When to Follow Up, Change Contacts, or Stop

Who this decision guide is for

This guide is for operators deciding what happens after delivery, silence, a bounce, a wrong-contact response, a referral, or an explicit stop. It applies to controlled B2B outreach, not to customer-support, legal-notice, or emergency communications.

Contents

1 Diagnose the current contact condition 2 Apply hard stops first 3 Choose follow-up only when four gates pass 4 Change contacts only for verified reasons 5 Pause when evidence or ownership is unresolved 6 Govern rounds, contacts, and time boundaries 7 Preserve company and person history 8 Use the decision matrix 9 Illustrative example 10 Common mistakes 11 What this does not mean 12 Common questions 13 Sources and methodology

1. Diagnose the current contact condition

Start from the last meaningful event, not from the number of days since sending. Build a Contact Decision Record with:

• account, person, pursuit, and channel;

• current lifecycle and outreach state;

• last message and its distinct purpose;

• delivery result and underlying provider code;

• reply, click, form, referral, or human-interaction history;

• contact validity and employment evidence;

• current suppression scope;

• contact count, round count, and pursuit-window dates;

• current owner, automation status, and next scheduled work;

• evidence date and next review. Mail failures are not interchangeable. RFC 3463 distinguishes persistent transient failure from permanent failure and provides more specific status categories. IETF, RFC 3463, "Enhanced Mail System Status Codes" Preserve the underlying result rather than applying one bounce label. A temporary system issue may support a controlled retry; a permanent address failure supports stopping that destination. Silence is an observation: no captured response within the stated interval. It does not establish whether the person saw the message, understood it, has the role, has interest, or is available. Keep those questions Unknown unless evidence answers them.

2. Apply hard stops first

The stop decision has precedence over scoring, cadence, and account attractiveness. Use at least these stop classes: Stop class Required action Scope question Unsubscribe suppress applicable marketing contact and cancel queued actions person, address, channel, brand, or broader stated scope? Company stop suppress the account within the stated scope all outreach, one offer, one legal entity, or one channel? Hard bounce / invalid address stop that destination; preserve diagnostic code is replacement permitted and independently supported? Departure or retirement mark the person ineligible for new rounds does a separate account pursuit remain eligible? Prohibited or protected contact stop under legal, policy, contractual, or account rule which entities and channels inherit the stop? Human takeover pause conflicting automation who owns the conversation and when can automation resume? The FTC's business guide states that covered U.S. commercial email needs a clear opt-out method and explains the obligation to honor requests. FTC, "CAN-SPAM Act: A Compliance Guide for Business" Canada's CRTC CASL FAQ discusses consent, identification, and unsubscribe mechanisms. CRTC, "Frequently Asked Questions about Canada's Anti-Spam Legislation" The exact applicable rule must be reviewed for jurisdiction, recipient, relationship, and channel. RFC 8058 defines a one-click list-unsubscribe signaling mechanism. IETF, RFC 8058 It does not replace all legal or operational unsubscribe requirements. Whether a stop arrives by header, link, reply, form, complaint, or staff entry, normalize it into one precedence-controlled suppression process.

3. Choose follow-up only when four gates pass

A follow-up is justified only when all four gates pass: 1 Eligibility gate: person, account, purpose, geography, and channel remain eligible and unsuppressed. 2 Usefulness gate: the next message has a distinct buyer-relevant purpose rather than restating the last note. 3 Boundary gate: contact count, round count, time window, and review date remain open under the approved policy. 4 Coordination gate: no human conversation, other owner, referral, complaint, or sensitive event conflicts with automation. Useful follow-up purposes include supplying requested context, clarifying a scope-matched problem, presenting a different piece of relevant evidence, asking one answerable routing question, or closing the loop. Just checking in does not by itself satisfy usefulness. Do not infer permission from a click. Do not mention tracking. Do not change channel merely because email was silent unless the new channel is separately eligible and appropriate. The UK ICO's B2B marketing guidance is a useful example of channel- and organization-type-specific rules: it notes that requirements can differ by method and type of business. Information Commissioner's Office, "Business-to-business marketing" Its jurisdiction and rules are not a North American template.

4. Change contacts only for verified reasons

Contact replacement is not a reward for exhausting a cadence. Replace or add a contact only when the operating policy permits it and one of these reasons is supported:

• verified wrong contact for the decision scope;

• explicit referral to a more appropriate person or team;

• invalid address or permanent delivery failure;

• verified departure, retirement, or changed responsibility;

• explicit inability or refusal to handle the topic without a company-wide stop;

• tenant or accountable human chooses to skip the person with a recorded reason;

• follow-up evidence establishes a material role or application mismatch;

• the defined time boundary expires without conversion and policy allows a new round. In the SellersGPT default operating concept, max contacts per company is five and the default is three rounds, but configured limits can vary. Treat the effective tenant setting as the rule, not the editorial default. A company without an active lead or meaningful human interaction can enter a later approved round; that does not mean the system should automatically fetch a new contact. Reuse the existing eligible pool unless a replacement reason applies. When a person is marked departed or retired, keep a durable ineligible status so the record cannot re-enter enrichment and new rounds. When an address hard-bounces, suppress the address; do not assume the person left the company unless separate evidence supports it.

