# Starter kit · Method 16 · Data-cleaning and spreadsheet repair

> Strength: High. That is our screening judgement of how well this method holds up for a beginner with $0, not an earnings forecast. Typical first-sale window: 7–21 days.

## The offer in one look
- Buyer: small organizations with messy customer, stock or expense sheets
- Deliverable: cleaned file, rules log and validation report.
- What AI does: suggest formulas and identify anomalies; deterministic code/formulas do the actual checks.
- $0 start: the free tiers of the tools used in the workshop below. Check current limits before you rely on them.
- Boundary: never upload sensitive records without permission; use synthetic samples in demonstrations.

## Practise it first
Best starter build: Build 04 · Clean a real spreadsheet safely
- Workshop: https://open-practice.zlnew.workers.dev/earn/builds/spreadsheet-cleanup/
- Practice pack: https://open-practice.zlnew.workers.dev/resources/earn/packs/04-spreadsheet-cleanup.zip

## First sale, step by step
Start here: make one private sample of the deliverable (cleaned file, rules log and validation report) for one real prospect (buyer: small organizations with messy customer, stock or expense sheets) and send it to them personally.

1. Write your offer sheet for one narrow buyer (Build 01 helps). Name five real prospects you can reach without buying a list.
2. Make one private sample with invented or public information. Never use someone's private data without permission.
3. Send five personalised messages, one at a time, each with a reason that is specific to that person. No mass messages.
4. Offer a small, bounded first job with a clear price, timeline and revision limit.
5. Deliver, run the acceptance checks below, get written approval, and ask for an honest testimonial only after real work.

## Acceptance checks (proof to keep)
- [ ] The work delivers what the offer says: cleaned file, rules log and validation report
- [ ] Boundary kept: never upload sensitive records without permission; use synthetic samples in demonstrations
- [ ] The client approved the work in writing
- [ ] No invented reviews, numbers or guarantees anywhere

## Five prospects (sales practice)
| # | prospect (role or business) | why them specifically | personal first line | sent | reply |
|---|---|---|---|---|---|
| 1 | | | | | |
| 2 | | | | | |
| 3 | | | | | |
| 4 | | | | | |
| 5 | | | | | |

## Economics
- Hours per delivery:
- Free-tier limits that could bite:
- Price, and why:
- I will reinvest when:

---

# The offer contract

Every Earn by Doing build ends with these 10 artifacts. They make very different businesses measurable, reviewable and safe to practise. Fill in one contract per offer.

## 01 · Buyer
_One specific customer and a painful job worth solving._

- Buyer: small organizations with messy customer, stock or expense sheets
- The painful, repeated job:
- Where I can reach five of them this week:

## 02 · Bounded offer
_Exact deliverables, exclusions, timeline and revision limit._

- Deliverables: cleaned file, rules log and validation report.
- Exclusions:
- Timeline (days):
- Revisions (rounds):

## 03 · Free stack
_Primary tool plus an offline or free alternative; note current limits._

- Primary tool:
- Free or offline alternative:
- Limits checked on (date):

## 04 · Evidence
_Source log, inputs, outputs, tests and acceptance checklist._

- Where the evidence log lives:
- Acceptance checklist:

## 05 · Human gate
_Approval before publishing, messaging, spending or deletion._

- Who approves what, and how (in writing):

## 06 · Privacy
_Synthetic practice data; permission and minimisation for real data._

- Practice data used:
- Real data, with whose permission, and what I removed:

## 07 · Rights
_Provenance and commercial-use check for text, image, audio and code._

- Every asset, where it came from and whether it may be used commercially:

## 08 · Honesty
_No fake proof, invented testimonials or guaranteed-income claims._

- Claims I make, and the evidence for each: only what the acceptance checklist proves.

## 09 · Sales practice
_Five personalised prospects, not mass messages._

- Five prospects are in sales-practice.csv, each with a reason that is specific to them.

## 10 · Economics
_Time log, free-tier limits, delivery cost, price and a reinvestment trigger._

- Hours per delivery:
- Free-tier limits that could bite:
- Delivery cost:
- Price:
- I will reinvest when:

---
Open Practice · Earn by Doing · CC BY 4.0 · https://open-practice.zlnew.workers.dev/earn/