Why the Invoice Method
Almost everyone agrees that AI saves them time. Almost no one can tell you what that time is worth. Return on investment is one of the hardest things to pin down with AI, because the value shows up as a hundred small moments instead of a single line on a budget. The Invoice Method gives you a baseline. It turns the work your AI actually produced into a conservative dollar figure, so you can finally answer the question your CFO, your board, and your own planning keep asking: what is this worth?
The trouble is that AI never sends you an invoice. A consultant does. A new hire shows up on a payroll line. AI just quietly produces a memo, a model, a deck, an analysis, and moves on. The value is real, but it is what you did not have to pay someone else to do, spread across dozens of tasks a week. That is the kind of value that is easy to feel and hard to measure, which is why most conversations about AI ROI end in anecdotes instead of numbers.
- It is counterfactual. The cost you avoided never appears anywhere, so there is nothing to point to.
- It is diffuse. A few minutes saved here and there does not add up to anything you can see.
- It is invisible. No invoice, no timesheet, and no line item is ever created.
- It is uneven. A quick answer and a board-ready analysis are worth very different amounts.
The Invoice Method solves this with one simple move. It prices each piece of AI-assisted work at what a qualified human would have charged to produce it, then sets that against the near-zero cost of the AI. The human-equivalent figure is your baseline. Do it once and you have a number for today. Do it continuously and that baseline grows, so you can watch your return climb week over week instead of guessing at it.
What it gives you
- A baseline for your AI ROI, where before you had only a hunch.
- A number that is defensible, built on conservative mid-market rates and the low end of the hours.
- A way to track the return over time, so you can show progress, not just a one-time claim.
- Something leadership can act on, a real input for budgets, staffing, and the case for wider adoption.
- A method that works for any role, because every job produces work a professional would otherwise be paid for.
It moves you from "I think AI saves me time" to "here is the measured, conservative return on every hour my AI works."
How it works, start to finish
Setting this up is genuinely simple. You paste one prompt, answer a few quick questions, approve a folder, and the assistant builds your tracker and puts it on autopilot. Here is everything you need and the path from start to finish.
What you need
- The Claude desktop app with Cowork mode, the mode that lets the assistant work with files on your computer and run tasks for you.
- A folder on your computer for your files. Your Documents folder is fine.
- About 20 to 30 minutes, most of it hands-off while the assistant works.
- No coding and no spreadsheets. You only ever type plain English.
super prompt
quick questions
folder
everything
you're done
Easier for you, and no harder for the assistant. The only parts that truly need a human are two quick clicks: approving the folder, and starting the task once. The prompt pauses and walks you through both. Everything else is automatic, and you can change anything later just by asking.
The super prompt
Copy everything in the box below and paste it into a new conversation in Cowork. Do not change anything. The assistant will take it from there.
You are setting up The Invoice Method for me, start to finish, in this one session. It measures the value of my AI-assisted work by pricing each deliverable at what a human professional would have charged, then comparing that to the near-zero cost of the AI. Work conservatively so every number is defensible. Follow these steps in order.
STEP 1 - ASK ME THESE QUESTIONS FIRST, THEN WAIT FOR MY ANSWERS.
Offer the default in brackets so I can simply reply "use the defaults":
1. What should we name the folder where everything is saved?
[a new folder called AI Value in my Documents]
2. Use your standard conservative rate card, or my own rates?
[use the standard conservative rates below]
3. What time should the daily update run? [6:00 PM]
4. What counts as work? [real deliverables only; skip quick questions, setup, and personal chats]
5. How far back should we look for past work? [all my past conversations]
STEP 2 - CONFIRM AND CONNECT.
Summarize the plan in two or three lines. Then ask me to connect the folder from step 1 so you can save files there, and wait for me to approve access.
STEP 3 - BUILD MY CUMULATIVE TRACKER in that folder:
a. Look back through my past conversations and list the real deliverables.
b. Price each one using the rate card and rules below. Use the low end of hours. Count each deliverable once. Skip anything that is not real work.
c. Create a ledger file as the single source of truth, and keep a list of every conversation already counted so nothing is ever counted twice.
d. Create a dashboard I can open in any browser showing: net value generated, return on AI cost, consultant-hours saved, total AI cost, a breakdown by type of work, and a daily list.
e. Label any older work you estimate as an estimate.
f. Show me the headline totals when you finish.
STEP 4 - PUT IT ON AUTOPILOT.
Create a scheduled task that runs every day at the time I chose. Each day it must: find the new work not yet counted, value it the same conservative way, add it to the ledger, update the running totals, refresh the dashboard, and send me a short end-of-day summary with what we did, today's invoice amount, and my new cumulative total and return on AI cost. Make the task fully self-contained and make sure it never counts the same conversation twice.
