Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →rampstackco/paid-media-strategy
A discipline for running paid media that does not light money on fire. Hypothesis writing for paid spend, channel selection, budget allocation, audience targeting, bid strategy, campaign types, what NOT to spend on, attribution reality, and the failure modes that produce expensive lessons.
Passed 20 of 22 test cases with the skill, and 17 of 22 without it.
3 fixed17 passed either way2 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 3 test cases the skill fixed. Same model, same prompt, with and without the skill.
We have a $30,000 monthly growth budget for a DTC home goods brand launching paid ads for the first time. The team proposes splitting the budget equally ($10,000 each) across Google Search, Meta Ads, and Pinterest Ads to see which performs best simultaneously. How should this budget be structured instead?
| Checked | Without skill | With skill |
|---|---|---|
| Rejects an equal budget split across three unproven channels. | Passed | Passed |
| Recommends allocating 60% to 70% of the budget to a single primary channel. | Missed | Passed |
The instructions your model reads when the skill loads.
A senior performance marketer's playbook for running paid media that produces real outcomes.
The default state of paid media is wasted spend. Most accounts have campaigns running because they always have, audiences targeting because the rep suggested it, bid strategies on auto because manual is hard, creative not refreshed because there is no system. The cost compounds. A 20% efficiency gain on a $500K-per-year account is $100K back to the business. A 50% gain on a $5M-per-year account is $2.5M.
This skill is the discipline that produces those gains. It assumes you have a paid media platform (Google Ads, Meta, LinkedIn, TikTok, or aggregators like Synter) connected. It assumes you have working analytics and conversion tracking. The hard part is the strategic discipline behind the spend, and that is what is here.
When to use this skill: any time you are designing a paid media plan, evaluating whether to scale or kill a campaign, allocating budget across channels, or auditing an existing account.
This skill spans paid media strategy and operations. It does not cover ad creative production (use ads-creative-development), result interpretation in depth (use ads-performance-analytics), or platform-specific MCP tooling (consult each ad platform's official documentation for current MCP setup, auth, and example prompts).
The audience is a performance marketer (in-house or agency), a growth lead allocating spend across channels, or a founder making early paid budget decisions. The voice is tactical. There is no "evaluate every option yourself with no opinion." Paid media decisions have shape, and a senior practitioner can map a situation to a defensible plan in an afternoon.
Most paid media failures start with a vague reason for spending. A real spend hypothesis has five parts: audience, offer, channel, outcome metric, and magnitude. Missing any of them and the campaign cannot be evaluated honestly.
A bad reason: "We need to scale Google Ads spend." No audience, no outcome metric, no magnitude. Nothing is falsifiable.
A good hypothesis: "Top-of-funnel SaaS prospects searching for project management tools convert from a free-trial CTA at 3.2% CAC under $80. Increasing Search budget from $40K to $80K per month should hold CAC under $80 and add roughly 500 trial signups based on Q3 search volume."
That hypothesis names the audience (top-of-funnel SaaS prospects on PM-tool keywords), the offer (free trial), the channel (Google Search), the outcome metric (CAC, trial signups), and the magnitude (500 signups, CAC under $80). It is falsifiable: if CAC blows past $80 or signups come in below 250, the hypothesis is wrong and you pull back.
Pre-commit the falsification rule. Decide before scale: at what CAC do we hold? At what CAC do we pull back? At what trial-signup count do we kill? Without pre-commit, every result becomes a debate. With pre-commit, the decision is mechanical.
Primary metric is the one you are optimizing for (CAC, ROAS, CPL). Guardrails are the metrics you do not want to break (LTV, retention, brand search lift). Scaling the primary metric while breaking a guardrail is a Pyrrhic win.
Pick the channel where intent matches your offer. Run the wrong channel and your CAC reads as a channel problem when it is actually a fit problem.
test cases passed, vs 17 / 22 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.