A hotel project in Doha needed 340 reading floor lamps installed before a soft opening. The buyer had nineteen days. Their first move was to email eight floor lamp manufacturers and ask for the best price.

That is the standard playbook. In 2019 it was also roughly the right playbook. In 2025 it is the right answer maybe half the time — and the other half costs people actual money.

I coordinate rush orders at a lighting sourcing company. I've handled 400+ rush orders in six years, including same-week turnarounds for hotel rollouts and retail fixture refits. When a buyer calls me with a deadline that's already too short, I don't start with suppliers. I start with a comparison.

The two routes, and the four things that actually differ

Route A is direct from floor lamp manufacturers — you place a production order, they build it, lead time is quoted in weeks. Route B is stock-and-ship — a distributor or trading company already holds inventory, and lead time is quoted in days.

Almost every comparison of these two routes turns into a price argument. Price is the least interesting axis. The four that decide whether your project lands on time:

  • Where the days actually go
  • Whether the dimming spec is real or nominal
  • MOQ behavior on small, mixed orders
  • What the paperwork actually proves

And one thing that isn't a dimension so much as a trap: total landed cost. More on that at the end.

It took me four years and roughly 300 rush orders to understand that the route matters less than where the bottleneck sits. The bottleneck just happens to sit in different places on each route.

1. Lead time: the days are almost never in production

A factory quoting 25 days is usually quoting 12 days of production plus 13 days of queue. Ask a specific question: on what date does my order enter the line? If the answer is vague, add two weeks and plan against that number.

The other thing worth checking is component availability, because it's invisible from the outside. For a dimmable floor lamp, the driver is the long pole. A factory can have the body, base, and shade sitting in stock and still be four weeks out on a 24W triac-dimmable driver in the right form factor. Nobody puts that on the quotation sheet (note to self: we still don't ask this early enough).

In March 2024 a client called on a Tuesday needing 600 tripod floor lamps in eleven days for a showroom opening. Normal turnaround on that model was 30 days. We found a factory holding the wood legs and metal components in stock with assembly capacity starting in six days, then paid $2,900 in air freight on top of a $34,000 order to compress transit from 32 days to 4. It landed on day nine. Their alternative was renting 600 lamps for three weeks, which would have cost roughly what buying them did.

Look at the shape of that solution. The money didn't buy speed from the factory. It bought a different shipping mode. Which is where most rush orders get saved or lost, and it has almost nothing to do with which supplier you picked.

2. Dimming: this is where factory-direct wins, and it isn't close

Dimming is where "in stock" quietly becomes a liability. Stock dimmable floor lamps are usually triac-only, and "dimmable" on a stock spec sheet frequently means 100% down to about 20% with visible flicker at the bottom of the range. That's fine for a lobby. It's not fine for a bedside scene in a hotel room where the lamp sits at 15% output for four hours.

What I mean is that the word "dimmable" describes a family of technologies, not a feature — triac/ELV, 0-10V, and DALI are not interchangeable, the depth of dim is a separate number, and the driver determines which one you actually got.

Three specs to put in writing before anyone quotes you:

  • CRI — Ra ≥ 90 for hospitality and retail, Ra ≥ 80 for general residential. Get it on the driver or module datasheet, not the marketing page.
  • Depth of dim and protocol — 1% or 10%, and which control protocol. Match this to the building's existing control system before you order, not after install.
  • Flicker — IEEE 1789-2015 is the reference most lighting practitioners cite. Percent flicker under about 8% at low output is a defensible spec floor for anything a person will sit next to for hours.

Safety and compliance paperwork should name the standard explicitly: EN 60598-1 for luminaires generally, EN 62471 for photobiological safety, EN 55015 and EN 61547 for EMC. A test report that doesn't carry the model number on it is decoration.

We didn't have a formal driver-approval step in our process until 2022. Cost us when 300 dimmable floor lamps arrived at a hotel and wouldn't talk to the building's lighting control system. The lamps were fine. The drivers were fine. They simply weren't the same dimming protocol as everything else on that floor. We air-freighted 300 replacement drivers and swapped them on site — $4,100 in freight, plus labor, on lamps that were already installed and photographed.

The most frustrating part of this category: two factories will both say "dimmable, no flicker," both mean it sincerely, and only one of them has actually tested it at 5% output with a camera pointed at it.

