The signal
A maker selling candles online and at weekend markets has a surprisingly awkward question to answer: how many are actually left?
An online shop tracks its own orders. A market sale, damaged item or box left at home may not reach that record. Selling through more than one channel creates another task: keeping the counts consistent. For someone running a business around work or study, that can mean an evening spent reconstructing the weekend.
Here is a small business idea worth testing: help one market seller turn scattered stock notes into a clear, repeatable count. Not an app. Not a warehouse system. A spreadsheet and a handoff they understand.
The problem is plausible, not automatically validated. Your first job is to find someone who experienced it recently and cares enough to pay for help. You do not need to spend your spare $50 to discover that.
How it actually works
Keep the offer narrow: “I’ll help reconcile one market’s stock movements and leave you a reusable counting sheet.” Start with a seller who has a manageable range of finished products, rather than hundreds of variations or materials that require complex calculations.
Ask how they handled their last market. Did they count everything again? Discover an online listing for something already sold? Spend time hunting through messages? A specific recent incident is stronger evidence than someone saying your idea sounds useful.
With permission, examine a small sample of their existing records. Then build a sheet in Google Sheets with fields such as:
- Product name and a consistent identifying code.
- Units taken to the market.
- Units sold, damaged or given away.
- Expected units coming back.
- Actual units counted and any unexplained difference.
The calculation is simple: units taken minus units sold, damaged or given away equals expected units returned. But the actual count matters. A spreadsheet cannot repair missing information by itself.
Test the sheet with invented numbers first. Include a case where the physical count disagrees with the expected count. Flag that difference for the owner; never quietly change an input to make the totals match.
Agree on a fixed pilot price and scope before touching the real records. Specify one event, the number of product lines and whether a walkthrough is included. Track your time: a tiny fee is not a workable business if every job becomes hours of detective work.
Avoid customer names, payment details and account passwords. Work from stock records, let the seller control access, and agree when you will delete your copy. Do not update live listings or claim to provide accounting or tax advice.
Real example
InvenTree is a real open-source inventory management system built around tracking parts and stock. Its project shows that inventory is an established software category, not a problem you have invented.
That does not prove a local maker wants your service. Nor does it mean you should install a full inventory system for your first customer. Software brings setup, maintenance and support responsibilities; “open source” does not mean effortless.
Your smaller opportunity is the work around the tool: choosing consistent product names, finding gaps in the records and making a routine someone will actually follow.
A useful pilot ends with the owner able to repeat the process without you. Ask whether it reduced recounting or exposed missing information, and whether they would pay for another session. Compliments alone are not validation.
Your small move today
Spend 15 minutes sending one seller you know this message: “After your last market, how did you work out what stock was left? I’m exploring a small stock-count cleanup service. Would you show me your process for ten minutes? No customer details needed.” Do not buy software first. Find out whether the problem is real for that person.
Drafted with AI from trusted public sources and checked against our editorial rules. Educational content only — not personalized investment advice. Investing has risks and returns are never guaranteed.