Two trays of cluster lashes were sitting on a table at a client exhibition in September. They are same as length, curl, and shade of black. Yet, at the base of one of them, the material is flattened into a strip; where the fibers have been pressed together, a faint seam is still visible. The other, a 3D printed laser cut cluster, looked drawn in a single line. The buyer picked both up, said “this one feels lighter”, and then asked the question the whole meeting turned on.
How do you put that difference into a purchase order?
That is the awkward part about laser cut cluster lashes. The difference is real. You can feel the difference just by rubbing the lashes between two fingers. Yet, the specifications buyers send us usually consist of only four parameters: length, curl, thickness, and quantity. Suppliers can fulfill orders based on these, and the paperwork might match up, but the essential qualities that originally drew the brand to the product in the first place are often missed.
This article aims to bridge that gap. We’ll explore shifts in the development process, identify which metrics are truly worth tracking, and discuss why certain claims on product pages eventually lose their effectiveness. The focus here is on B2B product development for both the professional and DIY (at-home) lash markets—not on how to apply the lashes, nor on the specific weight or volume that an individual client’s natural lashes can safely support.
First, what the process actually does to a lash
Stamping presses fibers together. Laser cutting separates them.
Most cluster lashes start as fibers laid in a row and bonded at the root. The root is then cut to the shape the machine was set up for. It works, and it has worked for twenty years. The limit is physical: a bonded root has a thickness made of adhesive plus compressed fiber, and the seam where they meet does not always disappear.
Laser cutting approaches the same tray from a different direction. A fiber sheet is cut by light into a finished cluster, so the base is formed by the cut instead of glued together afterward. On A-RIX cluster lines the resulting base measures 0.035mm, and the profile is curved rather than flat across the width of the eye.
That curve is the part buyers notice without being told. A flat base has to bend to follow a lid. A pre-curved one doesn’t.
What changes on the tray, and what stays exactly the same
Two things change. The base gets thinner, and the outline of the cluster gets cleaner, because a cut edge has no loose fibers hanging off it.
Everything else is ordinary. Length still comes in the usual 7 to 16mm spread. Curl chart is still your curl chart. Color is still black, brown, or whatever your range runs. Density is still a styling decision. If a supplier tells you laser cutting replaces any of that, ask what the machine is supposed to be doing to the curl.
One real limit: laser cut trays tend to live in the cluster category. A 0.035mm base makes sense for a product a consumer places and removes. It is a different conversation for a volume fan a lash artist builds by hand, because the handling is different, the bond is different, and the base has to do a different job.
The band number is a measurement, not a slogan
Our own product pages disagree by a factor of ten
Here is an uncomfortable example, and it comes from our own catalogue. One page for the self-adhesive laser cut cluster lash lists each cluster at 0.02g. A sibling page on the same site lists 0.002g. One of those is a typing error. Which one, we can’t fix by arguing on the internet, and it is a factor of ten.
The lesson is not that someone mistyped. The lesson is that weight and band thickness get repeated across listings, marketplaces, and distributor catalogues without anyone re-measuring them. By the time a figure reaches a brand’s packaging copy, nobody in the chain may have seen the original test.
So treat every number on a product page, ours included, as a claim to verify rather than a fact to publish. Ask for the method. Ask what instrument. Ask how many pieces were weighed, and whether the average includes the adhesive.
How to check a base without a lab
You will not measure 0.035mm with the calipers in a desk drawer. The jaws themselves are usually thicker than the thing you are measuring, and hand pressure moves the reading by more than the difference you are trying to see.
Do this instead. Put the sample on a clean white card in daylight and photograph it straight on from the side, with a coin or a ruler in frame for scale. Then photograph the same shot with your existing base. Two images, one scale, no magnification. If you cannot tell them apart at that size, then for your customer the difference may not be visible either, and that is worth knowing before you pay a premium for it.
The second test costs nothing. Take one cluster, hold it by the tips, and drag the base across the back of your hand. A thick bonded base drags. A cut base glides, or it catches on nothing at all. Crude, but it separates the two constructions faster than a spec sheet does.
At the bench, a thinner base mostly changes two things
Corners and the curve
Inner corner lift is the complaint that gets a cluster brand its first one-star review. The cause is usually mechanical. A straight strip is being asked to sit on a curved lid, and the tension has to go somewhere. It goes to the corner.
A pre-curved 0.035mm base has less stored tension to release, so the corner has less reason to lift. That is a geometry argument, and geometry arguments hold up better than most marketing claims. It still is not a guarantee. Lid shape varies, the adhesive matters more than the base, and how the wearer sleeps matters more than both.
Be careful about how you use this in your own copy. “Designed to sit closer to the lash line” is defensible. “Won’t lift” is a complaint waiting to be filed.
Pickup, and the pull
The other change is at the tweezers. A thin cut base gives the tweezer tip a defined edge to get behind, and that edge does not fray. In a tray with a sticky strip, that means fewer clusters torn in half during removal.
