Sample report

This is what you get

A real evaluation output — the same structure your ideas will receive. Every verdict comes with the full reasoning, not just a score.

Feature evaluation

AI smart reply

Acme Commerce · ecommerce platform for SMBs · ~$2M ARR

SHIP
8.4/ 10
High confidence

Idea submitted

"Add AI-generated reply suggestions to our merchant support inbox. When a customer sends a message, surface 3 draft replies the merchant can send with one click. Trained on their past reply history."

Bull Case — Idea Champion

  • +

    Addresses #1 pain in ICP interviews

    Support time ranked top-3 friction in 6 of 8 merchant interviews conducted Q1.

  • +

    No direct competitor at SMB price tier

    Gorgias and Zendesk AI both require $300+/mo plans. Acme's ICP is $49–$149/mo.

  • +

    Search volume rising — 3yr CAGR 42%

    'AI customer reply' and related queries up significantly; no sign of plateau.

  • +

    Estimated 2-week build

    OpenAI fine-tuning API reduces custom training complexity. Low integration risk.

Bear Case — Devil's Advocate

  • Shopify announced similar feature at Connect 2025

    Launches in GA estimated Q3. If Acme ships after GA, differentiation collapses for Shopify merchants.

  • Power users may prefer manual control

    High-volume merchants often have brand voice guides. Auto-suggestions risk off-brand replies.

  • Third-party API introduces SLA dependency

    OpenAI's 99.5% SLA means ~3.6hrs downtime/month. Support feature with known outages is a trust risk.

Ship only if

1

You can ship before Q3 — specifically before Shopify's feature lands in GA for your merchant tier.

2

Engineering confirms the OpenAI fine-tuning integration is feasible within the 2-week estimate during sprint planning.

3

You add a manual override — let merchants edit or dismiss suggestions. This preserves brand voice and reduces the trust risk.

Untested assumptions

HIGH

Merchants will discover the feature without onboarding prompts — activation depends on feature visibility in the inbox UI.

MEDIUM

OpenAI rate limits won't affect p95 reply latency at scale — needs load testing before launch.

MEDIUM

Past reply history is sufficient training data for most merchants — new merchants with <6 months of data may see degraded suggestion quality.

Score breakdown

Market demand
9.1
Competitive differentiation
7.8
ICP fit
9.4
Build complexity
8
Strategic risk
6.8
Revenue potential
8.5

Run this on your own ideas

Your first evaluation is free. Enter your company URL, submit an idea, and get a verdict like this in under five minutes.