Annual Planning With Claude: How I Built Our Entire 2026 Plan

This one document is running our entire business in 2026 — and I built all of it with Claude.

Here's the problem with how most businesses plan their year: the plan isn't aligned with the actual goals, it isn't built on what worked and what didn't last year, and it completely ignores the current team dynamics. So it becomes a document nobody opens after January.

I did annual planning with Claude across a few focused sessions, with my strategic thinking partner — I named the project Tars — and it was the easiest, most useful planning year we've ever had. It saved me a conservative 20 to 30 hours, and we've already nailed every one of our Q1 rocks.

If you're a coach, consultant, or operator who's been using AI as a glorified search bar, this is the post that shows you the real process. Let's get into it.

Why Architecture Comes First

The architecture is what makes all of this possible. Miss this part and the rest falls apart.

Most people open a regular chat — ChatGPT, Claude, whatever — and start asking. That's not what I did. I built a persistent project: a workspace where the AI retains context across every single conversation inside it. It knows my business, my team, my history, and my numbers (the ones I chose to give it).

Inside that project I configured Claude as my strategic co-pilot with custom instructions tied to that specific role. It pushes back on my bad ideas. It asks discovery questions before it proposes solutions — which is the opposite of what most people get, where the AI just fires off generic answers that aren't tied to your business at all.

The data is where the magic happens. Business model, team assessments, past financials, previous strategic plans, client data — all of it accumulates in that project's context. That's what makes the output useful instead of generic. My project, Tars, had that from day one because I'd been building it for months before I ever sat down to plan. If you haven't built that foundation yet, start there before anything else.

This is the reframe I want you to sit with: if your AI gives you generic planning advice, it's not because you asked wrong. It's because there's no architecture underneath it. That's a structure problem, not a knowledge problem — and you can fix structure on purpose.

"AI Didn't Write My Strategy — It Extracted My Brain"

This is the line I keep coming back to, and it's the whole philosophy in one sentence.

I didn't ask Claude to invent a strategy for me. I had it interview me — pulling from those project files — and then organize what came out of my own head into something that actually made sense. Honestly, it organized it better than I would have.

And I didn't type a word of it. I did the entire thing with talk-to-text, mostly on walks in nature and while walking the dog. That matters, because the excuse most people have ("I don't have time to sit and answer 50 questions") disappears when you realize you can do it out loud while living your life.

Here's the thing though: you still have to be a trained pilot. You'd never board a plane flown by some random guy who just stayed at a Holiday Inn Express last night. But that's exactly how a lot of people use AI — untrained, then surprised when it doesn't land them at the right destination. AI is a powerful tool. You still have to fly the plane.

What Is the 50-Question Business Interview?

It's a structured interview, run in three phases, that gets your real business out of your head and onto the page. I have the full protocol built into SPARC, our AI platform — it's how our clients run this without reverse-engineering it themselves — but here's the architecture so you understand what each phase is doing.

Phase 1 (Q1–20): The Current Reality Audit

This phase doesn't give you answers. It forces you to think clearly and confront the things you've been quietly ignoring. Representative questions:

• What are your revenue streams, and what percentage of revenue comes from each product or service? (Most owners genuinely don't know.)
• For each team member, what are their top two strengths and their biggest gaps?
• What's your actual weekly time allocation — tier-one work versus regular operations?
• Which offer would you kill tomorrow if revenue weren't a factor?
• What process in your business breaks most often? (We turn this into an "issue–solution map.")

For us, this confirmed with real historical data what we'd only suspected anecdotally — that certain offers weren't worth the time we poured into them. I'm not a numbers guy. Having that done quickly and confirmed with data was worth the whole exercise on its own.

Phase 2 (Q21–35): The Gap Analysis

This phase identifies what's actually standing between you and what you want. Examples:

• Where do you want this business to be in 12 months — revenue, team size, your personal role?
• If you could only work on three initiatives this year, what would they be?
• What's the single biggest constraint between where you are and where you want to be?
• What are you currently doing that you should stop doing entirely?

The "only three initiatives" question is the one that reorganized my year. I'm an idea guy — I show up to every meeting with "what if we did this?" So I capped myself at two big ideas a year. It feels limiting. But constraints created more bandwidth, not less. We finished our Q1 rocks early, which gave me more of the big-picture playtime I actually love.

Phase 3 (Q36–50): Blind Spots and Honest Assessment

This is the phase most people would skip — and it's the most impactful. It gets uncomfortable on purpose:

• What scares you most about the year ahead?
• What conversations are you avoiding that you know need to happen?
• Where are you most likely to self-sabotage?
• What does success actually feel like — not the number, the feeling?

A good coach speaks truth with love because they know the opportunity cost of staying quiet. Trained correctly, AI will tell you the truth too — about things you may never have asked yourself, and things even a coach you've worked with never raised. The biggest thing I tell clients: be your best ally, not your worst enemy.

