Start with your customers, not the logistics
Delivery is a retail question before it is an operations question. The useful starting point is how many of your existing customers would order from home if that option existed, and how many people in your trade area never walk in at all.
Stores that treat delivery as a new sales channel tend to plan it well. Stores that treat it as an errand tend to under-resource it and conclude it does not work.
The three constraints that decide the answer
Territory: how far can you reasonably serve before the economics stop making sense? Volume: is there enough demand to keep a delivery window productive? Capacity: can your staff prepare delivery orders without slowing the counter?
If territory and volume look workable, the remaining question is who runs the operation — your payroll or a delivery partner.
What to do next
Model a conservative scenario: a small share of current orders, a realistic territory, your existing operating days. If the channel is still meaningful at conservative numbers, it is worth building.