How to Work with Dropship Suppliers When You are Too Small to be a Priority

"Build a relationship with your supplier." Every sourcing guide says it. Check in often, be personable, learn the name of the rep's dog. That advice assumes your supplier is deciding whether they like you. They are not. At 6am, with the warehouse two days behind, they are deciding which accounts cost the least to serve. Warmth loses that decision. Predictability wins it. And predictability is something you can build in one afternoon.

FREE list of 250 Hot and Trending Products to Sell in 2026

Access a custom list of the hottest products to sell in 2026, hand-picked by eCommerce experts.

Full name

Your email

Access now

🔒 View our privacy policy to see how we protect your data.

Thanks for signing up!

Please check your inbox to access over 250 of the hottest selling products for 2026.

If you don't receive the email within 5 minutes, please contact support@salehoo.com

What your supplier is deciding at 6am

Your supplier cannot see your business and you cannot see theirs. You are reading a stock number that was true at 6am in a warehouse eight time zones away. They are reading an order that landed 40 minutes after cutoff, from an account that bought twice last month. Both of you are guessing. Most supplier advice tries to fix that with tone. Which is a bit like fixing a broken inventory feed by being nicer to it.

Disruption is not rare, and it is not only a small-seller problem. In RapidRatings' 2025 Annual Risk Survey of more than 100 senior supply chain professionals, fielded between December 2024 and January 2025, 81% said supplier disruption had hit their business in the previous two years. Those buyers have contracts, auditors and penalty clauses. You have a Gmail account and a Shopify admin. So preventing disruption is not a goal you can set. Shortening the gap between the disruption happening and you hearing about it is. Every step below is built around that one gap.

One worked example runs through all six steps below. It is arithmetic, not a company. Nothing in it comes from a real store: a shop doing 380 orders a month across 41 SKUs, with a main supplier in Yiwu, a second in Reno and four stragglers. Every figure in it exists so you can follow the maths on your own export.

Before step one, cut the list to two

You cannot run a routine with nine suppliers. There is not enough of you. Everything below costs roughly 90 minutes a month per supplier. So the first decision is which two accounts get that time, and which ones stay transactional.

Export the last 90 days of orders. One row per order, supplier name and order value against each. Pivot on supplier. Then read revenue share and order count as two separate rankings, because they say different things: forty small orders a month make you a familiar name in a warehouse, and four large ones do not. Anything under 5% of revenue stays transactional. You order from it. You do not manage it.

In the worked example the revenue split lands at 68% to Yiwu across 26 SKUs, 24% to Reno across 11, and 8% shared between four others. Two accounts carrying 92% of revenue is the shape you want. It means the routine is affordable.

Rank on tickets as well as on money. A supplier can be 6% of revenue and 40% of your support load. In that case, remove it. A better relationship does not fix a workload. If your top two are also the two you would struggle to replace, evaluating and choosing reliable suppliers is the read to do before you invest another month in them.

Step 1. Get seven terms answered in writing

Almost every supplier argument comes from a term that was assumed rather than agreed. What "in stock" means. Whether the 4pm cutoff is their clock or yours. Who eats a parcel a carrier loses. You will not get a signed contract at your volume. You do not need one. You need written answers, in a thread you can search.

Send one email with seven numbered questions and ask for the answers inline:

1. What time is the order cutoff, and in which time zone.
2. Does your lead time mean order to handover, or order to delivery.
3. How often does the stock feed refresh.
4. What happens to an order that backorders after I have taken payment.
5. What is the returns process, and who pays the return freight.
6. Who covers a parcel that goes missing in transit.
7. How much notice do I get before a price change.

Ask for one name and one address for exceptions while you are there. Then label the thread and archive it. You will read it again in eight months, at midnight, when something has gone wrong.

Question six pays for the whole email. If your buyer files with PayPal, PayPal's Seller Protection program, last updated on 26 January 2026, wants an online and verifiable tracking number as proof of shipment, so a supplier who ships untracked has quietly moved that risk onto you without either of you writing it down. In the worked example the Yiwu supplier answered five of seven and skipped both liability questions, which was useful information by itself.

Where it breaks: Suppliers answer the easy questions and go quiet on liability. Treat that silence as the answer, price it into your margin as though you pay, and ask again in 60 days. If you have never written this email, the wholesale ordering script for first-timers has wording you can lift.

Step 2. Publish a delivery range built on their lead time

Your product page converts your supplier's lead time into a promise to a stranger with a credit card. Get that conversion wrong and you do not have a supplier problem, you have a refund problem. Shipping is also where checkouts die: Baymard Institute's rolling average of 50 studies puts documented cart abandonment at 70.22%, with delivery cost and timing near the top of the stated reasons. Your delivery line is part of the price.

