Tell a café what actually sells: a £250 one-page report from their messy spreadsheet
The practice till export in this mission has 198 rows pasted in twice, quietly inflating every total the owner has ever looked at. Owners don't want a cleaner spreadsheet; they want to know what sells, what doesn't and when to put more staff on. Clean the data with a tested DuckDB or Python script, answer those questions in a one-page report, and sell it with a monthly refresh.
- A kit to download: discovery questions, data protection terms, outreach templates, pricing calculator, weekly scoreboard and a duplicate checker owners can run on their own file
- The technical lessons this builds on: Analyse a million-row spreadsheet in seconds (installs DuckDB) and Beautiful invoices and CVs from plain text (installs Typst)
- Optional, for the Python version: uv, which the marimo lesson installs
- A spreadsheet program to open CSV files: Excel, Google Sheets or LibreOffice Calc
- The kit: download it all as a zip or read the kit overview
Pick who you'll help, and what it takes
Small businesses are sitting on answers they can't see. Know what the job involves and who pays for it.
Walk into any independent café and ask the owner what their best-selling item was last month. Most will guess. The answer is sitting in their till export, under duplicates, typos, blank rows and a date column that changed format when they got a new card reader. Turning that into one page of plain answers is a service a small business can see the value of straight away, and it's built on free tools from this site.
What it takes:
- Sell answers, not cleaning. Cleaning on its own sells cheap: on PeoplePerHour, most Excel data-cleansing offers are priced at £10 to £25, and most freelancers listing that skill charge £6 to £30 an hour (PeoplePerHour ↗). What you sell is the answers: five questions in the owner's words, one page, explained in person, from someone local who understands their business.
- Lead with a sample. Nobody searches for "spreadsheet clean-up", so you show owners a sample report and ask the right question. That's why the sample is the first thing you build.
- Personal data comes with the job. Customer names, emails, members' addresses: you handle them under UK GDPR, and step 2 shows how to keep that risk small.
- Walk in with the sample. Hunter's State of Cold Email ↗ report (31 million emails sent in 2025) found an average reply rate of 4.5%, so cold email alone is slow. Walking in with a sample report works better for this service.
- Each client's file is a little different. The kit's scripts handle the common problems; each new client means adapting column names and building a spelling list, so plan for your first real job to take two to three times as long as later ones.
Timelinea week or two of evenings to get comfortable with the scripts, then several weeks of steady outreach before your first paying client. Keep your day job while you land the first few. Everything here is free.
TipWho this suits: you enjoy detail work (reading odd rows, checking totals), you'd rather ask than guess at a number, you keep other people's data private and secure, and you're happy talking to business owners face to face. Half of this job is conversations.Who actually paysPick one group to start. Your sample report and your messages get sharper when you do:
- Cafés, delis and bakeries: lots of small sales, a till that exports, and real decisions about staffing, opening hours and what to bake.
- Independent shops and market traders: gift shops, bookshops, plant shops. Stock ties up cash, so "what isn't selling" matters.
- Trades: plumbers, electricians, cleaners, gardeners with a jobs spreadsheet. Which jobs pay best per hour, which customers come back, which months are quiet.
- Clubs and societies: sports clubs, choirs, allotment societies with a membership list. Renewals, lapsed members, income by month.
- Small charities: donations, events and volunteers in a spreadsheet the treasurer inherited. Which events covered their costs, how giving changes through the year.
Signs a business needs youa job advert mentioning "keeping the spreadsheet up to date"; a new till or website in the last year (two systems, two formats); an owner who says "I keep meaning to look at the numbers"; a club treasurer who's just taken over; anyone who exports data and never opens it.
TipYour sister service, if their numbers are stuck in PDFs (supplier invoices, statements, old reports), is the PDF to spreadsheet service. The two fit together: get the data out of PDFs, then clean it and report on it.