How to Scope a Spreadsheet Cleanup Service for Small Product Catalogues

A small shop's product list can become awkward long before it becomes large. Categories drift, extra spaces appear, and the same item arrives under several slightly different names. The owner may need a cleaner working file without needing a new inventory system.

A spreadsheet cleanup service can address that specific problem. The useful promise is a reviewed catalogue with documented changes and unresolved questions. It is not a promise to repair the shop's entire operation or upload a file directly into its live store.

Choose a narrow first job

Start with a product catalogue containing descriptions and identifiers, rather than customer records, accounts, or payroll. For an initial pilot, you might accept one sheet, up to 150 rows, and six named columns. Inspect a representative sample before quoting; row count alone says little about the judgement required.

A manageable scope could include trimming unnecessary spaces in descriptive fields, applying an approved category list, flagging missing values, and identifying possible duplicates. Keep rewriting sales copy, researching specifications, changing prices, and store imports outside the initial offer.

The guide to validating a side hustle idea can help you test whether anyone wants the outcome. In this case, show a small before-and-after example using invented products. Demonstrate a clearer file without exposing another client's catalogue.

Agree on rules before touching the data

Use a brief that the client can complete in plain language:

• Input: one named file, the relevant sheet, and the six columns included.
• Protected fields: product identifiers, supplier codes, prices, and any formulas that must remain unchanged.
• Category rules: the approved list and examples of items that belong in each category.
• Duplicate rule: what combination of fields identifies the same product and variant.
• Unknown values: the label used for a question awaiting the owner's decision.
• Output: a cleaned copy, a change log, and a separate questions sheet.
• Acceptance: one reviewer checks the sample and provides one combined feedback list.

Do not assume two products are duplicates because their titles match. A candle may appear in several sizes, scents, or packaging options. Treat a suspected match as a review item until the agreed identifiers establish whether it is a duplicate.

Preserve an untouched source and test ten rows

Keep the source file unchanged and work on a copy. Add a working source-row reference so you can trace a cleaned record back to its origin. Record the initial number of data rows and the names of the protected columns.

Choose ten rows that include both ordinary entries and awkward cases. Apply the proposed rules, then ask the owner to approve the result. A category change that seems obvious to you may conflict with how the shop organises its shelves or website.

For example, the owner might approve changing “ Kitchen ” to “Kitchen” in a category field while rejecting a change from “Mug - Blue” to “Blue Mug” in the product title. The service should implement agreed rules, not impose your preferred naming style.

Use cleanup tools selectively

Google's official Sheets cleanup documentation explains how to split text, remove duplicates, and trim whitespace. It also notes that duplicate detection can treat values with different letter cases, formatting, or formulas as duplicates, and that trimming does not remove non-breaking spaces.

Those details matter when product codes are significant. Protect identifiers from automatic cleanup unless the owner has explicitly approved a rule for them. Preserve leading zeroes, and check the exported file if the handover format could change how a code is represented.

For a first pilot, flag possible duplicates instead of deleting them. Keep the row count unchanged unless deletion is an agreed part of the job. Reconcile every approved removal in the change log so the final total can be explained.

Make the handover inspectable

Use one change-log entry per rule or exceptional edit. Include the source row, field, original value, new value, and reason. For repeated spacing fixes, a rule summary plus an affected-row list may be easier to read than hundreds of nearly identical notes.

Finish with three checks: compare protected fields against the source, confirm every category belongs to the approved list, and list all unresolved questions. Review every changed row in a small pilot. A spot check alone is a weak substitute when the file is short enough to inspect fully.

Send the cleaned copy with a short note stating what is complete and what still needs a decision. Use a consistent file backup routine, and agree with the client how long you should retain working copies. The owner should review the file before using it in another system.

Price the uncertainty as well as the row count

Consider this hypothetical 120-row job: 20 minutes of intake and sampling, 50 minutes applying rules, 30 minutes of checking, and 20 minutes preparing the handover and handling feedback. That is two hours. An illustrative £70 fee works out to £35 per delivery hour before expenses, tax, and unpaid business time.

If ambiguous variants add another hour, the same fee becomes about £23.33 per hour. Neither figure is a market benchmark or an income expectation. The example shows why “up to 150 rows” also needs limits on fields, rules, and unresolved cases.

Complete one paid pilot and record where the time went. Use the results to tighten the brief, revise the quote, or decline files outside your competence. A repeatable service begins with a file you can explain and a scope you can finish.