Do this:

1. Write the arithmetic down as three numbers: supplier handover, carrier transit, and a buffer of two days.
2. Publish a range rather than a date, and put it under the add-to-cart button instead of in a policy page.
3. Repeat the same range, word for word, in the confirmation email.

In the worked example the Yiwu supplier quotes 48-hour handover and 10 to 14 days transit. So the page says 14 to 18 days, and the confirmation email repeats it word for word. Reno ships the same SKUs in 4 to 6 days at $2.10 more per unit. That is a trade you can price. Shopify will calculate and show that estimate at checkout for you if you configure fulfillment time and delivery dates, so this is a settings job rather than a theme edit. Ask your supplier for the handover figure as a number, and check whether they already publish one. Printful publishes 2 to 5 business days to make and fulfil an order before anything reaches a carrier, and a number like that is the start of your range rather than the whole of it.

Step 3. Run one channel and one Monday message

Now the communication step, and it looks nothing like "check in regularly". Check-ins with nothing in them train a supplier to skim you, and once that habit forms the urgent messages get skimmed along with the rest. One predictable message, in one place, carrying something they can act on. That is the whole design.

Pick one channel per supplier and abandon the others. Then send five lines every Monday:

1. What sold last week, by SKU, in units.
2. What you expect to move this week, by SKU, in units.
3. Any promotion, with the date it starts.
4. Anything you want them to hold or pre-pick.
5. One question, and only one.

Keep exceptions out of it. Problems go to the named contact from Step 1, separately, so the Monday message never becomes the place where bad news lives.

In the worked example the Yiwu supplier started pre-picking the top SKU on Fridays without being asked. Nobody requested it. It took nine weeks to appear. That slow arrival is what real priority looks like, as opposed to the version where a supplier agrees on a call and nothing changes. Forecast low on purpose, too. A forecast you miss twice gets discounted permanently. A supplier you surprise twice starts building slack in your direction.

If your threads keep dissolving into arguments about what was agreed, solving communication problems with suppliers works through the specific failure patterns.

Step 4. Score four numbers every month, then send them the score

This is the step that separates a routine from a feeling, and it is the one that gets skipped, because sending a supplier their own report card feels rude the first time. It is not rude. You are already scored on exactly this. Under the FTC's Mail, Internet, or Telephone Order Merchandise Rule, 16 CFR Part 435, you have to ship within the time you clearly stated, or within 30 days if you stated nothing at all. Your supplier's handover time is the only thing standing between you and that deadline. Nobody at the Federal Trade Commission is worried about hurting your feelings either. So borrow the mechanism and point it at your supplier.

Four numbers, on one page, every month. On-time ship rate is orders handed to the carrier by the promised day, divided by total orders. I would set that threshold at 95%. Stock accuracy is orders cancelled or backordered after you took payment, as a share of orders, and I would set that threshold under 2%. Average reply time on exceptions gets one business day. Defect rate, meaning any refund or replacement caused by the product or the packing, goes under 3%. Those four thresholds are mine, not an industry standard. They give you a line to cross rather than a feeling to have. Move them once you have three months of your own history, because your own trailing average is a better bar than my round numbers.

Then send it. Four lines in the body of an email, no attachment, no essay, no adjectives and no thinly veiled remark about how much business you could be sending elsewhere. "Here is how last month looked from our side, and here is the one we would like to move." A warehouse manager who receives four numbers and one sentence can act on it before lunch. A warehouse manager who receives a spreadsheet attachment and three paragraphs will read it on Friday, or not at all.

Defect rate is the number that costs the most and gets watched the least. Returns are neither cheap nor rare: the NRF and Happy Returns 2025 returns study, published on 15 October 2025 from a survey of 2,006 consumers and 358 ecommerce professionals, put the online return rate at 19.3% of sales. So a supplier who takes a point off your defect rate is worth more than one who takes a point off your unit cost. Only the scorecard tells you which one you have. For the deeper version of this, including what to inspect when quality is the failing number, read supplier reliability and quality.

Step 5. Escalate on a schedule, not on a feeling

Escalating at the moment your patience runs out teaches a supplier nothing. That moment lands in a different place every time. Decide the trigger and the response now, while nothing is wrong. Then follow the schedule you wrote.

1. One bad month. Send the scorecard with one named number and ask what changed. Nothing else changes.
2. Two consecutive bad months. Ask for a call, put the fix in writing with a date on it, and move 10% of orders to your backup even at a worse unit cost.
3. Three months, or one unresolved liability event. Move your top five SKUs and say plainly why.

