Media Planner

Test a budget change before it runs. Priced against your forecast.

Sam tests channel budget changes against Felix's forecast before a dollar moves, and every spend increase is checked against your break-even before it runs.

How Sam works
Leads withThe options as tested scenarios, each with the constraint it was tested under.
NeverRecommend a single path without saying what it was tested against.
Why this exists

Strategy by trial and error. Paid for in real budget.

Tests cost real money to run.

A new channel mix takes $30K to $50K to validate. Most brands can only afford two or three tests per quarter, so most strategies never get tested at all.

Cycles take weeks, not seconds.

Shift budget on Monday, wait two to three weeks for signal, analyze, decide. By the time the answer arrives, the market has already moved.

One variable at a time.

A team can realistically test one budget change at a time. Combinatorial questions like Meta up, Google down, TikTok added stay in the spreadsheet, never in the wild.

Cross-channel effects stay invisible.

What happens to Google when Meta shifts? What's the TikTok halo on organic? Channel interactions never show up in a single-axis test.

What Sam ships

Five capabilities. One scenario workflow.

Each capability maps to a specific job: price the plan, read the measured response, check the break-even, score the split, and stay inside your limits.

Budget plans against the forecast

You set each channel's change, and Sam prices the plan against Felix's forecast range.

  • Change several channels in one plan
  • Each plan carries the forecast range, shifted by the measured response where a channel has one
  • A changed channel with no measured response says so, never a default

Example. Example. Meta down and TikTok up, priced against next month's forecast

Measured response

Where a model run or one of Parker's lift tests has measured a channel, Sam shows where today's spend sits.

  • Response read from a passing model run or a lift test Parker completed
  • A channel with no measurement says so; no default elasticity is used
  • Plans stay inside the measured spend range

Example. Example. A channel with a completed lift test shows its measured return beside today's spend

Risk-free what-if

Every strategy gets simulated before commitment. Upside, downside, and break-even priced before any budget moves.

  • Low and high ends of the range shown
  • Break-even from Max's Profit page, once your margin is known
  • $0 spent before validation

Example. Example. A plan whose return falls below break-even reads Below break-even before anything is sent

Scored recommendations

The best split for a fixed budget, each channel stopped at break-even, within your Limits caps.

  • Chance of reaching the revenue target, where a target is set
  • Each figure names the curve and forecast behind it
  • Risk and reward shown side by side

Example. Example. The chance a plan reaches this month's revenue target, shown beside the plan

Agent-aware

Reads Felix's forecast, measured response from model runs and lift tests, and your Limits caps.

  • Scenarios start from Felix's forecast range
  • Limits caps bound every plan
  • Lift tests from Parker count as measured response

Example. Example. A plan over a Limits cap is held to the cap before it is priced

A typical week

What Sam runs while the team is in standup.

One workweek, hour by hour. Specific numbers vary by account.

Mon · 8:42 AM
Tested

A strategist moves budget from Meta to TikTok in the scenario builder.

Mon · 8:43 AM
Scored

The plan is priced against Felix's revenue and spend forecast, with a break-even verdict.

Mon · 8:44 AM
Compared

Sam shows the best split for the same budget alongside it, where the channels' response is measured.

Mon · 9:15 AM
Sent

The strategist files it as proposals in Approvals.

Tue · 11:02 AM
Approved

Approved changes go live; the prediction is stored so Measurement can check it.

What's underneath

Measured response. Not a recommendation wrapper.

Sam prices plans on measured channel response and Felix's simulated forecast range. Every scenario starts from the live forecast, not stale benchmarks.

Reads from

Felix's nightly forecast, plus model runs and lift tests. Sam never prices a plan against industry averages.

Felix forecasts + measured channel response

Reports to

Plans go to Approvals as proposals for Meta, or to Home as advice for Google and TikTok.

Nothing moves until a proposal is approved

Collaborates with

Every plan is checked against your Limits and the profit guardrail before it is sent. Lift tests from Parker count as measured response.

Shared memory, not message-passing

Built on

Measured channel response plus Felix's calibrated forecast range.

Response from model runs and lift tests

Ready when you are

See Sam test a budget change on your account, or talk to us about your stack.

The pilot connects your accounts to every agent. Bring a what-if question; Sam prices it against your forecast and break-even.

Looking for something specific? Felix forecasts the outcome →, Maya remembers the brand's decisions → or Parker sets up lift tests →.