How much stock should I hold?

The calculation is simpler than it appears. Take how much you sell in a typical week, multiply by the number of weeks your supplier takes to deliver, and add a buffer for the weeks that are busier than typical. That figure is your reorder point, and holding substantially more than it means paying for storage and risk you did not need to carry.

The buffer is where judgement enters, and it should reflect how bad a stockout actually is. For something customers will wait for, a small buffer is fine. For something they will buy elsewhere immediately, the buffer needs to be larger because the cost of running out is a lost sale and possibly a lost customer. Those are different situations and applying one rule to both is why businesses hold too much of some things and too little of others.

Buy shallow on anything unproven, which is the discipline that matters most in a first year. A product that has not sold yet has no demand history, which means any quantity is a guess, and a guess that arrives in a case of five hundred is a guess you will be living with for two years. Small initial orders cost more per unit and considerably less in total risk.

Understand what holding stock actually costs beyond the purchase price. Storage, insurance, the risk of damage or obsolescence, and the opportunity cost of cash tied up. That total is commonly estimated at a meaningful percentage of the inventory value annually, which changes the arithmetic on a bulk discount that looked attractive.

Watch the ratio between what you hold and what you sell rather than the absolute figure, since that is what indicates whether inventory is moving. Stock sitting for three times the period comparable items take to sell is heading toward dead, and catching that at three months rather than twelve is the difference between a discount and a write off.

Separate the products that behave differently rather than applying one policy. Fast moving items with reliable demand can be reordered frequently with a small buffer. Slow items should be held minimally or ordered on demand. Seasonal items need a plan for what happens to whatever remains, decided before the season rather than after.

Track it properly from the start, which means a system that adjusts as sales happen rather than a count somebody performs occasionally. Manual reconciliation stops being done by month three, at which point you are guessing, and the guess is always more optimistic than the shelf.

Then count physically against the system on a rotating schedule, since even a good system drifts through theft, damage, and miscounted deliveries. Counting a section at a time is considerably more sustainable than an annual count of everything, and a figure nobody has verified against a shelf is confidently wrong within a few months.

Negotiate shorter lead times before increasing quantities, since the amount you hold is driven directly by how long replenishment takes. A supplier who can deliver in one week rather than four allows you to hold a quarter of the stock for the same service level, and asking costs a conversation.

Watch seasonal items with particular care, since anything that does not sell within its window ties up cash until the following year and frequently has to be discounted heavily. Deciding what happens to the remainder before the season starts is what prevents that becoming a write off.