How to Choose a Supplier for Ecommerce Inventory

Two quotes side by side, and the only line you can compare without doing any work is the unit price. It is also the line least likely to decide whether the supplier works out.

Choosing a supplier is a judgement problem, not a calculation. This post gives you the criteria and the trade-offs behind them. It does not give you a scoring sheet that produces an answer, because any score you build before the first order is a weighted average of guesses.

What you are actually choosing

A supplier is three things you will live inside for as long as the relationship lasts: a lead time, a minimum, and a failure mode.

The lead time sets how long every mistake takes to correct. Order too little from a 30-day supplier and you are short for a month. The minimum sets how much cash the relationship demands per cycle, and how much stock you carry whether you wanted it or not. The failure mode is what happens on their bad week: a short shipment, a silent delay, a substituted component, a price that moves without notice.

The uncomfortable part: the criterion that matters most is reliability, and it is the one you cannot observe before the first order. Everything else on a quote can be read. Reliability can only be measured afterwards, from your own delivery history. So pick for the constraints you can see, and size the first order so that finding out the rest is affordable.

The criteria that matter

In rough order of how much they change your operation, not how easy they are to compare.

No formula exists because the weights are properties of your store, not of the suppliers. A store with spare cash and a stable seller can absorb a 300-unit minimum for a better price. A store whose cash is the binding constraint should pay more per unit and keep the flexibility. Same two quotes, different right answer.

Questions to ask before ordering

Ask these in writing so the answers exist in a form you can go back to.

The last one is worth asking even though many suppliers will decline. A supplier who cannot produce one small-customer reference is telling you either that they do not have small customers or that the small ones are not happy, and both are useful before you commit.

Sampling and the first order

Samples answer one question: does the product match what was described. They answer nothing at all about whether the supplier can hit a date, ship the right count, or handle a problem, because a sample is not run through the process a real order goes through.

So the first order is the actual test, and it should be sized to be corrected rather than to be right. The target is the smallest quantity that clears their minimum and that you can sell through inside a period you can afford to be wrong about. You are paying for information, and the price of that information is the cash tied up while you find out.

Two things to do before it goes out. Write down the date you expect it and the date it actually arrives, because that pair is the first row of the supplier's delivery history and it is worth nothing reconstructed from memory six months later. And count the receipt against the purchase order rather than eyeballing the pallet: a first-order short-ship that goes unnoticed sets the tone for every order after it.

If the product is new and you have no sales history to size the trial against, forecasting a new product without history covers the ways to get to a defensible number. And if their minimum forces a trial far larger than you would choose, that tension is worth pricing before you accept it: cash tied up in inventory is the constraint most likely to bite a small store first.

Comparing two quotes

Unit price is not the price. The number that belongs in a comparison is landed cost per unit: everything you pay to get one saleable unit onto your shelf, divided by the units you get.

Take Cedar & Fig, 250g, which sells 5 units a day at $18. Two quotes arrive.

The arithmetic, so you can check it. Quote A: 120 units at $7.00 is $840, plus $180 freight, giving $1,020 for 120 units, which is $8.50 landed. Quote B: 300 units at $6.40 is $1,920, plus $320 freight, which is $2,240, or $7.47 a unit, plus about $0.55 of duty and clearance, giving $8.02 landed and $2,405 of cash all in. Cover is just the units divided by 5 a day: 24 days from A, 60 days from B.

Quote B is genuinely cheaper. At 5 units a day the store moves roughly 1,825 units a year, and 48 cents on each of those is about $876 a year. That is a real saving and it should not be waved away.

It can still be the wrong first supplier. The trial costs $2,405 instead of $1,020, creates two months of cover instead of three and a half weeks, and every correction takes 30 days instead of 12. If the product sells at half the rate you expected, Quote A leaves you holding seven weeks of a mistake and Quote B leaves you holding four months of one. And a single drift from 30 days to 40 is a week of lost sales that takes a visible bite out of the year's saving.

One defensible reading of these numbers: start on A, measure, and move volume to B once the product has a history stable enough that 60 days of cover is a decision rather than a gamble. That is a reading, not a rule. Somebody with more cash and more confidence in the product could start on B and be right.

Once you have chosen, the lead time and the minimum stop being quote lines and become planning inputs. What they do to your reorder point is covered in how supplier lead times affect your forecast, and where the rest of the terms should live is in the supplier management guide.

Red flags

None of these prove a supplier is bad. Each one is a risk that should either change the terms you accept or change the size of the first order.

The pattern behind most of these is the same: an unwillingness to commit to specifics. A supplier who is comfortable writing down a date, a count and a price is a supplier who expects to hit them.

After the first few orders you stop guessing and start measuring, which is a different job with different tools: how to measure supplier performance.

Frequently Asked Questions

How many suppliers should I get quotes from?

Enough that you can tell whether a quote is normal, which usually means three or four rather than two. Two quotes give you a difference with no context: you cannot tell whether the cheaper one is competitive or whether the dearer one is quoting for something the other left out. Beyond four or five you are mostly spending time rather than learning anything new, unless the quotes are still landing far apart.

How big should a first trial order be?

The smallest quantity that clears their minimum and that you can sell through in a period you can afford to be wrong about. A first order is bought to produce information: whether they hit the date, whether the goods match the sample, whether the paperwork is right. If their minimum forces a trial larger than you can comfortably carry, that is itself a finding about whether the supplier fits your store.

Is a lower unit price worth a longer lead time?

Sometimes, and the way to tell is to convert both into things you can compare. A longer lead time means more cash held as stock, a slower correction when demand moves, and a bigger buffer to cover the same service level. Work out the annual saving from the price difference at your actual sales rate, then decide whether it covers those three costs. There is no formula that answers it for you, because the weights depend on your cash position.

What should I check before paying a deposit?

That the lead time, the minimum, the unit price, the freight terms and what happens on a partial shipment are all in writing, in the same document, with a date on it. Verbal terms are the ones that get remembered differently later. Also check whether the deposit is refundable if they miss the agreed date, because that answer tells you how confident they are in the date they just gave you.

One habit worth starting on the day you pick: keep the quote you accepted, dated, next to the supplier's record. Twelve months later, when the price has moved twice and the lead time is running four days long, the original document is the only thing that separates a renegotiation from an argument about what was said.