Contents
01The question that settles most cases
Ask what happens if you do this exactly like your competitors. If the answer is nothing much, buy the tool they use. Invoicing, payroll, helpdesk ticketing and email delivery are solved problems, and a custom version of a solved problem is an expense with no upside.
If the answer is that you would lose the thing customers choose you for, that process deserves software of its own. The route planning of a logistics company, the quoting logic of a manufacturer and the triage of a clinic are not interchangeable, and forcing them into a generic tool quietly removes the advantage.
02Symptoms of a tool you have outgrown
Outgrowing a tool looks the same across industries, and it is visible before anyone decides to replace it.
- Someone exports to a spreadsheet every week to answer a question the tool cannot answer.
- A person spends part of every day copying data between two systems.
- Your actual rule cannot be expressed, so the team keeps it in their heads instead.
- You pay for seats so that people can read, not work.
- A change you need waits on a vendor roadmap you do not influence.
One symptom is normal. Three at once is usually cheaper to fix with software than to keep paying for.
03The five year comparison
Compare like for like over five years. A ready-made tool costs licences per seat, which rise as you hire, plus the hours spent working around what it cannot do. Custom software costs the build, hosting, and the changes you will want in years two to five. Neither number is small, and the honest comparison usually turns on headcount growth and on how often the process changes.
Our build work starts at $18,000 for a first system, and that figure is a band rather than a promise: discovery sets the actual range before anyone commits.
04The middle path most companies miss
The answer is rarely all of one or the other. The common shape that works is a bought system of record, usually a CRM, with a small custom layer where your advantage lives: the calculator that prices a job correctly, the portal a client logs into, the automation that routes work the way your team actually works.
That keeps the boring parts standard and supported, and puts the money where it changes a number you care about.
Questions
Will you tell us not to build?
Yes, and we do. It is on the short list of work we decline, alongside apps that should be a mobile web page and campaigns run without working measurement.
What if we start with a tool and outgrow it?
That is a normal path and a cheap one, provided the data stays exportable. Check before you commit that you can get your records out in a usable format without paying for the privilege.
How do you estimate custom work?
Fixed price discovery first, at $2,400 over three working days. It produces a written scope and a price band with the factors that move it. We absorb the first 15% of an overrun, then stop and re-plan with you rather than quietly continuing.
Where we can help