Claude Prompt for Pipeline Gap Diagnostic and Paid Media Response Plan

Build a paid media response to a pipeline coverage gap by classifying each underperformance signal — paid-addressable vs. tracking artifact vs. capacity constraint — before proposing paid actions. The "What NOT to Do" section is what separates this brief from a list of spend recommendations.

Matt Danese

Senior Demand Generation Manager · 8+ years building B2B demand gen programs at Meta, Webflow, Medely, and Regal.ai. Specializes in AI automation for paid media, lead scoring, attribution, and marketing ops. · LinkedIn

Pipeline gap diagnostic: Build a structured paid media response to pipeline underperformance by pasting the pipeline report, current paid program overview, and any known tracking issues into Claude with this prompt. It classifies each signal as paid-addressable, sales process, tracking artifact, or capacity constraint — then returns specific paid actions, an explicit "What NOT to Do" section, and leadership framing guidance.

The Prompt

Production Prompt — Copy and use verbatim
You are a senior B2B SaaS demand gen director responding to a pipeline coverage report that shows underperformance. Leadership needs a concrete paid media response plan — fast. You know that the wrong response to a pipeline gap (cutting working channels because of broken tracking, for example) is worse than no response at all.

INPUTS

I will paste the pipeline report below — typically as a coverage update, QBR slide content, or RevOps dashboard export.

{PASTE_PIPELINE_REPORT_HERE}

{PASTE_CURRENT_PAID_PROGRAM_OVERVIEW_HERE}
(Example: "Active paid channels: LinkedIn (Document Ads, Sponsored Content), Google (NonBrand Search, Brand Search, PMax), Meta retargeting. Q1 paid contribution: $X to pipeline.")

{OPTIONAL_PASTE_KNOWN_TRACKING_ISSUES_HERE}
(Examples: "Warm MQL tracking has been miscounting for 6 weeks," "LinkedIn Conversions API is partially broken." Leave blank if no known issues.)

WHAT I NEED FROM YOU

Produce a structured response brief in this exact order:

1. Signal Diagnosis
For each underperformance signal in the pipeline report, classify it:
- Paid-addressable (paid media can directly impact this metric)
- Sales process issue (paid can support indirectly, not the primary lever)
- Tracking artifact (the metric is broken, not the performance — paid response would be the wrong fix)
- Capacity constraint (BDR or AE headcount issue, not a paid issue)

This is the most important step. Misclassifying a tracking artifact as a paid problem causes the wrong budget call.

2. Paid Response Plan
For each paid-addressable signal, propose specific paid actions. Examples:
- Late-stage win rate support: LinkedIn retargeting of open pipeline accounts by company name, multi-threading paid at Stages 5 and 7 with social proof and customer story creative
- BoFu volume gap: shift budget toward higher-intent creative ("Talk to our enterprise team" vs. "Get a demo")
- ICP fit issues: tighten audience filters on LinkedIn, refine Google audience signals
- Channel mix rebalancing: move budget toward channels with proven SAO contribution

Each action must include: specific channel, specific tactic, expected timeline, and the metric it's targeting.

3. What NOT to Do (critical section)
Explicit call-out of likely-wrong moves. Examples:
- "Do not shift budget away from [channel] due to [reported metric] — that metric is currently miscounted; the underlying performance is healthy."
- "Do not cut [campaign] based on the QoQ comparison — the prior-period number included a one-time effect that won't repeat."

This section is what separates a real response brief from a generic one. If leadership makes a wrong budget call based on broken data, the brief failed.

4. Framing Guidance for Leadership
Before any Q-over-Q budget decision is made based on the flagged signals, what should the conversation with leadership address? Examples:
- "Confirm Warm tracking is fixed before reading anything into Warm channel performance."
- "Validate SAO definition against the most recent operational change — Sales Ops updated criteria on [date]."

JUDGMENT RULES

- The first job of a pipeline response brief is to distinguish what paid can fix from what paid cannot. Misclassifying signals leads to wasted budget shifts.
- "Tracking artifact" is the most common misdiagnosis. If a metric drop coincides with a known tracking change, weight that heavily before recommending paid action.
- Late-stage support (Stage 5+) is real paid work. LinkedIn retargeting of open pipeline accounts, ABM display, and multi-threading creative are paid plays that influence win rate — they are not "demand gen" in the traditional MQL sense, but they are paid media plays.
- BoFu messaging shifts ("Talk to our enterprise team" vs. "Get a demo") materially affect lead quality. Lower-intent CTAs produce more leads; higher-intent CTAs produce fewer leads but higher SAO conversion. State the trade-off explicitly when recommending a shift.
- The "what NOT to do" section is what makes this brief useful. A response brief without it is just a list of paid actions — and leadership often acts on instinct before reading the actions. Lead with the wrong moves to prevent them.
- If the pipeline report is too thin to diagnose, say so. "The pipeline report does not break down SAO conversion by lead source — cannot determine whether the win-rate drop is paid-attributable" is better than guessing.

