Knowing What to Reorder: Demand-Driven Purchasing
Walk the warehouse of almost any distributor and you will see the same two problems side by side. One aisle holds pallets of something that has not moved in eighteen months. A few rows over, a fast seller is empty, and the customer has gone elsewhere. Both are purchasing decisions that went wrong, and both are costing money at once. Most buyers can feel that the stock is in the wrong shape; what they lack is the clean information needed to fix it without simply guessing harder.
Purchasing is the quiet engine of a distribution business. Buy too much and cash is trapped on the shelf, ageing toward write-off. Buy too little and you miss sales and train customers to call a competitor first. The job is to sit in the narrow band between those two failures, across thousands of lines, while lead times and demand refuse to stay still. Done on instinct, it drifts; done with good data, it becomes a repeatable discipline.
What “demand-driven” actually means
Demand-driven purchasing is a plain idea in a slightly formal name. Instead of reordering to a habit, a hunch, or whatever the supplier rep is pushing this month, you reorder in response to what is genuinely selling and what it takes to replenish it. Three questions sit underneath every line you stock:
- What should be reordered, and what should not?
- When does each line need reordering, given how long the supplier takes?
- How much should the order be, so you neither run dry nor overbuy?
None of these can be answered well for the whole catalogue in your head; a distributor carrying a few thousand active lines makes thousands of these small judgements a week. The aim is not to remove the buyer - experience and supplier relationships still matter - but to give a reliable starting position for each line, so attention goes to the exceptions rather than to re-deriving the obvious every time.
Reorder points, lead times and the min/max band
A reorder point is the stock level at which you should place a new order. Set it too low and the line runs out before the replenishment arrives; set it too high and you carry more stock than you need, all the time. It depends on two things: how fast the line sells, and how long the supplier takes to deliver - the lead time. A line that sells ten units a day from a supplier who delivers in seven days must reorder while there are still well over seventy units on the shelf, because seventy will be gone before the replacement lands. Most stockouts are not demand surprises; they are lead time underestimated or ignored.
Sensible buyers add a buffer on top - safety stock - to absorb the weeks when demand runs hot or the supplier slips. The size of that buffer is a deliberate trade between the cost of holding a little extra and the cost of disappointing a customer: generous for a cheap, fast-moving line, slim for an expensive, slow one.
Minimum and maximum levels put a sensible band around each line: a floor you do not want to drop below, and a ceiling that stops a well-meant bulk order from burying cash in stock you will be selling off in two years’ time. Set these per line, from real sales history rather than a flat rule across the catalogue, and do not freeze them - the demand and lead time behind them both move.
Demand patterns, seasonality and the lines that lie still
Averages flatten exactly the information a buyer needs. A line that sells twelve a month on average might sell two in the quiet months and forty in the run-up to a seasonal peak; order to the average and you are short in the busy weeks and overstocked in the slow ones. Looking at the actual shape of demand across the year, with last year’s pattern in front of you, lets you bring stock in ahead of a known peak and ease off before a known lull.
The other end of the catalogue deserves as much attention and usually gets far less. Slow-movers and dead stock are where cash quietly goes to die. Every line that has not sold in months is money already spent, sitting still and inching toward obsolescence. The first step is simply seeing it clearly: which lines have not moved, how much capital they represent, and how long they have been dormant. You cannot decide what to discount, return, bundle or stop reordering until you can see the dead weight plainly, and a purchasing approach worth its name surfaces these lines automatically rather than waiting for a stocktake to embarrass you. This is one of the things our SFA and logistics platform is built to make visible - the buying and stock picture together, so reorder decisions are made against what is genuinely moving.
Supplier reliability is part of the calculation
A reorder point built on a quoted lead time is only as honest as that lead time. Suppliers who promise seven days and routinely deliver in twelve are quietly forcing you to carry extra buffer or accept regular stockouts. Tracking the gap between promised and actual delivery, supplier by supplier, turns a rough sense of who is dependable into a longer effective lead time and a buffer sized to their real behaviour rather than their sales pitch. It also gives you something concrete to raise with them.
What data-driven purchasing actually needs
The methods above are not new, and they are not the hard part. The hard part is that every one of them depends on two things being true, and in many distribution businesses they are not.
The first is clean stock data. If the system says you have forty of something and the shelf has twelve, every reorder calculation built on that number is wrong before you start: the line looks healthy on screen while it is actually empty. Accurate, current figures - kept honest by disciplined goods-in, picking and counting - are the foundation.
The second is reliable sales history. Reorder points, seasonality and min/max levels are all derived from what has actually sold, over a long enough period to show the pattern. If sales data is fragmented across spreadsheets, lost when a line is renamed, or not captured at line level, there is nothing solid to reason from.
This is the unglamorous truth behind demand-driven purchasing: the clever bit is not the reorder formula but having stock and sales records you can trust, in one place. Get that foundation right and the rest is largely arithmetic the system can do for you, leaving the buyer to handle the judgement calls and supplier conversations that genuinely need a human.
Frequently asked questions
Do we need to replace our whole stock system to buy this way?
Usually not. What matters is that your stock and sales data are accurate and accessible. In many cases the existing records are sound enough to build on, or can be tidied and connected rather than discarded. The bigger lift is the discipline - keeping counts honest and capturing sales at line level - not buying new software for its own sake.
How do we deal with brand-new lines that have no sales history?
New lines are the one place where judgement has to lead, because there is no history to derive a reorder point from. Start conservatively, watch the early weeks of demand closely, and adjust the reorder point and min/max levels as a real pattern emerges. Within a season or two the line earns its own history and is treated like any other.
Will this take buying decisions away from our experienced staff?
No - the opposite. Calculating reorder points and flagging slow-movers across thousands of lines is mechanical work that buries good buyers in routine. Handing that to the system frees them for the exceptions, the negotiations and the seasonal calls that genuinely benefit from experience.
Demand-driven purchasing is not a clever trick. It is the steady habit of buying against what is really selling, with honest numbers for lead times and stock, and the discipline to look squarely at the lines that are not moving. Businesses that get this right do not eliminate every stockout or slow-mover, but they stop funding both at once, and free up the cash and shelf space that gut-feel buying quietly ties down.