Module 4 — Edge cases

Module time ~2 hours (reading + exercises) · ~3,400 words · 8 cited primary sources · Tool-independent · v1.0 · 2026-06-09

4.0 Why edge cases are the competence test

Modules 1–3 covered the discipline's normal operation: theory, workflows, process. This module covers the abnormal — the situations that arrive without warning, don't fit the playbook, and disproportionately determine reputations. The premise of including edge cases in a Foundations tier is one of this track's deliberate design choices: routine competence is table stakes; judgment under exception is what distinguishes a function someone can trust. A coordinator who executes calendars flawlessly but freezes when an ad account gets suspended is half-trained. The goal of this module is pattern recognition: you will likely face each of these situations rarely, but you should never face one for the first time conceptually.

Each section follows the same anatomy: the situation, the underlying pattern, the canonical response, and the escalation line — the point at which a Foundations-level operator stops acting alone and gets help (a manager, the founder, a lawyer, a platform rep). Knowing where the escalation line sits is itself Foundations-level competence; pretending it doesn't exist is how small mistakes compound into public ones.

4.1 Crisis communications — the SCCT framework

The defining marketing edge case is the crisis: something has gone publicly wrong — a product failure, an offensive post, a data breach, a viral complaint — and the next communication will either contain the damage or multiply it.

The pattern: response strategy must match crisis responsibility

The dominant academic framework is W. Timothy Coombs' Situational Crisis Communication Theory (SCCT), developed across two decades of empirical work (consolidated in Ongoing Crisis Communication, now in multiple editions). SCCT's central finding: the public's reaction is driven by how much responsibility they attribute to the organization, and the response strategy must be matched to that attribution. Three crisis clusters:

The classic failure is the mismatch: treating a preventable crisis with victim-cluster messaging ("we're saddened by these events") reads as evasion and converts a product story into a character story. The reverse mismatch — over-apologizing for a victim-cluster crisis — manufactures responsibility the public hadn't assigned. Diagnosis precedes drafting.

The canonical first response

Across the crisis literature the first-response checklist is stable: fast beats perfect (silence is read as guilt; an initial holding statement — what you know, what you're doing, when you'll update — buys time honestly); one voice (a single designated speaker; everyone else routes inquiries); facts before fault (never speculate about causes you haven't verified — corrections compound crises); and paused calendars (every scheduled post and campaign is suspended within the hour; the cheerful product promo landing mid-crisis is a self-inflicted second crisis — this is a standing line in your process from Module 3, §3.5's launch controls).

The reference case remains Johnson & Johnson's 1982 Tylenol response — full national recall, transparent communication, visible structural fix (tamper-evident packaging) — taught for four decades because it chose long-term trust over short-term cost, and won both. The escalation line for a Foundations operator: any crisis touching law, safety, or data goes to the founder and counsel before any public statement; your job is the holding statement and the paused calendar, not the strategy.

4.2 Platform dependency — disapprovals, deliverability, and concentration risk

The second edge-case family is mundane until it isn't: the platforms your marketing runs on can throttle, suspend, or change the rules at any time, without appeal to fairness.

Disapproved ads and suspended accounts

Every ad platform enforces policy through automated review, and automated review produces false positives routinely. The pattern: disapproval is information, not insult. The response workflow: read the cited policy (platforms name it), diagnose honestly (three common true causes: prohibited-category language, unsubstantiated claims — "best," "guaranteed," health/income promises — and landing-page mismatch with ad copy), fix and resubmit once, and only then appeal with a factual note. The escalation line: repeated disapprovals across campaigns signal a category or claim problem — stop resubmitting variants (which platforms read as evasion and punish with account-level review) and reassess the offer's framing. Account suspension is the severe form: respond only through official channels, factually, never with workaround accounts, which convert a recoverable suspension into a permanent ban.

Email deliverability red flags

Email's equivalent failure is silent: messages stop reaching inboxes. The red flags a Foundations operator monitors: open rates dropping sharply across campaigns (not one), spam-complaint rates above roughly 0.1–0.3% (the thresholds mailbox providers publish), bounce rates climbing, and spam-folder placement on your own test addresses. The standing causes: purchased or stale lists, missing authentication (SPF, DKIM, DMARC — ask whoever manages your domain; the concepts matter even if the configuration is delegated), sudden volume spikes, and consent shortcuts. The recovery pattern is always the same direction: send less, to fewer, more-engaged people, with cleaner consent — deliverability is reputation, and reputation rebuilds slowly. Module 6 covers the legal layer (CAN-SPAM, GDPR consent); this section is the operational layer.

Concentration risk — the strategic version

Disapprovals and deliverability are acute symptoms of a chronic condition: channel concentration. The PESO analysis from Module 2 (§2.3) becomes a risk register here. The test: if your single largest channel halved overnight — algorithm change, policy change, cost spike — what fraction of your pipeline disappears, and what's your recovery time? Businesses built on one shared or paid channel have experienced exactly this on every major platform across the last decade; the pattern repeats often enough that planning for it is Foundations-level prudence, not paranoia. The standing mitigations: the owned-channel conversion discipline from Module 2 (every rented-channel win feeds the email list), a tested second channel held warm (Bullseye's middle ring never fully closes), and the WIP-limited reactive lane from Module 3 so a platform shock doesn't consume the whole function.

4.3 Regulated-category marketing — knowing when you're in one

The third edge case is the one with legal teeth: some product categories carry marketing rules far stricter than the general FTC truth-in-advertising baseline (which Module 6 covers in depth). The Foundations competence is not knowing every rule — it is reliably detecting that you are in a regulated category and escalating before publishing.