STEP 5 - HAND IT BACK TO ME.
Tell me the one thing to do by hand: open the Scheduled panel and click Run now on the new task once, so it approves the tools it needs and future runs do not pause for permission. Remind me tasks run while the app is open.
RATE CARD - conservative mid-market independent rates, by category. Match each deliverable to the closest role below. Use my own rates instead if I gave you any.
STRATEGY & EXECUTIVE
Executive / strategy advisor ............ $325/hr
Senior management consultant ............ $275/hr
Management consultant (mid-level) ....... $200/hr
Org design / change management .......... $250/hr
HR & PEOPLE
Senior HR / CHRO-level advisor .......... $250/hr
HR consultant (mid-level) ............... $175/hr
Compensation & benefits analyst ......... $175/hr
Employment law / compliance advisor ..... $275/hr
Talent acquisition strategist ........... $175/hr
Learning & development designer ......... $150/hr
HR systems / HRIS consultant ............ $225/hr
FINANCE & ACCOUNTING
Fractional CFO / finance advisor ........ $300/hr
Financial analyst / modeler ............. $175/hr
FP&A / budget analyst ................... $175/hr
Bookkeeper / staff accountant ........... $125/hr
LEGAL & COMPLIANCE (non-attorney advisory)
Regulatory / compliance advisor ......... $275/hr
Contract / paralegal specialist ......... $150/hr
(Licensed-attorney work is $350+/hr; use only if the deliverable truly required an attorney.)
MARKETING & COMMUNICATIONS
Marketing strategist .................... $200/hr
Executive communications ................ $175/hr
Content strategist / writer ............. $150/hr
SEO / digital marketing specialist ...... $150/hr
Brand / creative designer ............... $150/hr
RESEARCH & ANALYSIS
Industry analyst (specialized) .......... $200/hr
Senior research analyst ................. $150/hr
Data analyst ............................ $125/hr
TECHNOLOGY & OPERATIONS
Solutions architect ..................... $250/hr
Process / operations consultant ......... $200/hr
Project manager ......................... $175/hr
Business analyst ........................ $150/hr
Dashboard / BI developer ................ $150/hr
CREATIVE & DESIGN
Presentation designer ................... $125/hr
Graphic / visual designer ............... $125/hr
HOW TO CHOOSE A RATE:
- Match the deliverable to the closest role above.
- Use a junior or routine rate (about $100-125) for templated work; the senior rate only when the task needed real expertise or executive judgment.
- When a deliverable spans roles, blend the rates and lean toward the lower-cost role to stay conservative.
- These are independent rates. Brand-name firms (Big Four, MBB) charge 2 to 4 times more, so this is a floor.
HOURS BY COMPLEXITY (use the low end):
Simple (a standard email, a small edit) ............. 1 to 2 hours
Moderate (a memo with analysis, a short post) ....... 2 to 4 hours
Complex (a briefing deck, a research summary) ....... 4 to 8 hours
Highly complex (an enterprise strategy, a policy) ... 8 to 16 hours
THE RULES:
- Use mid-market rates, not premium firm rates.
- Estimate hours at the low end and round down.
- Count each deliverable once, not once per draft.
- Only count real work. Quick questions and setup are worth zero.
- Price the AI generously, about 50 cents per session, so the comparison cannot be challenged.
- Be transparent about estimates. A number I can defend to my CFO is the goal.
What happens after you paste it
- It asks you five quick questions (folder name, rates, daily time, what counts, how far back). To keep it simple, just reply use the defaults.
- It confirms the plan and asks you to connect a folder. Approve access when prompted. This is where your files will live.
- It reviews your past conversations and builds your tracker, valuing each deliverable conservatively, and shows you the headline totals.
- It sets up the daily task at the time you chose, so the tracker updates itself every evening.
- It asks you to click Run now once. Open the Scheduled panel, click Run now on the new task, and you are finished.
1. Approve folder access when the assistant asks for it.
2. Click Run now on the task once after setup, so future runs do not pause for permission. Scheduled tasks run while the app is open, so if it is closed at the scheduled time, the task runs the next time you open it.
