A knowledge library, not a blog

Symptoms

• Next month's channel split looks like last month's, every month.
• Nobody can state the payback window of the second-biggest channel.
• Budget conversations are about totals, never about reallocation.

Why it happens

Channel budgets calcify because reallocating requires an argument, and without payback data every argument is a stalemate. The default wins, and the default is whatever history left behind, not what the math would choose today.

How to detect it this week

Ask for each channel's cost per closed customer, not per lead. If the answer takes more than a day to produce, the budget is running on habit.

The systemic fix

A tracking contract per channel: no source gets spend without click-to-revenue traceability. Then re-rank quarterly by payback, and let the ranking, not the meeting, move the money.

Symptoms

• Creative ships when it's ready, not when the last one stops paying.
• Nobody can name which ad variant is currently winning, or why.
• The same hero visual has run for months because everyone likes it.

Why it happens

When creative is treated as artwork, it gets judged by taste and retired by boredom. Ad fatigue is a measurable decay curve, but without versioning and per-variant numbers, the decay is invisible until the blended CAC has already absorbed it.

How to detect it this week

Pull the age of your currently running creatives and their CPA trend over that lifetime. If age is unknown or the trend line doesn't exist, the leak is active.

The systemic fix

Creative as testable units: versioned, launched against a control, measured on cost per outcome, retired by rule. The calendar stops deciding; the numbers do.

Symptoms

• Ads promise one thing; the page opens with something else.
• Paid landing pages share one generic template across campaigns.
• Bounce on paid traffic is far above bounce on organic.

Why it happens

A click is a question. When the page answers a different one, the visitor doesn't debug the mismatch, they leave. Message match decays silently because ads iterate weekly while pages iterate quarterly.

How to detect it this week

Take your top three ads, click each, and read the first screen of the page as if you were the person who clicked. Score claim, language and proof for match. Don't assume you're the exception until you've scored it; both funnel teardowns we've published found this leak.

The systemic fix

A message matrix: every ad group maps to a page section that repeats its claim, its language and its proof. Pages become modules of campaigns, not brochures that campaigns point at.

Symptoms

• The conversion rate is reported, never targeted.
• The last A/B test anyone remembers shipped two quarters ago.
• Funnel changes happen inside redesigns, so nothing is attributable.

Why it happens

A rate nobody owns becomes weather: observed, discussed, never changed on purpose. Redesigns bundle dozens of changes into one release, so even wins teach nothing, and the organization learns to treat conversion as fate.

How to detect it this week

Ask who owns the conversion rate and what experiment is running right now. Two blank answers equal one active leak.

The systemic fix

A standing experiment loop with a ranked backlog: research, hypothesis, test, ship, log. The rate becomes a managed metric with an owner, a cadence and a paper trail.

Symptoms

• Two people describe the same deal's stage differently.
• A spreadsheet exists next to the CRM, and it's the trusted one.
• Forecast meetings begin by reconciling numbers, not deciding.

Why it happens

Stages without entry criteria are adjectives, and adjectives drift by person. Once the CRM stops matching reality, people route around it, the workaround becomes the system of record, and every downstream decision inherits the noise.

How to detect it this week

Sample five open deals and ask two people to place each one in a stage independently. Disagreement on more than one is the tell.

The systemic fix

One pipeline with enforced entry criteria per stage and hygiene rules the system checks automatically. The spreadsheet gets a retirement date, not a sync.

Symptoms

• Hot signals (demo request, pricing page, trial spike) reach a human by email digest, or by luck.
• Median first-response time is measured in days, or not measured.
• Follow-up sequences exist in someone's head, not in the system.

Why it happens

Interest decays in hours, and manual follow-up runs in days. The gap isn't laziness, it's architecture: when routing depends on someone noticing, the system's response time equals the busiest person's backlog.

How to detect it this week

Time-stamp the path of your last ten inbound leads from signal to first human touch. The distribution, not the average, shows the leak: the long tail is where deals died.

The systemic fix

Automated routing with an SLA: signal fires, owner is assigned, context is attached, clock starts. Humans keep the judgment; the machine keeps the clock.

Symptoms

• Churn is discovered in the billing report, after the fact.
• Onboarding is a PDF or a single welcome email.
• Nobody can list the three signals that precede cancellation.

Why it happens

Retention fails upstream of the cancellation: in an onboarding that never reached the first value moment, in usage that faded for weeks unobserved. Billing is the last place churn appears, and the only place many systems look.

How to detect it this week

Ask what changes in a customer's behavior in the 60 days before cancellation. If the answer is a shrug, the signals exist but nothing is listening.

The systemic fix

Instrument the leading indicators (activation milestones, usage decay, support tone), wire them to owners and playbooks, and treat onboarding as a designed flow with a measurable finish line.

Symptoms

• Upsells happen when a customer asks, not when the data suggests.
• Referrals arrive occasionally and nobody knows what triggered them.
• Expansion revenue isn't a line anyone forecasts.

Why it happens

Expansion is the cheapest revenue a business has, and the least engineered. Without triggers, playbooks and owners, it runs on customer initiative, which means it runs at a fraction of its rate.

How to detect it this week

Compare expansion revenue as a share of new revenue against your own last year. Flat share while the base grows means the motion is passive.

The systemic fix

Expansion as a pipeline: usage-based triggers, a playbook per trigger, an owner per account tier, and a referral loop that fires at measured moments of success instead of hoping for goodwill.