The detection list — categories where marketing claims trigger specific regulatory regimes (United States; analogues exist in most jurisdictions):

• Health and medical — claims that a product treats, cures, or prevents any condition bring FDA and FTC scrutiny; supplements have their own claim rules; HIPAA constrains how patient information is used in marketing.
• Financial products — investment marketing is governed by SEC and FINRA rules (performance claims, testimonials, fair-and-balanced presentation); lending marketing by Truth in Lending requirements.
• Alcohol, tobacco, cannabis, gambling — platform bans layered on federal and state regimes; cannabis adds state-by-state variance and federal conflict.
• Children's audiences — COPPA governs data collection from under-13s; advertising to children carries its own FTC attention (Module 6, §6.5).
• Employment, housing, credit — ad targeting in these categories is constrained by anti-discrimination law; platforms enforce special restricted processes for them.
• Income and business-opportunity claims — "make $X/month" marketing is among the FTC's most-enforced areas.

The pattern for a Foundations operator: maintain the category awareness, write the conservative version by default (claims you can substantiate, qualified honestly), and treat "is this claim allowed?" as an escalation trigger, not a judgment call to make alone at 11pm before a launch. The cost asymmetry justifies the caution: a delayed campaign costs days; a regulatory action costs the company.

4.4 Attribution disputes — when sales and marketing both claim the win

The fourth edge case is internal: a deal closes, and two functions each believe they produced it. Marketing points to the campaign the buyer engaged with; sales points to the relationship and the close. The dispute sounds petty and is actually structural — it decides budgets, headcount, and whose model of the customer wins.

The Foundations-level understanding: the dispute is unresolvable at the level of single deals, and that's a measurement fact, not a diplomatic one. Module 1 (§1.4) established that buying journeys are non-linear and multi-touch; Module 7 will formalize why single-touch attribution models (first-touch, last-touch) are systematically biased toward whoever owns the touch being measured. Most real deals had many influences, several invisible to your systems (a colleague's recommendation, a podcast mention, three months of brand exposure). Anyone claiming clean credit for a multi-touch journey is choosing a flattering model, usually without knowing it.

The operational resolution is procedural, from Module 3's handoff contracts: a jointly owned definition of stages, disposition data flowing both directions, and — the piece that defuses most disputes — agreement that channel-level decisions use channel-level evidence (cohorts, incrementality, the Module 7 toolkit) rather than deal-level anecdotes. The cultural resolution is framing: pipeline is a shared production system (Grove again), not a relay race with a single baton pass. When the dispute recurs anyway — it will — the DACI from Module 3 names who decides budget reallocations, and the retro asks what data would have settled it faster. Module 5 picks up the relationship side of this border in depth.

4.5 Founder-led vs brand-led — the transition problem

The fifth edge case is specific to this track's audience. Early-stage marketing is usually founder-led: the founder's face, voice, network, and credibility are the channel. It works — audiences trust people before they trust logos, and the founder's authenticity is unfakeable. Then it becomes the bottleneck: every asset needs the founder, the audience follows the person rather than the company, and the marketing function cannot scale past one human's calendar — or survive their vacation.

The pattern: founder-led marketing is a loan, not a foundation. The strategic work is converting borrowed personal trust into owned brand assets before the loan comes due. The conversion checklist, drawn from the distinctive-asset discipline of Module 1 (§1.2):

1. Codify the voice. The founder's way of explaining things becomes a written voice guide — vocabulary, stance, taboos — so others can produce in it (the Brand Strategist track treats this in depth).
2. Build non-founder distinctive assets. Visual identity, named frameworks, a recognizable content format — things the audience can recognize in 100ms that aren't a face.
3. Migrate the audience to owned channels. The founder's personal following converts to the company list (Module 2's owned-channel discipline) while the relationship is warm.
4. Stage the substitution. Founder presence narrows to the highest-leverage moments (launches, flagship content) while the brand carries the routine cadence.

The edge-case judgment is timing: substitute too early and you discard your best channel; too late and the company's marketing has a single point of failure with a personal brand it doesn't own. The honest test: could the function hit its targets for one quarter with the founder fully absent? If the answer is no and the company is past its first marketing hire, the conversion work is overdue.

4.6 Endings — sunsets, discontinuations, and graceful exits

The final edge case is the one nobody plans for because it feels like failure: discontinuing a product, killing a service tier, leaving a market, or winding down a brand. Marketing usually inherits the announcement after the decision is made — and the announcement is where customer trust is either banked or burned.

The pattern: an ending is a marketing event with a longer memory than most launches. Customers forget your campaigns; they remember how you treated them on the way out — and they tell others. The canonical sunset communication, assembled from the same principles as §4.1 (honesty, speed, one voice):

1. The reason, honestly compressed. One sentence of true explanation beats three paragraphs of corporate fog. Customers detect fog and read it as disrespect.
2. The timeline, with margin. Dates for end-of-sale, end-of-support, end-of-life — set with enough runway that no customer is stranded mid-commitment.
3. The bridge. What happens to their data, their balance, their subscription; the migration path or the refund; what you are doing to make leaving easy. The bridge paragraph is the whole announcement's reputation payload.
4. The thank-you that doesn't grovel. Gratitude, brief and adult. The audience for a sunset notice includes every current customer of your other products, watching how you behave.

The same anatomy serves smaller endings: closing a channel ("we're leaving X; here's where to find us"), ending a partnership, retiring a content property — and the content-audit Retire lane from Module 2 (§2.4) at its largest scale. The escalation line: endings with contractual or refund implications go through counsel and finance before announcement; marketing owns the words, not the terms.