The questions it will ask you
You can answer all five with use the defaults. Here is what each one means, in case you want to change it.
| It asks | Easy default | Change it if |
|---|---|---|
| Folder name | A new AI Value folder in Documents | You want it in a specific or shared location. |
| Which rates | Standard conservative rates | You know your market's consulting rates and prefer them. |
| Daily time | 6:00 PM | You would rather get your summary at another time. |
| What counts | Real work only | You want to include or exclude certain conversations. |
| How far back | All past conversations | You would rather start fresh from today. |
Reading your dashboard
Open the dashboard file from your folder in any web browser. Four numbers tell the story.
| Number | What it tells you |
|---|---|
| Net value generated | Your total human-equivalent value, minus the small AI cost. The headline. |
| Return on AI cost | How many times over the value exceeds what the AI cost you. |
| Consultant-hours saved | The professional hours the work would have taken a human. |
| Total AI cost | What the AI actually cost. Usually a rounding error by comparison. |
Below the headline, the dashboard breaks your value down by type of work and lists each day. It is a single file, so you can open it on any computer, print it, or hand it to a leader as is.
The rate card, in full
You never manage the rate card by hand, it is built into the prompt. Here it is in full, so you can see exactly how your numbers are built and swap in your own market rates if you prefer. These are conservative independent-consultant rates; brand-name firms charge two to four times more, so this sits at the floor of the real market.
| Category | Role | Rate |
|---|---|---|
| Strategy & executive | Executive / strategy advisor | $325/hr |
| Senior management consultant | $275/hr | |
| Management consultant (mid-level) | $200/hr | |
| Org design / change management | $250/hr | |
| HR & people | Senior HR / CHRO-level advisor | $250/hr |
| HR consultant (mid-level) | $175/hr | |
| Compensation & benefits analyst | $175/hr | |
| Employment law / compliance advisor | $275/hr | |
| Talent acquisition strategist | $175/hr | |
| Learning & development designer | $150/hr | |
| HR systems / HRIS consultant | $225/hr | |
| Finance & accounting | Fractional CFO / finance advisor | $300/hr |
| Financial analyst / modeler | $175/hr | |
| FP&A / budget analyst | $175/hr | |
| Bookkeeper / staff accountant | $125/hr | |
| Legal & compliance | Regulatory / compliance advisor (non-attorney) | $275/hr |
| Contract / paralegal specialist | $150/hr | |
| Marketing & comms | Marketing strategist | $200/hr |
| Executive communications | $175/hr | |
| Content strategist / writer | $150/hr | |
| SEO / digital specialist | $150/hr | |
| Brand / creative designer | $150/hr | |
| Research & analysis | Industry analyst (specialized) | $200/hr |
| Senior research analyst | $150/hr | |
| Data analyst | $125/hr | |
| Technology & operations | Solutions architect | $250/hr |
| Process / operations consultant | $200/hr | |
| Project manager | $175/hr | |
| Business analyst | $150/hr | |
| Dashboard / BI developer | $150/hr | |
| Creative & design | Presentation designer | $125/hr |
| Graphic / visual designer | $125/hr |
How a rate gets chosen
- Match each deliverable to the closest role, then take that rate.
- Use a junior or routine rate, about 100 to 125 dollars, for templated work. Reserve the senior rate for work that needed real expertise or judgment.
- When a deliverable spans several roles, blend the rates and lean toward the lower one.
- For anything that truly required a licensed professional, such as an attorney, CPA, or actuary, use 350 dollars or more and note it.
- Swap in your own market rates anytime by asking. The structure stays the same.
Make it your own
Everything is adjustable after setup. You change it the same way you built it, by asking in plain language.
Update my value tracker to use these rates instead, and recalculate my totals: [list your roles and the hourly rates you want].
Change my daily value task to run at 5:00 PM, and stop counting personal or test conversations from now on.
Add a monthly summary I can print, and group my dashboard by project.
Questions and troubleshooting
| If this happens | Do this |
|---|---|
| It paused and asked for permission. | That is normal the first time. Approve it, and click Run now on the task once so future runs do not pause. |
| The task did not run at the set time. | The app must be open for scheduled tasks to run. Open it and the task runs. |
| A number looks double-counted. | Ask: please check the list of conversations you have already counted and remove any duplicates. |
| The total feels too high or too low. | Ask for the conservative version, or give it your own rates. You control the assumptions. |
| I am brand new with no history. | Reply that there is nothing to look back on yet. The tracker starts today and grows from here. |
| Is this only for HR? | No. Legal, finance, marketing, operations, sales, and founders use the same method. |
| Can other people see my data? | Everything stays in the folder on your own computer unless you choose to share the file. |
A note on using this well
The point is not to brag. It is to make a quiet daily reality visible: that one person with an AI assistant is now producing work that used to take a team and a budget. That changes how you plan, how you staff, and how you talk about value.
So under-count on purpose. Round down. Label your estimates. Build a number you would be comfortable putting in front of your most skeptical finance partner. When the number is honest, it does the arguing for you.
Price each piece of AI-assisted work at what a human would have charged, set it against the near-zero cost of the AI, and let the gap tell the story, conservatively, every single day.