There is a version of this article that says "always build custom," written by people who build custom. There is another that says "always buy," written by people who sell subscriptions. Here is the one that tries to be useful.
Buy when
The job is generic. Email, accounting, scheduling, a CRM. Thousands of businesses need exactly the same thing, so the software is good and cheap. Do not build a calendar.
You need it tomorrow. Off-the-shelf works today. A custom tool takes weeks. If the problem is urgent and a product solves 80 percent of it, buy the product and revisit later.
You are still figuring out the process. Software forces a process on you, which is sometimes exactly what a new business needs. Once you know how you actually work, then you know what to build.
Build when
The job is specific to how you work. Your intake, your qualifying questions, your pricing logic, the three systems that have to talk to each other in your particular order. No product was made for that, so every product makes you adapt to it.
The generic product keeps embarrassing you. A chat widget that gives the wrong hours, a bot that cannot say whether you take an insurance, an autoresponder that sounds like a bank. Customers notice. A custom tool trained on your actual content answers the way you would.
The subscriptions have piled up. When you are paying for five tools to do one job, and a person to glue them together, building the one tool that does the job is usually cheaper inside a year.
You want to own it. Off-the-shelf means their servers, their pricing changes, their feature removals, their data policy. A custom tool runs on accounts in your name and does not disappear when a vendor gets acquired.
The four questions that settle it
- Does a product already do this well for businesses like mine? If yes, buy it. Honestly check first; a lot of "we need custom" is "we did not look."
- Would I have to change how I work to use it? A little is fine. A lot means the product is fighting you.
- Is the job tied to systems only I use together? Integrations are where products fail and custom tools earn their keep.
- Does this touch the customer in my voice? If a tool speaks for your business, generic is a liability.
Two or more answers pointing toward build, and it is worth a conversation.
The trap on each side
The build trap is building something a product already does well because custom sounds better. It costs more, takes longer, and you end up maintaining a worse calendar.
The buy trap is stacking subscriptions until your business runs on duct tape between apps, then paying someone to hold the tape. That is the most common state we find companies in when they call.
What a custom build looks like here
At Skald Studio a custom AI tool is scoped around one job after a free consultation, built on your content and your systems, and handed over with full ownership plus a monthly plan for usage and upkeep. If a product would serve you better, we say so on the call. The details are on the custom AI tools page.
Written by Coley Grantham
Founder of Skald Studio. Army EOD technician turned builder. He designs, codes, and answers the email himself. About the studio