customer communication needs its own outcome, eligibility, risk exclusions, and owner.
The product-adoption expansion playbook begins with verified account-level value and a specific adjacent need. That is a different state contract from educating an unknown prospect.
Likewise, the retargeting expiration policy applies the same principle to paid audiences: inclusion should expire, and current CRM or product state should override a stale site visit.
Diagnose a broken nurture system
| Symptom | Likely failure | First repair |
|---|---|---|
| Contacts receive emails after requesting a demo | Target transition is not an exit trigger, or the sync is late | Trace the request event, route, and next scheduled action |
| Customers receive acquisition offers | Customer state loses to a weaker inclusion rule | Move customer suppression above campaign eligibility |
| The program grows but rarely shrinks | Completion and expiration do not remove eligibility | Add an end state and require new evidence for re-entry |
| Sales rejects many handoffs | Entry event shows activity but not fit or readiness | Audit account, role, use case, and target-transition evidence |
| “Not now” contacts restart immediately | Cooling state has no controlling date or rule | Add reason, review date, and explicit re-entry trigger |
| Opens are healthy but opportunities are weak | Content engagement is being treated as commercial progress | Report accepted transitions and mature outcomes |
| Several workflows message the same person | Journey priority and concurrency are undefined | Create a cross-workflow conflict order |
| Exit counts suddenly fall to zero | Source fields, workflow settings, or syncs changed | Reconcile record samples against current state and rule version |
The first debugging move should be a record trace. Pick several people who received, exited, and were suppressed. Reconstruct the source event, account state, workflow eligibility, exit event, and next route as they existed at decision time.
Copy this nurture brief
Before launch, complete one record:
| Field | Decision to document |
|---|---|
| Program owner | Person responsible for the state contract |
| Source event | Exact event, timestamp, system, and evidence |
| Person and account eligibility | Required identity, fit, role, and use case |
| Target transition | State the program is meant to produce |
| Message job | Buyer question each message helps resolve |
| Allowed claims and offer | Approved promise, proof, destination, and next step |
| Positive exits | Progress events that end or reroute the sequence |
| Negative exits | Suppression, disqualification, conflict, or preference states |
| Time exits | Expiration, completion, and cooling rules |
| Conflict order | Which state wins when several rules match |
| Next routes | Sales, trial, customer, expansion, cooling, complete, or suppressed |
| Re-entry | New evidence required before another enrollment |
| QA | Test records, state samples, timing checks, and owner |
| Outcome evidence | Accepted transition, mature commercial result, and guardrails |
| Version | Effective date, changed fields, approver, and rollback condition |
The goal is not to keep every prospect warm forever. It is to send a useful message while the person remains in the state that makes the message true, then stop or reroute the moment that state changes.
Co-founder and CEO of Stackmatix, startup advisor, and former Head of Sales at MightyHive. Connect with Matt on LinkedIn ↗ · More about Matt →