---

name: marketing-audit
description: "Trace actual pipeline to actual spend — not last-touch attribution — and find which channels you're over-investing in and which you're starving."

/marketing-audit

Last-touch attribution is fiction. It systematically overvalues the channel that closes deals (SEM, sales outreach) and undervalues the channels that create demand (content, events, word of mouth). Most marketing audits confirm what the CMO already believed because they start with the attribution model instead of the pipeline. This skill starts with pipeline and works backward to spend, using deal-level data to find the actual acquisition story — including the hidden issues no attribution model surfaces.

Spend Baseline

• Total marketing spend last 12 months, broken down by: paid media (by channel), content/SEO, events (field + virtual), marketing headcount loaded cost, agency/tools/infrastructure
• What percentage of spend is fixed (headcount, retainer) vs variable (campaign spend)?
• Which channels have been increased or decreased in the last 6 months — and why?
• What is the split between brand spend (no direct response signal) and demand gen spend?

Pipeline Tracing — for each channel/campaign:

Do not use last-touch. Use multi-touch or sourced-only analysis. For each:

• Opportunities sourced (not influenced): how many deals trace their origin to this channel?
• Pipeline value sourced
• Average deal size from this channel vs company average
• Win rate from this channel vs company average
• Average sales cycle length from this channel vs company average
• CAC: total spend on this channel divided by closed-won customers sourced from it

The channels that look expensive on CAC but have higher win rates and larger deals are often undervalued. The channels that look cheap on CAC but have low win rates and small deals are often overvalued.

Segment Fit Analysis

• Which channels are generating your best customers (highest NRR, highest expansion, longest retention)?
• Which channels are generating customers that churn fastest or have lowest NRR?
• Is there a segment mismatch — channels that reach an audience that isn't your ICP, producing pipeline that sales closes but CS later regrets?

Hidden Issues — the things attribution won't surface:

• Are there channels where sales team ignores the leads? (High MQL volume, low SQL conversion — find the specific channel)
• Are there segments where demos book but deals don't close — a product-market fit problem the marketing numbers are hiding?
• Are there geographic or vertical pockets where performance is strong but spend is low?
• What does the sales team say about lead quality by channel — and does it match the data?

Benchmark Comparison

• Blended CAC vs industry benchmarks for your segment (SaaS, e-commerce, B2B services)
• MQL-to-SQL conversion rate vs benchmark
• Marketing-sourced pipeline as percentage of total pipeline vs benchmark
• Content/SEO: organic traffic to pipeline conversion vs peers

Reallocation Recommendation

• Which channel to increase (highest pipeline quality, untapped capacity)
• Which channel to decrease (low pipeline quality, high spend relative to outcomes)
• Which channel to cut (no traceable pipeline and no brand-building rationale)
• The one experiment to run next quarter (a channel or format you haven't tried with your ICP)

Rules

1. Every spend category must produce a CAC number or a documented reason why it can't (brand spend requires different accountability).
2. Win rate and deal size are as important as CAC. A 2x CAC channel with 1.5x deal size and 1.3x win rate is your best channel.
3. The sales team's opinion of lead quality is data — collect it systematically, not anecdotally.
4. Benchmarks are reference points, not targets. Your benchmark is your best-performing channel, not the industry median.
5. Attribution models should be used to see what's unmeasurable, not to replace pipeline tracing.
6. Segment fit issues are not marketing problems — they're a product-market fit conversation. Name them as such.

The output of this skill is a reallocation brief: the actual CAC by channel, the 2 channels to increase and 1 to cut, and the segment fit issue (if one exists) that the marketing data is revealing about the product.