The pain

You meant to do a weekly review every Friday. You did one in January. Maybe one in March. The reason isn't laziness — it's that a real review requires thinking out loud with someone who knows the context. Solo operators don't have that someone. So Friday afternoon becomes "clean the inbox and call it a week."

Three months pass. You realize you've been working hard on the wrong thing.

Who has it

Anyone running a solo business doing meaningful revenue ($100k+/yr). The bottleneck isn't time — it's the absence of a thinking partner. Coaches, consultants, agency-of-one operators, indie SaaS.

The fix

A Claude Project loaded with your business context — goals, current quarter, recent retros, KPIs — that you talk to every Friday. Not a prompt. A partner you've been training.

Setup (90 min, once)

  • Build the context document. One file. Sections:
    • - Business model (offers, prices, current MRR/revenue, biggest customer types) - Current quarter goals (specific, with numbers) - Known bottlenecks (the two things you suspect are most blocking you) - Decisions in flight (the open questions you haven't resolved) - Voice notes — how you make decisions, what you tend to over/underweight
  • Drop it into a Claude Project called "Friday review."
  • Build the review prompt:

    • > You are my weekly review partner. You know my business and my goals (in the context). I'll dump this week's calendar, revenue, key conversations, and what I worked on. Your job is NOT to summarize. Your job is to ask me 3 sharp questions: one about whether this week moved the actual goal, one about the biggest bottleneck I'm avoiding, one about a decision I should make before Monday. End with one observation I might be missing. Never be flattering.

  • Test it once. Tune the prompt until the questions feel uncomfortable in a useful way.

    • Running it (30 min, weekly)

    Friday afternoon. Open the project. Paste the week dump. Read the questions. Answer them in writing. The act of answering is the review.

    What it looks like running

    You paste in your week. The partner replies:

    1. Your goal this quarter is $5k/mo recurring. You added 1 retainer this week ($800/mo). At this pace you hit $4.2k by end of quarter — short. What changes next week to close the gap, or is this goal becoming aspirational rather than active?
    >
    2. Three of your high-energy hours this week went to portfolio site updates. Fair to call that the bottleneck of the week, or was it avoidance of the [enterprise pitch you mentioned in March]?
    >
    3. You've had Karen's contract draft sitting for 11 days. What's the actual block — pricing, scope, or the conversation you don't want to have?
    >
    Observation: Your last three weeks all show heavy creative work and zero outbound. That's a pattern, not a coincidence.

    You write a 200-word response to each. Forty minutes after starting, you have your three Monday actions.

    Why this works

    This is a Partner-tier workflow — the heaviest trust ask: "Can AI think with me about my business?" The leverage is the context document plus the constraint to ask, not summarize. Without those two, you get generic management-consulting bullshit.

    It works because the AI partner doesn't have the conflict of interest a friend, partner, or mentor has. It will not soften feedback to preserve the relationship. It just asks the next obvious question.

    Run this for 12 weeks straight and you will know things about your business you currently don't.