The second order is a different decision
A first order is a hypothesis. It says: this market will buy this range, at this price, in these finishes. A second order is the moment the hypothesis meets evidence, and most of the evidence is sitting in a spreadsheet nobody has opened since the container arrived.
The tempting move is to repeat the first order with a few models swapped by instinct. It is also the move that keeps a wholesale range from ever settling, because it replaces data with memory and memory is generous. What actually sold is rarely what the buyer expected to sell, and what broke is rarely what they expected to break.
Before the second order is placed, four things need to be extracted from the first one: what sold and at what rate, what failed, how the delivery compared to the plan, and which models the supplier can actually repeat. Each is a different kind of information, and they should not be averaged into a single feeling about the range.
Turn the first order into data
The data does not have to be sophisticated. It has to exist in a form that can be sorted. A single table, one row per model, with the same columns for every shipment, is enough to answer every question the second order raises.
- Units received, by model and finish — from the packing list, not from the invoice total.
- Units sold, by month, by model and finish.
- Units broken in transit, and units found defective after unpacking — recorded separately, because they are different problems.
- Units returned, and the stated reason for each return.
- Units still in stock, by model and finish, and how long they have been there.
- The dates the order was confirmed, the dates the container was loaded, and the dates it arrived.
Two of those columns are usually missing and both are decisive: the finish, and the arrival date. A range that sells well in white and badly in black is a different range from the one the sales report describes, and a shipment whose arrival drifted from the plan has taught you something about the supply chain rather than about the market.
Sell-through by model, not by category
Category-level reporting hides the only facts a reorder needs. “Toilets sold well” is compatible with two very different situations: every model moving steadily, or two models carrying the range while the rest sit. The second situation is much more common, and it means the second order should be narrower than the first, not broader.
Sort the range into three groups by model. The models that moved without prompting. The models that moved only after a discount, an installation, or a display change. The models that have not moved at all. The first group is the second order. The second group is a question about positioning, finish or price, and it should not be reordered until that question has an answer. The third group is a decision about whether the model belongs in the range at all.
Finish behaves like a separate model in this analysis, and it should be counted that way. A coloured or matt finish carries its own minimum at the factory, which means a finish that sells slowly in a broad range is more expensive to keep than its unit sales suggest.
Breakage and claims as range data
Damage is not only a claims event. It is a signal about the load, and it points at specific parts of the range. A model that breaks repeatedly while its neighbours arrive intact is telling you something about its carton, its weight distribution, or the way it sits in a stack.
Record damage by model and by position in the container. If the failures cluster at the door end, the problem is movement in transit. If they cluster at the bottom of stacks, the problem is compression and the stack limit is wrong. If they cluster around one model regardless of position, the problem is the packing of that model.
That distinction changes what you ask for on the second order. A compression problem is a packaging specification conversation. A single-model problem is a carton design conversation. A claims problem is neither: it is about evidence, and it may have nothing to do with the goods at all.
One more column belongs in the table: whether the damaged units were replaced, credited or written off. That record is what tells you the real cost of the range, which is never the unit price alone.
What the delivery actually taught you
Compare the plan with what happened, in three places. Confirmation to loading. Loading to departure. Departure to arrival. Whichever of the three moved most is where the second order should change something.
If the gap sits between confirmation and loading, the constraint is production or materials, and the answer is earlier confirmation or a longer planning horizon. If it sits between loading and departure, the constraint is booking and documentation, and the answer is a document set prepared before the container is stuffed. If it sits after departure, the route is the variable, and the answer is a conversation with the forwarder rather than with the factory.
None of those three are the supplier’s fault by default, and treating them as one number hides which party can actually fix the delay. The second order should carry the conclusion, not the complaint.
Cutting is the harder half
Adding models to a second order feels like growth. Removing them feels like admitting a mistake, which is why ranges accumulate and capital sits in finishes nobody asks for.
The test for keeping a model is not whether it sold at all. It is whether it sold without help, at a margin that covers the space it occupies, and whether it can be replenished on its own. A model that needs a discount to move is a model that is teaching you the price is wrong; ordering it again at the same price produces the same lesson.
There is also a range-level argument for cutting: a narrower range is easier to display, easier to stock, easier to train staff on, and easier to replenish in a single shipment. Wholesale catalogues that sell well tend to be shorter than the ones that do not, and the shortening is usually done at the second order rather than at the tenth.
“Sells” and “can be replenished” are different tests
The most expensive mistake in a second order is treating demand as the only constraint. A model can sell beautifully and still be a bad line to reorder, for reasons that have nothing to do with the market.
Four of those reasons are worth checking model by model. Whether the model is a standard line or a finish with its own minimum. Whether the components — seat, hinges, flush plate — are available in the same lead time as the ceramic. Whether the model is certified for the market you are selling into, since a certificate covers a model and a market rather than a range. And whether the model shares parts with others in the range, because a shared part is a smaller spare-parts burden.
A model that passes the demand test and fails the replenishment test will produce a second order that arrives and cannot be repeated. Better to know that before the container is booked.
What the second RFQ should contain
A second enquiry is not a copy of the first with new quantities. It carries the lessons, written down, so the reply can be compared with what actually happened.
- The shortlist by model and finish, with the quantity for each, and the models being dropped listed explicitly rather than omitted.
- The damage record from the first shipment, by model and position, with a request for the packaging change that addresses it.
- The three delivery intervals from the first order, and which one the buyer needs to change.
- Any finish that will be added, with the minimum quantity that finish carries, confirmed in writing.
- The document set required for the destination, restated, because it may have changed since the first shipment.
Sent that way, the second quotation is a response to a specific history rather than a fresh catalogue price. That is the difference between a range that settles and a range that keeps being re-guessed.
Frequently asked questions
Should the second order repeat the first exactly?
Only if the first order’s data supports it. Repeating is a decision, not a default: it should follow from sell-through by model and finish, the damage record, and the delivery record. In most ranges the second order is narrower than the first, not identical to it.
What is the minimum data I need before reordering?
Units received, sold, damaged and returned, by model and finish, plus the dates the order was confirmed, loaded and delivered. Six columns and one row per model are enough to make the decision defensible, and most of it comes from documents you already hold.
How do I decide which models to drop?
Drop the models that only moved with help — a discount, a display change, an installer’s persuasion — and the ones that never moved at all. A model that needs a discount to sell is telling you the price is wrong, and reordering it at the same price repeats the lesson.
Why does the finish matter so much?
Because colour and matt finishes carry their own minimum quantity per model at the factory. A finish that sells slowly in a broad range costs more to keep than its unit sales suggest, and it should be analysed as its own line rather than folded into a model total.
What if the first order arrived late?
Find out which interval moved: confirmation to loading, loading to departure, or departure to arrival. Each points at a different party and a different fix. Carrying the complaint forward without that split produces a second order with the same problem in it.
Can I mix new models into a repeat order?
Yes, and a small trial quantity is a reasonable way to test a new model alongside proven ones. Keep the trial explicit in the enquiry, with its own quantity, so the range analysis for the following order can separate the trial from the repeat.
Related on this site
Documents and certificates
Related reading
- Consolidating suppliers into one catalogue
- Where the lead time actually goes
- Receiving a Container: The First Hour