Ask how much custom software costs for a small business and you will hear everything from a few hundred dollars to numbers with too many zeros, and both can be honest quotes for different jobs. The useful skill is knowing what moves the price, so you can scope small and get a flat number you can trust.
Two businesses ask for "a simple tool to track jobs". One gets a quote for a few hundred dollars, the other for tens of thousands. Neither quote is dishonest; the two requests just unpacked into wildly different amounts of work. Understanding what moves the number is the difference between buying the tool you need and funding a project you never wanted.
Words like "simple", "basic", and "just" hide scope. "Just a form that saves to a list" is small. "And it should email the customer, sync with QuickBooks, work offline in the field, and let three employees log in with different permissions" is five projects wearing one sentence. The builder who asks lots of annoying questions before quoting is protecting you; the one who quotes instantly off one sentence is guessing, and you will pay for the guess later.
The choice is rarely "buy a big platform or commission a big system". Most small businesses need something in between: one chatbot, one scraper, one tracker, one automation that removes one specific chore. Small tools like that are days of work, not months, which is why they can be quoted flat and delivered fast. The pattern shows up across cheap busywork automation: the highest-value builds are usually the least impressive-sounding ones.
This is the space Keelflo works in: single-purpose tools with a flat quote agreed upfront, built once, owned by you. No hourly meter running, no surprise line items. For a sense of real numbers on one common tool type, see what a website chatbot costs.
Off-the-shelf subscription software is the right answer when your need is standard, invoicing, payroll, email. Custom wins when the need is specific to how you work and the subscription tools make you bend your process around theirs. The math is straightforward: total the subscription over the years you expect to use it, add the hours of workarounds it forces, and compare against a one-time build. A build you own stops costing money when it is done; a subscription never does. Neither answer is always right, which is why the comparison is worth five minutes with a calculator.
Describe the problem, not the feature list. "I lose track of which quotes I have not followed up on" gives a builder what they need to propose the smallest tool that fixes it. Then ask three questions: Is the price flat or hourly? What exactly is included? What happens after delivery if something breaks or the world changes? Clear answers to those three tell you more about the builder than any portfolio.
Shrink the scope, not the builder quality. A tiny tool that solves one real problem, built well, beats a sprawling system built cheap. Cutting features is free; fixing a bad build is not.
Hourly shifts the risk of surprises onto you; flat pricing forces the builder to scope carefully upfront and absorb their own misjudgments. For small, well-defined tools, flat pricing is fairer to the buyer and keeps incentives honest.
Ask this before anything else. Some shops build on platforms you must keep paying to use, or keep the code themselves. A build you own outright can be hosted, modified, or handed to another developer later. Get ownership in writing.
There can be legitimate small ones: hosting, a domain, a third-party AI or API fee if the tool uses one. Those belong to you directly and should be named upfront, at cost, with no markup. Anything that only appears on the final invoice was hidden, and that is a builder problem, not a software problem.