One inventory or one per clinic? A practical answer for multi-location groups

Group practices outgrow per-clinic inventory the same way single clinics outgrow Excel. There is no specific moment. One day there is just too much rework, too many "we are out of size 3 brackets at Indiranagar but we have plenty at Koramangala" conversations, and the owner realises the system has stopped scaling.

The question is whether to fix it by going to a single shared inventory across locations, or by keeping per-location inventory but tightening the visibility between them.

The honest answer depends on three numbers. This post walks through which is right when.

The three numbers that decide it

Skip the philosophy. Pull these three numbers from your last 6 months:

1. Inter-clinic transfer frequency. How many times per month do you physically move a stock item from one location to another to cover a shortfall? Count the actual transfers. Be honest.
2. Stock-out cost per event. What does it cost when a clinic runs out of an item mid-procedure? This is hard to measure directly. A reasonable proxy: doctor time wasted plus patient inconvenience plus reputational cost. For a clinic running ₹1.5L per chair per day, a 30-minute stock-out costs roughly ₹7,500.
3. Distance and travel time between locations. A 15-minute drive is different from a 3-hour drive. A same-day transfer is feasible at the first, painful at the second.

With these three numbers in hand, the decision becomes mechanical.

The decision matrix

Transfers per month	Distance between clinics	Recommendation
Less than 4	Any	Keep per-location inventory. The transfer pain is too low to be worth the centralisation cost.
4 to 12	Under 30 minutes	Per-location inventory with a shared visibility layer. Each clinic owns its stock. The system shows what every other clinic has. Transfers happen on demand.
4 to 12	Over 30 minutes	Per-location inventory with reorder discipline. Centralising is too painful given the distance. Tighten reorder points instead.
More than 12	Under 30 minutes	Shared inventory with a central hub. One reorder process, one supplier relationship, daily transfer routes.
More than 12	Over 30 minutes	This is unstable. Either reduce transfers (better forecasting) or accept that each clinic must hold larger buffer stock.

What "shared inventory" actually means in practice

When clinics ask about shared inventory, they often mean different things. Three real models:

Model 1: shared visibility, separate ownership

Each clinic still buys, holds, and accounts for its own stock. But every clinic can see every other clinic's stock levels in real time. When clinic A is low, the front desk at A can see clinic B has plenty and request a transfer.

This is the right model for most 2-clinic and 3-clinic groups. Low operational change. Big reduction in stock-out friction.

Model 2: central hub, spoke clinics

One location is designated the hub. The hub holds the bulk of inventory. Spokes hold a small working stock for the day or week. A daily or twice-weekly route restocks the spokes from the hub.

This is the right model for groups of 4+ clinics within a small geographic area (one city, broadly). It cuts total inventory holding by 30 to 40 percent because the spokes do not need to carry full safety stock. The savings on capital tied up in inventory are real.

Model 3: fully pooled, no per-location accounting

Every item is owned by the group, not by any specific clinic. Clinics consume from the pool. P&L per clinic still works because consumption is logged against location.

This model is rare and usually a mistake at fewer than 8 clinics. It removes per-location accountability for waste, which is a real problem. Clinics that should never go to this model often try and then quietly revert.

The accounting wrinkle

If your group is structured as separate legal entities per clinic, shared inventory triggers transfer pricing concerns. Each transfer between clinics is a sale from one entity to another. Your accountant will have opinions.

For most family-run groups in India, this is solvable with internal transfer-priced billing. Set a fixed internal price per item, log every transfer, settle quarterly. Talk to your CA before you go down this path. It is not optional.

If your group is structured as one entity with multiple branches, this disappears. There is no inter-entity transfer, just internal logistics.

The reorder discipline question

Every model above lives or dies on reorder discipline. The cleanest design fails if no one is actually re-ordering on time.

The mechanism that works:

• Each item has a reorder point (the level at which a new order is placed).
• Each item has an order quantity (how much to buy when reordering).
• The system raises a flag when stock crosses the reorder point.
• One named person (per clinic, or for the group) reviews the flag list daily.

Both numbers (reorder point and order quantity) come from real consumption data, not gut feel. Most clinics overstock by 30 to 50 percent because they set reorder points based on a bad week three years ago. Reset annually based on the last 6 months of actual usage.

Kythro's inventory module does this automatically once you have a few months of consumption history. Most clinics see a 25 to 35 percent reduction in inventory cash tied up after the first reset.

What changes operationally as you grow

A 1-clinic practice can run inventory by feel. The owner walks past the bracket drawer and notices it is low.

A 2-clinic practice needs a shared visibility layer or it will accumulate transfer pain.

A 3 to 5-clinic practice needs reorder discipline backed by a system. By feel does not scale to 5 clinics. Mistakes compound.

A 6+ clinic practice needs either a central hub or rigorous per-location process. The economics force the choice. Holding 6x of full safety stock at every clinic ties up too much capital.

The summary

The "one inventory or many" question has a numerical answer, not an opinion. Count your actual transfers, measure your stock-out cost, look at the geography. The matrix above tells you which model fits.

Pick a model that matches where you are now, not where you might be in 5 years. The cost of switching models later is small. The cost of running the wrong model now is large and recurring.