3. MOQ: the case against factory-direct, and it's stronger than the sales pitch admits

Here's the conclusion that surprises people. For small, urgent orders, the stock route beats factory-direct on the metric that matters, and it isn't close either.

Factory MOQs on floor lamps typically run 200–500 pieces per model per finish. That last part is what trips people up. An order of 60 black and 60 brass of the same led floor lamp design is two MOQs, not one.

Below roughly 150 units per SKU, or below about $8,000 in order value, the sample cost, setup, and consolidation overhead usually eats the factory-direct saving — and that's before you count the 25 days you don't have.

Samples settle a lot of these arguments on their own. Stock route puts a physical unit in your hands in 3–5 days. A factory sample is 7–15 days plus transit. If the decision-maker needs to see the finish before signing off, that gap alone can decide the whole question.

Most "always go factory direct" advice floating around was written for buyers with 40 days and 500 units. Plenty of real rush orders are 40 units and 9 days. For those, no unit-price advantage compensates for a route you can't physically complete.

4. Paperwork: both routes are equally capable of telling you what you want to hear

"CE certified" is not a document. A test report is a document. It names a model, a standard, and a date, and you can check whether the model number on it matches the model number on your purchase order. Roughly half the certificate PDFs I get sent in the first round of a rush project don't survive that check.

Two regulatory points that matter more than they used to, and that stock-and-ship sellers rarely volunteer:

EU Ecodesign regulation (EU) 2019/2020 and energy labelling (EU) 2019/2015, both in force since September 2021, changed what's required of light sources sold into the EU — including replaceability of the light source and availability of spare parts. If your client is buying 400 reading floor lamps for a property with a five-year maintenance contract, an integrated non-replaceable LED module is a contractual risk, not a nice-to-have.

Finish consistency runs on the same logic as print color. Delta E under 2 is the tolerance for brand-critical colors; 2–4 is noticeable to trained observers; above 4 is visible to most people. That applies to powder coat as much as it does to Pantone matching. Get a Delta E value on the lab dip, in writing, or accept that "black" and "black" may be two different blacks across two shipments.

And on quality control: pre-shipment inspection catches cosmetics — scratches, weld spatter, packaging damage. It does not catch a driver that fails at 8,000 hours. Burn-in a sample for 4–8 hours at low dim, and check the driver brand against the BOM you approved.

The trap: total landed cost on a rush order

The unit-price spread between two factories on a $30 led floor lamp is typically $2–4. Air freight on a 5.5 kg lamp carton runs roughly $28–38 per unit at early-2025 rates, against $7–10 by sea. The shipping decision moves the number about ten times more than the supplier decision does.

What I mean is that in a rush, arguing over $2 of unit price while shipping air is a bit like negotiating the taxi fare after you've decided to charter the plane. The line item that decides whether the project is profitable is usually mode and consolidation, not which factory won the quote.

Add the quiet costs while you're at it: factory overtime premium of 15–30% for a compressed production window, and rework cost at destination if a 2% defect rate lands on a job where the installers have already gone home. Timestamp this — freight rates moved 30%+ twice between 2021 and 2024, so any number anyone quotes you is a snapshot, not a rule.

What I'd actually do

Run the scenario, not the ideology:

  • 150+ units per SKU, 30+ days available, dimming spec matters — factory-direct, no contest. Send the full spec sheet, not a photo and a hope.
  • Under 100 units, standard spec, 10 days or less — stock-and-ship. Don't overthink it and don't let anyone shame you into a MOQ you'll regret.
  • Short deadline plus a custom spec — split the order. Take 20–30 units from stock to hit the opening date, run the balance through the factory by sea for phase two. This is the one I use most, and it's the reason I stopped treating this as an either/or question.
  • Mixed finishes on a small order — take one finish now, schedule the second finish after the deadline. Splitting finishes across time is cheaper than splitting them across MOQs.

The fundamentals haven't changed. You still need a spec sheet, a test report that matches the model number, a firm date for when the order enters the line, and a shipping mode you can afford. What changed between 2019 and 2025 is that "go factory direct" stopped being a universal answer. It's one of two routes now, and the right one depends on whether your bottleneck is the production queue or the calendar.

If you're doing floor lamps sourcing against a date you can't move, start with the bottleneck. The supplier list is the easy part.