Should a Small Company Build Its Own Internal Tools or Keep Buying Generic Software?

Almost every company of a few dozen people runs the same loop: spreadsheets get messy, generic software does not fit, an agency quotes a fortune, and everyone falls back to spreadsheets plus a group chat.

The cause is simple. Building an internal tool that matched your process used to cost far more than tolerating one that did not. When that cost drops to one person for one day, the trade-off flips.

Why generic software does not fit

It is not badly built - it has to serve ten thousand companies. Its leave-approval flow is the average of all their flows, so your one unusual step (say, the shop supervisor confirming the shift roster first) has nowhere to live.

The result: half the process is in the system and half is in a chat group. Neither side has complete data, so management reports cannot be trusted.

Three real advantages of building it yourself

1. The process follows your reality

"Leave requests go to the shop supervisor to confirm the roster, then to the plant manager, and anything over three days is copied to HR" is a one-sentence requirement. You stop reshaping the company to fit the software.

2. The fields are the ones you report on

Generic software reports the metrics it thinks matter. A tool you built captures the handful of numbers you actually review at month end - exportable as-is, with no manual rework.

3. Changes do not wait for a release

The process changed? Say one sentence and it is changed. No support ticket, no waiting for a vendor's next quarterly release.

Which internal tools to build first

Ranked by payoff, usually:

• Anything still totalled by hand on paper or in spreadsheets: inspection sheets, shop-floor reports, stock counts. Moving them to mobile forms saves a fixed number of hours every month.
• Approvals currently passing through group chat: leave, expenses, stamp requests. Even a basic system gives you a searchable record.
• Forms external customers fill in: follow-up surveys, appointment intake. Send a link - nothing for the customer to install.

Conversely, do not start with core systems for financial settlement or tax compliance. Those accuracy requirements belong with specialist systems and specialist people.

Three pieces of advice

Build one step, not a platform

Anyone trying to build "the unified company platform" in one go does not finish. Completing the inspection sheet and getting the frontline to use it beats drawing a grand diagram.

Let the people doing the work request changes

The biggest gain here is that the person with the requirement can make the change. The shop supervisor saying "we never use this field, drop it" is far more accurate than the same request relayed through three layers.

Run it for two weeks before rolling it out

One step, five people, two weeks. Copy it to other departments after that - by then you know which of your assumptions were wrong.

The cost comparison

Before: a custom build from an agency, two to three months, and changes billed separately.
Now: one person who understands the business ships a first version in a day, iterates to company-wide use within two weeks, for a few hundred dollars a year.

The difference is not only money. It is that the small tools never worth a formal project finally get built - and most of a company's wasted effort hides in exactly those small places.