OUTPUT FORMAT

Return as {OUTPUT_FORMAT}.

If "markdown": full structure with ## headers per section.
If "slack": condensed for a leadership thread — signal diagnosis as a bulleted summary, paid response as 3-5 specific actions, "what NOT to do" prominently flagged.
If "html": exec-ready brief with clear section navigation.

Begin.

How to Use It

This prompt is built for the moment leadership presents a pipeline coverage report and asks "what are we doing about this?" The most expensive mistake in that moment is responding to a tracking artifact as if it were a performance problem — cutting budget on a channel that appears to be underperforming because its attribution is broken, not because the channel isn't working. The Q1 FY27 example in this prompt's PRD source: pipeline at 70% of plan ($25.2M vs $36.7M), Hot SAO win rate dropping from 12.8% to 8.8% QoQ, SDR SAO creation down 60%+ QoQ — and the first question before responding to any of those numbers should be "which of these signals are trustworthy?" Claude handles the signal classification more precisely than ChatGPT — Claude is more calibrated about labeling a metric as "likely tracking artifact" when a recent system change coincides with the drop.

The four-bucket classification (paid-addressable, sales process issue, tracking artifact, capacity constraint) is the entire logic of the prompt. Get it right and every subsequent paid action is targeted at a signal that paid media can actually move. Get it wrong and you burn budget on the wrong lever while the real problem (broken tracking, SDR capacity, win-rate degradation from a sales process change) stays unaddressed. The prompt instructs Claude to classify signals before proposing actions — the order matters.

The "What NOT to Do" section is the most distinctive and most valuable part of the output. It explicitly surfaces likely-wrong moves — "do not cut [channel] based on [reported metric] — that metric is currently miscounted" or "do not read this QoQ comparison at face value — the prior-period number included a one-time effect." Leadership often acts on instinct before reading the full brief; putting the wrong moves first means they see the most important information first. Every response brief should have a section like this — most don't.

Example Output

Live Example

Example output coming soon — currently running this prompt against live data and will publish the redacted output once it's ready.

Common Failure Modes

Variations

Two variations of this prompt are worth knowing.

Variation 1: Mid-Quarter Triage Version

Condensed for a mid-quarter leadership thread — signal diagnosis as a three-bullet Slack summary, paid response as five specific actions, and "What NOT to Do" framed as a single priority call-out. For when the pipeline gap surfaces at a QBR and you have 20 minutes to prep a response.

Coming soon

[PROMPT GOES HERE]

Variation 2: Late-Stage Win Rate Support Version

Focused specifically on win-rate degradation at late-stage pipeline — LinkedIn retargeting of open pipeline accounts by company name, multi-threading paid creative at Stages 5 and 7, and social proof creative recommendations for influencing deals already in-flight.

Coming soon

[PROMPT GOES HERE]

Get one new prompt every Monday.

Plus the system behind it. Free. Built for in-house demand gen managers at B2B SaaS companies.

Subscribe free →

Frequently Asked Questions

What's the difference between a "paid-addressable" signal and a "sales process issue"?

A paid-addressable signal is one that paid media can directly improve — low MQL volume, poor ICP fit in the lead mix, low BoFu conversion rate from paid traffic. A sales process issue is one where paid can support indirectly at best — slow follow-up speed after MQL creation, low contact rates from BDRs, declining win rates at late stage. The distinction matters because proposing a paid budget increase in response to a sales process issue wastes money without fixing the problem. The pipeline gap response brief is only as useful as this classification is accurate.

What does "tracking artifact" mean in the context of pipeline signals?

A tracking artifact is when a metric looks worse than actual performance because of a measurement problem rather than a performance problem. Common examples: Warm MQL tracking miscounting for several weeks due to a UTM convention change, OCI match rate drop due to a Salesforce sync rule change, or LinkedIn conversions undercounting because the Conversions API is partially broken. If you respond to a tracking artifact by cutting budget on the affected channel, you're responding to bad data with a real budget decision — the worst possible outcome. The prompt checks for known tracking issues explicitly in the inputs.

When should I include the "known tracking issues" input?

Always. If you know of tracking issues, include them — they're the most important context for the signal classification. If you don't know of any tracking issues, paste "none known" rather than leaving the field blank — it signals to the model that you've checked, rather than that you didn't fill in the field. Tracking artifacts cause more wrong budget decisions than actual performance problems, because they're invisible to the people making the decisions.

How does the "framing guidance for leadership" section work?

It surfaces the questions leadership should answer before making any budget decisions based on the flagged signals. Examples: "Confirm Warm tracking is fixed before reading anything into Warm channel performance," or "Validate the SAO definition against the most recent operational change — Sales Ops updated criteria on [date]." The framing guidance is the brief's way of buying time for correct diagnosis before a budget call locks in. Use it verbatim in the QBR discussion or in the leadership thread where the pipeline report was shared.