Buy Plan
by
Quarterly buy decisions for ecommerce teams: size each SKU's order against demand, lead time and the cash available, then recommend the quantities.
The quarterly buy is the largest cash decision most ecommerce businesses make, and it is usually made from a demand forecast alone. The forecast says what would sell; it does not say what the business can fund, and the gap between those two is where the trouble starts. Buying to the forecast puts working capital into whatever the model was most confident about, which is rarely the same as what earns most per dollar committed. This skill sizes each SKU against demand, lead time and the cash actually available, then recommends the quantities.
What It Does
- Forecasts demand per SKU from its own history rather than a category curve.
- Ranks SKUs by contribution per dollar of working capital committed.
- Fits the total buy inside the cash available in the quarter it lands.
- Accounts for lead time, so the buy covers the period it will actually serve.
- Flags the SKU whose forecast rests on a single unusual month.
- Recommends the quantity per SKU, with the cash and the cover it produces.
Use Cases
- Size a quarterly buy against the cash that will actually be there.
- Rank SKUs by return on working capital rather than by forecast volume.
- Catch a forecast built on one promotional month.
- See the cover each quantity produces before the order goes out.
Example Session
Build the Q3 buy. We have about $1.4M of working capital for it.
The demand forecast wants $1.82M. Here is the $1.4M version.
- SKU | Forecast qty | Recommended | Cash | Contribution per $ | Cover
- Core tee, black | 12,000 | 12,000 | $108,000 | $2.90 | 91 days
- Jacket, navy | 3,400 | 3,400 | $302,600 | $3.40 | 104 days
- Cap, charcoal | 9,000 | 6,200 | $74,400 | $1.70 | 68 days
- Hoodie, oatmeal | 5,500 | 2,100 | $61,000 | $0.90 | 41 days
These are the four largest lines of the buy. The jacket and the tee are funded in full because they return most per dollar committed. The cap and the hoodie are cut, contributing $132,000 of the $420,000 the whole range has to give back, and the hoodie takes most of that: at $0.90 of contribution per dollar of stock it is the weakest use of working capital you have. Its forecast also leans on one promotional month, which the buy should not repeat.
Why This Beats Prompting It Yourself
A Demand Forecast vs The Cash To Fund It: The Hidden Risk
A generic prompt builds the buy from the forecast, which is how most buys get built. The forecast has no view of what is fundable, so the plan quietly commits more working capital than the quarter has, and the correction arrives as an emergency: a canceled order, a discount to free cash, or a supplier waiting on payment.
When You Still Need Contribution Per Dollar Committed
Ranking by forecast volume funds whatever sells most units, which is usually your cheapest product. Ranking by contribution per dollar of stock is a different order entirely, and it is what decides which SKU absorbs a shortfall. The hoodie at $0.90 per dollar is not a bad product; it is the wrong place for the last $60,000.
Known Limitations
Forecasts built on a short history or on one promotional period are flagged rather than silently used, but the skill cannot know about a demand change you have not recorded. Working capital fitting assumes the payment terms you supply, and a supplier renegotiation changes the answer materially. It recommends the quantities and does not issue purchase orders.
Install It
- Download the ZIP.
It is free and there is no account to create. - Unzip it into your agent's skills folder.
Claude Code reads ~/.claude/skills/, which is hidden by default: the command in the folder block opens it. Other agents scan their own directory, so drop the same folder there instead. - Ask your agent to use it.
Restart the agent if it was already running, then it picks the skill up with no config.
skills folder
Copy
~/.claude/skills/buy-plan/
SKILL.md
references/forecast-notes.md
quarterly/
# one file per buy, with the cash fit
# macOS: create the folder and open it in Finder
mkdir -p ~/.claude/skills && open ~/.claude/skills
# Windows: paste in the Explorer address bar
%USERPROFILE%\.claude\skills\Frequently asked questions
How is this different from buying to the forecast?
A forecast says what would sell. This adds what you can fund and what each dollar of stock returns, then fits the buy inside the cash available. When the two disagree, and they usually do, it says which SKUs absorb the difference and why.
Which agents does the skill run in?
Any agent that supports the open SKILL.md format: Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, Manus, Grok and others load it unmodified. The format is portable, the location is not. Each agent scans its own skills directory, so you drop the same folder into whichever one yours uses. There is no config file to edit and no API key to provision.
What does it need connected to work?
Sales history per SKU, product cost, supplier lead times and payment terms, and your available working capital for the period. Without payment terms it fits against a cash number you supply and labels that as an input rather than a projection.
Why rank on contribution per dollar rather than margin percentage?
Because margin percentage ignores how long the cash is tied up and how much of it each unit consumes. A 60% margin item with a four month cover can be a worse use of working capital than a 40% margin item that turns twice as fast.
What does it do with a forecast it does not trust?
It flags the SKU, says what the forecast rests on, and sizes the buy more conservatively. A forecast driven by one promotional month is the most common case, and repeating that quantity without repeating the promotion is a reliable way to create overstock.
What is included with the skill?
The skill itself, the forecast notes reference, the quarterly buy format, and the working capital fitting rules. It is free and security scanned. We re-publish the ZIP when the model changes, so download it again if your quantities start shifting.
Quarterly buy quantities are a bet made with cash.
Polar reads sell-through, margin and cash position on one definition, so the quantities you commit are sized against what the business can fund.
<script type="application/ld+json">{"@context":"https://schema.org","@type":"FAQPage","mainEntity":[{"@type":"Question","name":"How is this different from buying to the forecast?","acceptedAnswer":{"@type":"Answer","text":"A forecast says what would sell. This adds what you can fund and what each dollar of stock returns, then fits the buy inside the cash available. When the two disagree, and they usually do, it says which SKUs absorb the difference and why."}},{"@type":"Question","name":"Which agents does the skill run in?","acceptedAnswer":{"@type":"Answer","text":"Any agent that supports the open SKILL.md format: Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, Manus, Grok and others load it unmodified. The format is portable, the location is not. Each agent scans its own skills directory, so you drop the same folder into whichever one yours uses. There is no config file to edit and no API key to provision."}},{"@type":"Question","name":"What does it need connected to work?","acceptedAnswer":{"@type":"Answer","text":"Sales history per SKU, product cost, supplier lead times and payment terms, and your available working capital for the period. Without payment terms it fits against a cash number you supply and labels that as an input rather than a projection."}},{"@type":"Question","name":"Why rank on contribution per dollar rather than margin percentage?","acceptedAnswer":{"@type":"Answer","text":"Because margin percentage ignores how long the cash is tied up and how much of it each unit consumes. A 60% margin item with a four month cover can be a worse use of working capital than a 40% margin item that turns twice as fast."}},{"@type":"Question","name":"What does it do with a forecast it does not trust?","acceptedAnswer":{"@type":"Answer","text":"It flags the SKU, says what the forecast rests on, and sizes the buy more conservatively. A forecast driven by one promotional month is the most common case, and repeating that quantity without repeating the promotion is a reliable way to create overstock."}},{"@type":"Question","name":"What is included with the skill?","acceptedAnswer":{"@type":"Answer","text":"The skill itself, the forecast notes reference, the quarterly buy format, and the working capital fitting rules. It is free and security scanned. We re-publish the ZIP when the model changes, so download it again if your quantities start shifting."}}]}</script>