5. Pause when evidence or ownership is unresolved

Pause is the correct output when action may be legitimate but material uncertainty would make continuation unsafe or incoherent. Examples include:

• two records may represent the same person;

• the person replied, but the content cannot be safely classified;

• an existing-customer collision needs account-owner review;

• a distributor and direct-sales team both claim ownership;

• a complaint or sensitive issue is under review;

• a referral conflicts with the current buying-group evidence;

• the applicable channel or suppression scope is unclear;

• a tenant handoff is offered but not yet accepted. A pause needs an owner, reason, start time, review deadline, and allowed actions. Paused must not become a hidden graveyard. If the deadline passes, escalate or apply a defined terminal disposition.

6. Govern rounds, contacts, and time boundaries

Use explicit settings:

• pursuit window: default 30 days, configurable to 30, 60, 90, 180, or 365 days;

• maximum rounds: default three, configurable within approved limits;

• maximum contacts per company: default five;

• per-contact message limit and minimum spacing;

• conditions for starting another round;

• replacement reasons that permit a new person;

• reset, draining, and scope-change behavior. The numbers are SellersGPT defaults, not industry benchmarks or legal safe harbors. A tenant can configure a broader boundary only where the channel, geography, data, purpose, and operating capacity support it. Use cohort dates. Round 2 should not silently restart the original 30-day clock without a recorded rule. If Primary ICP changes, apply the approved draining policy rather than instantly repurposing old contacts under new criteria. If a company produces an active lead or meaningful human interaction, stop automatic next-round entry and let the accountable owner decide.

7. Preserve company and person history

Suppression and evidence must survive deduplication, re-enrichment, imports, and contact replacement. The ICO's guidance on respecting preferences recommends retaining details on a suppression or do-not-contact list rather than simply deleting them, so future marketing does not restart by mistake. Information Commissioner's Office, "Respect people's preferences" Apply the relevant local legal and data-governance review; the operational point is to maintain the minimum durable control needed to respect the preference. For each terminal or replacement decision, append:

• normalized reason code and original evidence;

• scope and effective time;

• deciding rule version and human reviewer if applicable;

• canceled jobs and propagation result;

• eligibility for another channel, person, or future round;

• next review, if any. NIST's log-management guide supports formal policies for storing, accessing, and disposing of logs. NIST SP 800-92 It does not prescribe retention for sales records; determine retention from purpose, law, contract, risk, and company policy.

8. Use the decision matrix

The SellersGPT Follow-Change-Pause-Stop Matrix is a deterministic review aid: Evidence condition Follow up Change contact Pause Stop Valid, eligible contact; distinct useful purpose; boundaries open Yes No No No Silence only; existing pool still eligible Maybe, if all four gates pass No Maybe at review boundary No Explicit referral No conflicting follow-up Yes, to referred path if eligible If identity or ownership conflicts If referral includes stop request Wrong contact No Yes, if account remains eligible If replacement basis unclear Stop person if requested Hard bounce / invalid destination No Yes, only with permitted independent evidence If provider evidence conflicts Stop destination Departure or retirement No Yes, if account remains eligible No Stop person permanently from new rounds Unsubscribe No No circumvention Only for scope review, with no contact Yes within applicable scope Evidence condition Follow up Change contact Pause Stop Company stop No No Only for scope review, with no contact Yes for stated account scope Human interaction or takeover Human-owned action only Only by owner decision Pause automation Stop automation if required When more than one condition applies, the more restrictive result wins until reviewed. Record the chosen cell, evidence, owner, and rule version.

9. Illustrative example

Illustrative example - explicitly fictional. NovaSeal Components runs one fictional 30-day pursuit across 24 eligible accounts. At the first decision review, 40 contacts have these conditions:

• 22 valid, eligible, silent contacts with an approved distinct follow-up purpose;

• 5 valid contacts whose next message would only repeat the first note;

• 4 explicit wrong-contact replies, including two referrals;

• 3 permanent address failures;

• 2 verified departures;

• 2 unsubscribe requests;

• 1 company-stop request;

• 1 substantive human conversation already owned by the tenant. The decisions are: Decision Count Calculation Follow up 22 all four follow-up gates pass Retain without follow-up until review 5 usefulness gate fails Change contact 9 4 wrong contacts + 3 invalid destinations + 2 departures Stop person/address 7 3 invalid destinations + 2 departures + 2 unsubscribes Stop company 1 explicit company stop Human-owned; pause automation 1 active conversation Decision counts overlap by design: a departed person is stopped while the company may qualify for replacement, so Change contact and Stop person/address are not additive categories. The team creates two referral routes immediately and researches seven other replacements only under the approved account rules. These fictional counts are not response benchmarks.

10. Common mistakes

• Treating silence as proof that the person is wrong.

• Fetching a new contact after every unreturned message.

• Letting a score override unsubscribe or company stop.

• Treating hard bounce as proof that a person left the employer.

• Sending another automated message after substantive human interaction.

• Erasing departed or retired contacts so they can be rediscovered later.