Test it during sampling rather than trusting the description. Pick up ten clusters with the same tweezer, in the same way, at the same angle. Count how many come up clean and how many come up bent or split. Ten out of ten is a good tray. Seven out of ten is a tray that will generate returns.
While you are there, look at the removal. A base that releases cleanly from the strip usually releases cleanly from the lash line too. Not always. Usually.
PET memory, and the 10 to 15 use claim
What that number actually depends on
Reuse figures are the least standardized number in this category. Ten to fifteen wears per cluster is achievable with PET fiber, and it is also achievable with a well-made PBT cluster. The fiber is not the only variable.
Three things move the number more than the material does. How the wearer removes the cluster, because peeling from the tip breaks the spine. Whether the adhesive is cleaned off the base or allowed to build up, since a crusted base stops sitting flat. And storage, because a cluster loose in a makeup bag is a cluster with bent tips by week two.
So when a supplier quotes a reuse figure, ask what conditions produced it. If the answer is a lab in controlled handling, the number describes the product and not the customer’s bathroom.
A curl test a brand can run in an afternoon
You need six clusters, a kettle, and a phone camera.
Photograph the six from the side against a grid, all at the same distance. Steam them for thirty seconds. Let them dry at room temperature for an hour. Photograph again from the same position. Then do the same with your current base as a control.
If the curl angle has moved on the laser cut and not on the control, or the reverse, you have something to discuss with the supplier that is not opinion. If both moved the same amount, then the “curl memory” line in your marketing should probably come out, because your customer will find out.
Writing the process into a purchase order
Fields that should be on every cluster
Most lash POs carry six fields. For a laser cut cluster, add these:
- Base thickness in millimeters, with the measuring method named
- Base profile: flat, curved, or specified radius
- Fiber material and designation (PET, PBT, and the exact grade if the supplier will name it)
- Cluster weight, with the sample size the figure came from
- Strip type: sticky strip, slide-out tray, or loose in a compartment
- The revision number of the approved sample
That last one does more work than the rest combined. Two trays can arrive in identical packaging with a revised base between them. A revision field on the artwork and the PO gives the person opening the carton something to check before the first cluster hits a lash line.
Where batch variation actually shows up
Not usually in the base thickness. Cutting is a controlled step, and a 0.035mm target holds better than most buyers expect.
Variation tends to appear in three places. First, the cut edge itself, where a worn blade or a blade approaching the end of its service life leaves a slightly ragged outline rather than a clean one. Second, cluster-to-cluster alignment in the tray, because a cut sheet that shifts during handling produces a tray that looks uneven even when every cluster is individually fine. Third, and most common, the base curvature drifting slightly between lots, which nobody notices in a photo and everybody notices in a set.
Ask your supplier how they inspect for the third one. If there is no answer, that is your answer.
MOQ, cost, and where laser cut belongs in a range
100 boxes as a first run
Laser cut cluster lash work at A-RIX starts at 100 boxes, low enough to test a launch without committing a season’s cash to it. Standard private label work on other lines typically starts at 1,000 boxes, so the two numbers get confused in conversation. Check which one applies to the product you are actually quoting.
Sampling usually runs around seven days from confirmation. Add your own review time, then a revision round if the base needs one, then production. A brand that books sampling in week one and expects stock in week four is planning a rush, and rush plans are where artwork approvals get skipped.
When the premium is not worth it?
The honest answer: sometimes it isn’t.
If your customer is price-led, buys one pack at a time, and mostly compares on count per box, a thin base is a feature you will pay for and then have to explain. Standard clusters work. The market has proved that for two decades.
The premium earns its place in three situations. A brand selling reusability, where the base has to survive removal and reapplication. A brand whose reviews are already carrying corner-lift complaints. And a brand competing in the premium DIY segment, where the unboxing and the first touch are part of what the customer paid for.
Notice what is not on that list: the word “innovation”. Nobody pays for a process. They pay for the thing the process fixes.
Questions buyers keep asking
Is a laser cut cluster automatically better than a bonded one?
No. It is better at a specific set of jobs. Thin base, curved profile, clean edge, low tension at the corner. If your customer’s main complaint is price or volume per pack, laser cutting does not address it.
Can we mix the two in one collection?
Yes, and it is common. Many ranges use laser cut for the hero SKU and a bonded base for the volume SKU. Just label the difference on the tray, because the two will otherwise look identical on a shelf.
What should a brand send with the first inquiry?
The intended use, a reference for the texture, the base construction you are considering, and the fields above with the ones you are still deciding marked as open. Marking a preference as a decision is how a sample round gets wasted, and sample rounds are the cheapest part of the project to get right.
One last thing to do before you send the brief
Hand the brief to someone who has never seen the mood board. Ask them to identify the approved sample from the description alone. If they can point to it, you have a specification. If they hesitate, or if their answer depends on the word “wispy”, you have a description of a feeling. Those two documents produce very different trays.
For development questions on 3D printed and laser cut cluster lines, the A-RIX team can send physical samples and the base construction used on them. Physical samples are still the fastest way to settle an argument about 0.035mm.