Never threaten a stage you will not run. Once is enough to make every future warning sound like noise.

Time it as carefully as you word it. Run a stage two in your quietest week, never in the fortnight before Black Friday. A supplier under peak load reads a scorecard as an attack, and you lose the one conversation that might have worked. In the worked example, stage two arrives in month two of a slide, 10% of orders move to Reno at a higher unit cost, and the on-time rate recovers the month after. Whether that was the conversation or the volume, you will never know. It does not matter.

When it does not recover at all, what to do when a dropshipping supplier lets you down covers the exit properly. Escalating on the wrong metric is the own goal here. A supplier who is late but honest is a scheduling problem you can plan around, and a supplier who is on time and wrong about stock is a trust problem you cannot. That second one skips the ladder and goes straight to stage three.

Step 6. Keep a backup live before you need it

A backup supplier you have never ordered from is not a backup. It is an untested contact. Switching only takes a day if the account already exists, the seven terms are already answered, and a real order has already moved through it.

1. Pick your top five SKUs by revenue. Those are the only ones that need a second source.
2. Find two candidates per SKU, then run the Step 1 terms email at both of them.
3. Route roughly 10% of real orders through the better one, permanently, even at a worse unit cost.
4. Match the backup on your published delivery range rather than on price.
5. Re-score it quarterly using the same four numbers from Step 4.

Put the quarterly re-score in the calendar, because a backup nobody re-scores goes stale in about five months.

In the worked example only five SKUs were second-sourced, not 41. Reno was already the live backup for three of them before the slide in Step 5 started. Stage two therefore took a day.

Doing this for free means working supplier lists, trade show exhibitor directories and marketplace seller pages, then verifying each candidate yourself: business registration on the state or national register, a phone call, a sample order. Budget a week per category and expect two of your ten candidates to survive it. A vetted directory like SaleHoo's collapses the verification part, which earns its money when you are second-sourcing several categories at once and does not when you need one supplier for one product. The free route in full is in finding reliable dropshipping suppliers.

Ten percent of your orders, permanently, at a slightly worse unit cost. That premium is the price of the switch taking a day.

The whole routine on one page

Six steps sounds like a lot until you see how little of it recurs. The terms email and the concentration pivot are annual jobs. Everything else collapses into three repeating slots.

That is the 90 minutes a month, laid out. It is also why the cap is two suppliers rather than nine.

Four bad mornings, and what to do in the first hour

Four failure modes account for most of it. Here is what each looks like from your side of the screen, and the move to make before you make any other move.

They ship late and blame the carrier

The tracking number tells you. USPS shows "Shipping Label Created, USPS Awaiting Item" and states plainly that the package has not yet been given to USPS, so a label created on day one with no carrier scan until day four means the parcel sat in the warehouse. Pull five recent orders and measure the gap between label creation and first scan. Then send the gap, not the accusation. It is a fact, and facts do not start arguments.

The stock feed says 40 and the shelf has two

You find out from a customer, which is the worst available way. Cancel and refund the same day. Do not wait for the supplier to confirm, and absorb the cost yourself. Then go back to the feed-refresh answer from Step 1, and if it is daily rather than hourly, raise your safety stock buffer until the gap stops being your problem. Late deliveries cost more than the order they ruin: in Narvar's State of Post-Purchase 2025 report, a survey of 3,461 US online shoppers run in August 2025, 74% had experienced a late delivery in the past year and 60% of 18 to 29 year olds said they would not shop with that retailer again. The refund is not the loss. The customer is.

A defective batch reaches 30 customers at once

Stop selling the SKU before you contact anybody. Then message every affected customer first and the supplier second, because your customers are deciding whether to file a chargeback while you wait for a reply from a different time zone. Ask the supplier for the batch number and a replacement run. If the return terms turn out worse than you assumed, handling dropshipping returns has the recovery options that do not destroy the margin.

They stop replying

Three business days on the named contact, then a second channel, then your stage three move. Silence has a few common causes and you will rarely learn which one you got, so do not spend three weeks finding out. The backup you kept live in Step 6 is the reason you can afford not to.

If you only do one of the six

It is Step 1, and it is not close. Open the thread with your biggest supplier, paste the seven questions, ask for inline answers. Eleven minutes. That email will surface at least one term you had wrong, and every other step here runs on those answers, so none of them work until the answers exist. The 90-day pivot can have your Saturday instead.

Somewhere in that thread is a question your supplier has never answered in writing, and the one about lost parcels is usually it. Which term has yours been quietly dodging?