• Starting a new round without an explicit boundary and review.

• Counting overlapping stop and replacement dispositions as distinct people.

• Pausing without an owner or deadline.

11. What this does not mean

This guide does not prescribe a universal number of messages, contacts, days, or rounds. SellersGPT defaults are configuration baselines, not legal safe harbors, response benchmarks, or promises. It does not provide legal advice or authorize contact in any jurisdiction. Changing contacts is not multithreading by default. Buying-group outreach requires role-relevant purpose and coordination under Module 03. Nor does a follow-up, replacement, or reply create a qualified lead unless the applicable qualification criteria are satisfied.

12. Common questions

Should we replace a contact after the first hard bounce? Stop the failed destination when the evidence indicates a permanent failure. Replace the contact only if the account remains eligible and the replacement policy permits an independently supported person. Do not infer departure from bounce alone. Does no response after 30 days mean stop forever? Not necessarily. The current pursuit window can expire without creating a permanent company stop. A later round requires the effective configuration, a fresh eligibility review, and no suppression or active human interaction. Can we contact another person after an unsubscribe? Do not use another person to circumvent the stated preference. Interpret whether the request applies to the person, channel, brand, offer, or company, apply the more restrictive safe state while unclear, and follow applicable law and policy. What if the contact says they are not responsible but gives no referral? Record wrong-contact evidence and stop follow-up to that person for the pursuit. If the account remains eligible, use approved public or first-party evidence to identify a replacement within account limits; do not guess or pressure the person for an introduction.

13. Sources and methodology The four follow-up gates, replacement reasons, default SellersGPT boundaries, and decision matrix are SellersGPT editorial tools. Sources below support limited points about email status, unsubscribe, suppression, record controls, and jurisdiction-specific direct marketing. They do not define a universal cadence. 1 PP04-FTC-CANSPAM - U.S. Federal Trade Commission, "CAN-SPAM Act: A Compliance Guide for Business." https://www.ftc.gov/business-guidance/resources/can-spam-act-compliance-guide-business Accessed August 25, 2026. 2 PP04-CRTC-CASL - Canadian Radio-television and Telecommunications Commission, "Frequently Asked Questions about Canada's Anti-Spam Legislation." https://crtc.gc.ca/eng/com500/faq500.htm Accessed August 25, 2026. 3 PP04-ICO-B2B - UK Information Commissioner's Office, "Business-to-business marketing." https://ico.or g.uk/for-organisations/direct-marketing-and-privacy-and-electronic-communications/business-to-business-marketin g/ Accessed August 25, 2026. 4 PP04-ICO-PREFERENCES - UK Information Commissioner's Office, "Respect people's preferences." http s://ico.org.uk/for-organisations/direct-marketing-and-privacy-and-electronic-communications/direct-marketing-guid ance/respect-peoples-preferences/ Accessed August 25, 2026. 5 PP04-RFC3463 - Internet Engineering Task Force, RFC 3463, "Enhanced Mail System Status Codes." https://www.rfc-editor.org/rfc/rfc3463 Accessed August 25, 2026. 6 PP04-RFC8058 - Internet Engineering Task Force, RFC 8058, "Signaling One-Click Functionality for List Email Headers." https://www.rfc-editor.org/rfc/rfc8058 Accessed August 25, 2026. 7 PP04-NIST-LOGS - National Institute of Standards and Technology, SP 800-92, "Guide to Computer Security Log Management." https://csrc.nist.gov/pubs/sp/800/92/final Accessed August 25, 2026. Related Knowledge What to Do After a Click, Reply, or Form Submission Classify the event and apply the correct stop or routing action before deciding on another contact. How to Define Entry and Exit Criteria for Every Pipeline Stage Encode contact boundaries and terminal states as testable stage criteria rather than informal judgment. How to Build a Pipeline Evidence Ledger Preserve the event, rule version, suppression scope, and decision behind every follow, change, pause, or stop. Make the next contact decision reviewable SellersGPT helps B2B teams continue, change, pause, or stop outreach from evidence and accountable rules. No subscription. No fixed monthly fee. Pay per qualified lead. Editorial illustration brief

• Asset type: Four-way evidence decision matrix.

• Aspect ratio: 16:10.

• Primary visual: One contact record passing through eligibility, usefulness, boundary, and coordination gates into Follow up, Change contact, Pause, or Stop outcomes.

• Color direction: SellersGPT yellow and warm white on charcoal, muted blue-gray for review states, green for permitted follow-up, amber for pause, and restrained red for hard stops.

• Prohibited visual tropes: Countdown clocks implying pressure, endless email sequences, robots, target icons, generic inbox screenshots, and contact-volume leaderboards.

• Alt text: A four-way contact decision with evidence gates for follow up, replace, pause, and stop. Implementation notes

• Render one H1 only; use the YAML title as the document title.

• Emit Article structured data from reviewed metadata and visible content; do not generate FAQPage schema automatically from Common questions.

• Store stop scope, person ineligibility, company eligibility, replacement permission, and automation status separately; do not collapse them into one status.

• Hide unpublished Related Knowledge cards; leave unassigned publication, reviewer, primary CTA, and asset fields null.