Briefing — Portfolio Mode

Growth target

How much do you want to grow new customers? Defaults to YoY (annualized) — that's how operators actually frame growth. MoM at 20% would be ~9× annual, rarely realistic. Math accounts for the fact that new winners deliver only ~65% of their lifetime monthly avg in month 1 (learning phase + ramp).

+ %

Enter a growth target above to see the recommended spend + brief volume…

Growth basis

Risk profilei

Proven / data-backed Unproven / exploratory

55% proven · 45% unproven moderate

Calendar state

Compliance ⓘ

Show the math

Loading… —

Advanced math overrides

Target CPA ($)

Confidence target (%)

Target ROAS

Variants per concept

Month-1 ramp factor (0.1–1.0) Fraction of a winner's lifetime monthly avg it delivers in month 1. Default 0.65 (median first-30-day / first-60-day per industry data).

Spread recommendation

Lane mix is the venture-portfolio layer (which lanes get budget); the dials below are the within-lane spread. Angle, ad type, and execution style get chosen by the planner based on lane + format. Offer mix is opt-in — leave closed for evergreen briefs.

Lane mix

100%

concept briefs in this batch

— × 3 variants = — total ads Reset to recommendation

Tell me what you want.

Pour in raw thoughts, fragments, references — typed, dictated, or dragged in. I'll go generate an intelligent batch from the same evidence pack you'd get in Portfolio Mode, anchored to whatever you put in here.

🎙 Brainstorm with GPT (voice)

Live voice conversation with GPT. It pulls real reviews, ad performance, research findings — and suggests where to look next using the 4Cs framework. End the call and the conversation becomes your briefing seed.

Realtime diagnostics

Drag images / screenshots here or Cmd-V paste · max 10 MB each · png/jpg/webp/gif

Default: 1 family with 8 variants — for a single tight idea you want explored from multiple angles. Bump Families up when your seed contains several independent ideas.

Concept families

Opus drafted these independent exploit-weighted families across the full evidence pack. Approve the ones you want to keep, reject the rest, edit any fields inline, or regenerate individual families. When you're done, click "Generate briefs from approved" to fire Stage 2.

Cluster by

0 / 0 approved