Confession time.
A lot of clients come to me with the same loop — and honestly, I get caught in it too.
It goes like this: you've got your list. You know what you should be working on. Then you see a new tool, a new strategy, a new platform, a new "okay this is the thing that will finally work" — and your brain lights up.
You start it. You spend two days on it. Maybe a week.
Then something else catches your eye. And the half-built thing joins the graveyard of half-built things.
You're busy. You're working constantly. You're learning new stuff every week.
And your business… isn't really moving.
This is shiny object syndrome, and it isn't a discipline problem.
Why it actually happens...
Your nervous system likes novelty. New things give you a dopamine hit. Finishing things is harder — because finishing means the thing has to be judged. Finishing means you have to find out whether it worked. The half-built version is safe. It still could be the thing.
So you keep starting. Because starting feels like progress. And as long as you're starting, you don't have to face the discomfort of finishing.
That's not laziness. That's protection.
🛑 The old way
Too many productivity gurus tell you to "just focus." Discipline. Time-blocking. A bigger to-do list with prettier columns. None of that addresses why you keep wandering off in the first place.
You don't need more discipline. You need a filter.
✨ The new way
Here's what I've been doing — and what I'm now recommending to clients.
Every Monday morning, before I open Slack, before I touch email, before I look at the inbox of things people want from me — I sit with Claude and ask one question:
👉 What are the three things this week that will actually move the needle?
But here's the part that makes it work: I don't ask cold. I give Claude my business context first.
I have a file called my business-snapshot.md that lives in my Claude project. And it's not the surface-level "what I sell, who I serve" version — that's not enough for Claude to actually help. Mine captures: - What I sell — offers, prices, delivery format
- Who I serve — my ICP, not in the abstract but specifically
- How many active clients I have right now
- My current monthly revenue range
- What's actually in my pipeline this month (warm leads, conversations in motion, proposals out)
- Which marketing channels are bringing people in, and audience size on each
- My 90-day goal — the specific outcome, not a vibe
- The metric that tells me I'm winning that goal
- What's worked to bring me money before (the actions, not the theories)
- What I've tried that hasn't worked
- The current hurdles I keep running into
- What I keep avoiding
- What's already half-built
- How many hours a week I actually have for business-building (vs. client delivery)
- My budget for tools, help, or contractors right now
It's a long list. But here's why it matters: if Claude only knows what you sell and who you serve, the "top three" you get back is generic advice. If Claude knows you have two active clients, $4K MRR, one warm lead, no audience on LinkedIn, three half-built lead magnets, and eight hours a week — now we're talking about your business.
Then every Monday, I dump my current list into the conversation — everything I'm considering doing this week, including the shiny new thing that just caught my attention.
And I ask:
‼️PROMPT‼️
Given my business context (attached), I want to know which 3 things this week will actually move the needle on my 90-day goal. Before you answer, ask me whatever follow-up questions you need to give me the best answer for my business success — capacity, energy, recent wins or stalls, what's happening with clients this week, anything that would sharpen your recommendation. Then give me your three. Don't be nice — if something on my list is a shiny distraction, name it.
Two things matter in that prompt. "Ask me follow-up questions" stops Claude from guessing. "Don't be nice" stops Claude from hedging. Without those two lines, you get a confident-sounding answer based on incomplete information. Which is exactly what shiny object syndrome runs on.
What I get back is a short conversation, then a short list. Three things. The shiny thing usually gets named as the shiny thing — and sometimes that's a relief. Other times Claude pushes back and tells me actually, this fits, here's why. Either way, I leave that conversation knowing exactly what to work on this week.
Ten to fifteen minutes. That's the whole ritual.
👋 Want to try it this week?
If you don't have a business-snapshot.md file yet, here's the interview to build one. Open a new Claude chat, paste this in, and let it walk you through: ‼️PROMPT‼️
I want to build a my-business.md context file I can save and reuse for weekly planning. The goal is to give you enough information about my actual business — not the abstract version — so you can help me decide what will move the needle each week. Interview me one question at a time. Wait for my answer before moving to the next. If my answer is vague, ask one follow-up to sharpen it. When we're done, give me a clean markdown file I can save.
Cover these areas, in order:
- What I sell — every offer, the price, and how it's delivered
- Who I serve — my ideal client, specifically (not "women entrepreneurs" — which women entrepreneurs)
- Current client load — how many active clients I have right now
- Current monthly revenue range — rough is fine
- Pipeline — warm leads, conversations in motion, proposals out
- Marketing channels — which ones I'm actively using, audience size on each, which ones have actually brought in clients
- My 90-day goal — the specific outcome I want, not a vibe
- The metric that tells me I'm winning that goal
- What's worked before — actions that have brought me money or clients
- What hasn't worked — things I've tried that didn't pay off
- Current hurdles — what's actually in the way right now
- What I keep avoiding — the thing I know I should do but haven't
- Half-built projects — what's started but not finished
- Time — how many hours per week I realistically have for business-building (separate from client delivery)
- Budget — what I can spend on tools, help, or contractors right now
Save the output as business-snapshot.md. Drop it into a Claude project. Update it at the start of every quarter — or sooner if something material changes (new offer, big client signs, a channel stops working). ---END OF PROMPT
Then every Monday, you've got a 10–15 minute filter for whatever new thing tries to grab you.
Shiny object syndrome doesn't fully go away — the dopamine hit is real, the nervous system response is real. But you can build a system that catches it before it costs you another half-built project.
💬 Drop a comment: what's one half-built thing on your list right now that would actually move the needle if you just finished it? Sometimes naming it out loud is the whole unlock. 🩵
P.S. An "md" file is a markdown file that AI tools like ChatGPT and Claude can read really easily without using a ton of tokens. You can read them in a text reader or a free Markdown File editor of your choice (for actual formatted content).