agentleFS
kastriasanikastriasani/google-ads-skills/skills/google-ads-forecast/SKILL.md

google-ads-forecast

Skill1 starsChanged 5 months agoMITView on GitHub

Forecast Google Ads budget, clicks, conversions, revenue, ROAS and CPA with best, likely and worst scenarios before you commit spend. Use this whenever someone asks to forecast or project a Google Ads plan, model a budget, estimate results or ROAS, size the spend needed to hit a conversion or revenue goal, build a best/likely/worst or scenario forecast, check whether a plan clears break-even ROAS, run a what-if on CPC or conversion rate, plan a quarterly or seasonal budget, or decide where the next budget dollar should go across campaigns. Triggers include forecast, projection, budget plan, media budget, scenario planning, ROAS forecast, CPA target, results estimate, what-if, ramp, and goal coverage. Runs standalone on pasted numbers or a CSV; optional Google Ads API, Keyword Planner, GA4 and Sheets make it sharper.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 2 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Check whether a plan clears break-even ROAS at a given margin.

    line 24
  2. Distinctive:

    Run a what-if on CPC, CTR, conversion rate, or value per conversion.

    line 25
  3. Very specific:

    If the request is to recommend a bid strategy and targets, use google-ads-bidding.

    line 29
  4. Very specific:

    If it is to design an experiment, use google-ads-ab-testing.

    line 30
  5. Very specific:

    Forecast the funnel and scenarios: Run scripts/budget_forecaster.py --budget 25000 --cpc 1.80 --ctr 0.05 --cvr 0.04 --value 120 --margin 35.

    line 39
  6. Very specific:

    Forecast a multi-line plan with blended rates: When the plan spans several campaigns or campaign types and you have history, run …

    line 46
  7. Very specific:

    Check unit economics: Run scripts/roas_calculator.py --spend ...

    line 59

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Google Ads Forecast

Turn a Google Ads plan into a defensible set of numbers: impressions, clicks, conversions, revenue, ROAS, CPA and profit, each as a best / likely / worst scenario with the assumptions on the table. A forecast that hides its assumptions is theatre, so every script here prints them.

When to use

Reach for this skill when the user wants to:

  • Forecast or project results for a new or changed Google Ads plan.
  • Size the budget needed to hit a target number of conversions or a revenue goal.
  • Produce best/likely/worst (scenario) numbers, not a single optimistic figure.
  • Check whether a plan clears break-even ROAS at a given margin.
  • Run a what-if on CPC, CTR, conversion rate, or value per conversion.
  • Plan a quarterly or seasonal budget, including the Smart Bidding learning ramp.
  • Decide where the next budget dollar should go across existing campaigns.

If the request is to recommend a bid strategy and targets, use google-ads-bidding. If it is to design an experiment, use google-ads-ab-testing.

Workflow

  1. Gather inputs. You need a budget (or a conversion/revenue goal) plus four rates: expected CPC, CTR, conversion rate (CVR) and value per conversion. Pull them from account history first, then Keyword Planner / GA4, then the bands in references/benchmarks.md. Confirm the gross margin so break-even ROAS is real.
  2. Forecast the funnel and scenarios. Run scripts/budget_forecaster.py --budget 25000 --cpc 1.80 --ctr 0.05 --cvr 0.04 --value 120 --margin 35. It derives impressions, clicks, conversions, revenue, ROAS and CPA, produces best/likely/worst by flexing CVR and CPC, and adds a multi-month ramp that discounts the learning period plus optional --seasonal multipliers. To solve for spend from a goal, pass --target-conversions instead of --budget.
  3. Forecast a multi-line plan with blended rates. When the plan spans several campaigns or campaign types and you have history, run scripts/results_forecaster.py --input plan.json (or --demo). It blends recent and long-run CVR per line (70/30 toward recent), rolls up a 3-tier account forecast, computes goal coverage, and flags high-variance plans. See references/forecast-method.md for the blend and scenario logic and why.
  4. Decide where the next dollar goes. For an existing account, run scripts/channel_roi_analyzer.py --input roi.json to score each campaign on cash ROI, LTV ROI and marginal ROI, and get a DOUBLE-DOWN / MAINTAIN / DEFUND / EXIT verdict plus the inflection spend - the point where the next dollar stops paying for itself. This is how budget reallocation gets justified.
  5. Check unit economics. Run scripts/roas_calculator.py --spend ... --revenue ... --conversions ... --margin ... on any single campaign to confirm ROAS against break-even and surface CPA, CPC, CTR and CVR with recommendations. Use it to validate the forecast against reality after the first weeks of spend.
  6. Present with assumptions and a gap path. State the likely number, show the worst/best band, name the data window and variance used, and if coverage is below 100% give the gap-closing path (CVR, then budget, then goal) from references/forecast-method.md.

Inputs and data sources

Runs standalone: paste the numbers or hand the scripts a small JSON/CSV. Each script has a --demo mode and a --json flag for piping into a report or sheet.

Optional sources that improve accuracy:

  • Google Ads API / UI - historical CPC, CTR, CVR by campaign, match type and device.
  • Keyword Planner - auction-aware impression, click and CPC forecasts for a keyword set.
  • GA4 - value per conversion (AOV), assisted conversions, post-click behaviour.
  • Google Search Console - organic query and seasonality context for demand.
  • Google Sheets / BigQuery - store plan inputs and pipe --json output into a living forecast model.

Outputs

  • A funnel forecast (impressions, clicks, conversions, revenue, ROAS, CPA, profit).
  • Best / likely / worst scenarios with the variance band disclosed.
  • A monthly ramp accounting for the learning period and seasonality.
  • A 3-tier multi-line forecast with blended conversion rates and goal coverage.
  • Per-campaign reinvestment verdicts (DOUBLE-DOWN / MAINTAIN / DEFUND / EXIT) and inflection spend.
  • A non-optional assumption block to present alongside every number.

References

  • references/forecast-method.md - the funnel identity, scenario logic, blended conversion, learning ramp, coverage and gap analysis.
  • references/benchmarks.md - Google Search / Shopping / PMax / Display benchmark bands and seasonal factors, marked as planning estimates.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.