How Do You Use Claude to Align With a Business Partner?

This was the most powerful — and unexpectedly intense — part of the process, and it's a pro tip if you run your business with a partner.

After I finished my 50 questions, I had Claude generate follow-up questions for my wife and business partner, Kim, based on my answers. We packaged them up, handed the context off to her Claude account (hers is named Rosalina), she answered, and sent it back to mine (Tars).

Claude then acted like a neutral moderator — it immediately surfaced where we aligned and flagged where we diverged. Getting out of a business partnership is harder than getting out of a marriage, and the point isn't to go looking for problems. The point is that misalignment you catch in January is a conversation. The same misalignment discovered in June is a six-month detour. This forced the conversation early, on purpose.

From Interview to Operating Plan

At this point Claude had hundreds of pages of raw input — my answers, Kim's answers, where we agreed and where we didn't. Then we synthesized it into the actual 2026 operating plan.

• Revenue architecture. Realistic, stretch, and dream goals — each with the math behind it. Not "six figures," but how many clients at what price point. That makes the number measurable and actionable, and shows you how realistic your stretch goal really is.

• The one thing. Borrowed from Gary Keller's The One Thing — the domino that knocks down the rest. For us, it was getting Kim out of operator mode by building our internal AI operating system in ClickUp. That freed her to focus on marketing, which fed our content engine, which improved our leads. Within a month she was spending about a third of the time in the business she used to.

• Quarterly rocks. Built on Gino Wickman's Traction. We broke the annual plan into focused quarterly chunks on a theme cadence: Foundation → Acceleration → Scale → Optimize. (I'll cover the rock-building process, and how I connected Claude directly to ClickUp with MCP to auto-build and assign tasks, in upcoming posts.)

We also wove in the Prosper Plan — Kim's annual planning framework that our community and clients ask for every year. If you've got an existing planning protocol you love, you layer it in; you don't throw it out.

The bottom line: Claude forced us to focus on what had the biggest downstream impact, which let me put every person on the team in a position to win — and bought me back the bandwidth to do the big-picture work I'm actually good at. That's what a strategic thinking partner does. It didn't hand me a plan. I'm still the pilot. It just helped me think through my own plan more rigorously than I ever would have alone.

Key Takeaways

• Architecture first. A persistent Claude project loaded with your real business data is what makes the output useful instead of generic.
• AI didn't write the strategy — it extracted it. Use it to organize your own thinking, not to invent thinking for you.
• Run the 50-question interview in three phases: Current Reality Audit, Gap Analysis, Blind Spots. Do it with talk-to-text so "no time" stops being an excuse.
• If you have a business partner, fork the answers between accounts and let Claude moderate. Catch misalignment in January, not June.
• Turn the interview into a real plan: revenue math, the one thing, quarterly rocks.
• Constraints create bandwidth. Focusing on three initiatives got us further, faster — and freed up time, not less.
• You still have to be a trained pilot.

Frequently Asked Questions

Can you use Claude for annual business planning?

Yes — and it works best when you treat it as a thinking partner, not a plan generator. Load a persistent Claude project with your business model, financials, team assessments, and prior plans, then have it interview you. Used this way, annual planning with Claude extracts and organizes your own strategy rather than producing a generic one.

What is the 50-question business interview?

It's a structured planning interview run in three phases: a Current Reality Audit (questions 1–20), a Gap Analysis (21–35), and Blind Spots and Honest Assessment (36–50). The first phase doesn't hand you answers — it forces clearer thinking about revenue, team, time, and the things you've been avoiding. The full protocol is built into SPARC for clients.

How is a Claude project different from a normal AI chat?

A regular chat forgets everything between conversations. A Claude project is a persistent workspace that retains context — your business, team, history, and numbers — across every thread inside it, plus custom instructions for how the AI should behave. That accumulated context is what makes its planning output specific to you instead of generic.

How do you use AI to align with a business partner on the plan?

After each partner completes the interview in their own Claude account, you fork the answers between accounts and have the AI generate follow-up questions and act as a neutral moderator. It surfaces where you align and flags where you diverge — letting you resolve misalignment early instead of discovering it mid-year.

Did AI write the whole plan for you?

No. Claude organized and pressure-tested my thinking, but I'm still the pilot. The strategy came out of my head and my business data — AI structured it more rigorously than I would have on my own. That distinction is the whole point: powerful tool, trained pilot.

Want Help Building Your Plan This Way?

Knowing the process is one thing. Building it for your specific business — your data, your team, your constraints — is another.

If you're a coach or service provider running an online business and you want to stop relying on your own memory to plan the year, we can help two ways. You can work with us inside our coaching programs to build your strategic, marketing, and AI-integration systems — or if you just need a one-off to get unstuck, you can book an AI strategy call with our team.

Don't rely on your own brain to hold the whole plan. Build the system.

— Chris