Buying a product gets you working quickly and shares the cost of maintenance with every other customer. Building gives you an exact fit and full control. Neither is right by default. These questions help decide.
The five questions
Is this process a source of advantage? Payroll is the same everywhere; buy it. The way you quote, schedule or serve customers may be why they choose you; that is worth owning.
How far would you bend to fit a product? If adopting the product means changing a process for the better, good. If it means degrading something customers value, the licence fee is not the real cost.
What will the product cost in five years? Add per-user fees as you grow, the integrations you will need and the price of leaving. Compare that with building and maintaining your own.
Who will look after it? Custom software needs an owner, in-house or contracted. If nobody will hold that responsibility, buy.
Can you do both? Often the answer is a standard product for the commodity parts, with a small custom layer or integration around the part that is uniquely yours.
Decide with numbers and a prototype
Write the answers down with rough costs beside them. Where uncertainty remains, a two-week prototype or a product trial with real data will settle more than another month of discussion.