Skip to content
Skills from A Smart Bear

Observations: The raw, unjudged facts about who you actually are

Surface the raw material your ideal-customer definition is built from — what customers praise, complain about, and leave over — through twelve unsparing question categories, recorded vividly and judged later. For a company already operating online, it first scans public reviews and press to seed the walk. Adapted from Chapter 2 of Hidden Multipliers: Carol Loves You for You.

Input
Your company or product — plus your team's write-storm notes, if you ran one, or just yourself and an honest hour.
Output
An OBSERVATIONS.md of numbered, vivid, specific observations (O1, O2, …) organized by the twelve question categories — deliberately unevaluated.

What this is about

You can't define the customer who loves you for who you are until you know — honestly — who you are. But "write down our strengths and weaknesses" is an impossible instruction: nobody knows where to start, everyone rationalizes, and the loudest voice fills the whiteboard. The book's answer is to start a level lower, with raw observations, gathered through twelve detailed, evocative question categories: what customers praise when they choose you despite your foibles, the complaint you have no defense against, what you're proud of whether customers care or not, what distinguishes your most profitable customers from the rest, what you claim to be great at but aren't, what you envy in competitors, what you'd honor even if it lost you money.

This skill runs that fact-finding as a working session. Bring a team's write-storm output and it processes the notes one observation at a time — clarifying, deduplicating, sharpening; or bring nothing but yourself and it walks you through the twelve categories directly, pressing every vague answer into a specific, vivid fact ("customers post screenshots of support wait times on Reddit," not "support could be better"). When your company already operates and strangers already talk about it online, the skill opens by scanning that public record — reviews, forum threads, press, competitor comparisons — and files what it finds in an External Research section that seeds the walk and sharpens its prompts; those findings stay candidates until you confirm them, and a pre-launch company with no public footprint simply skips the scan. Two rules hold throughout, both from the book: observe like an outside consultant — behaviors and outcomes, never justifications — and don't evaluate, don't blame, don't act. Judging what's a strength or weakness is the next step's job; this one just gets the truth on the record.

Example invocation

You can invoke the skill like this:

/asb-carol-observations We're a 6-person bootstrapped SaaS doing email
deliverability monitoring, 4 years in, ~$50k MRR. I want to figure out
who our ideal customer really is — start with the honest look at us.

The skill walks you through the twelve question categories one at a time, pressing each answer into specific, vivid observations and recording them in OBSERVATIONS.md as you go. You walk away with the raw material the strengths-and-weaknesses step distills next.

One practical note: say where you're working. Name a directory (or point at your existing files) when you invoke, and the skill keeps the method's files together there — if you don't, it asks before creating anything.

From the source

Two sources form the foundation of this skill:

Supporting sources each supply a piece of the mechanism:

  • Write-Storming Workshop (appendix) in Hidden Multipliers — hiddenmultipliers.com — brainstorm alone, process together: the workshop mechanics this skill's team mode follows and its solo mode adapts.
  • Specificity: A weapon of mass effectiveness — why every observation must be specific and vivid: generic words are interpreted differently by every reader and carry no evidence.
  • Fail to face the truth, and you fail — the posture of the whole exercise: an honest look at what is, before deciding what to do about it.