Call(WhatsApp Us)
+8615360806100Email Us
cherry@laeyo.comSee the factory by live video walkthrough
Book an on-site factory visit in GuangzhouAlmost every skincare brief comes back with the same line: “We do everything in-house — mixing, filling, and packing.” That sentence is doing a lot of work. It can mean a single site with…
Almost every skincare brief comes back with the same line: “We do everything in-house — mixing, filling, and packing.” That sentence is doing a lot of work. It can mean a single site with dedicated tanks, its own filling lines, and a packing hall under one quality management system. It can also mean a trading office that buys bulk from one plant, fills at a second, and cartons at a third. Both can deliver a good product — but only one of them gives you the traceability, change control, and accountability you are actually paying for. This guide gives you the specific questions, documents, and acceptance criteria to separate the two before you commit a deposit, a mould, or a launch date.
Before you ask “do you own the factory?”, define the scope you are verifying. For skincare, in-house control normally needs to cover five things, not three:
If any of these sit with a subcontractor, that is not automatically a rejection — but it must be declared, and the subcontractor must appear in the documentation chain. Undeclared outsourcing is the problem, not outsourcing itself.
Mixing is where a lab sample becomes a product — or stops being one. Ask how the plant controls the transition from bench batch to production batch, and what it measures in-process.
Filling is the step buyers most often verify emotionally (“I saw the line”) and least often verify numerically. Numbers are what protect you.
Final packing is usually where in-house control is thinnest, because it is labour-heavy and easy to sub-contract quietly. Verify it directly.
| Process stage | What “in-house control” must mean | Evidence to request | Acceptance criteria |
|---|---|---|---|
| Compounding | Own tanks and homogeniser, own batch records, in-process pH/viscosity checks | Redacted executed batch record; equipment list with capacities | Every critical parameter has a recorded value, limit, timestamp, and QC signature |
| Water system | On-site purified water generation and monitored loop | Water trend report; sanitisation log; microbial and conductivity limits | Results within specification across the last 3 months with no unexplained excursions |
| Cleaning / changeover | Validated cleaning with verification before next batch | Cleaning SOP plus a completed verification record | Acceptance criteria defined (e.g., rinse water, visual, TOC/conductivity) and met |
| Filling | Line qualified for your format and viscosity | Line list; trial report for your container; in-process control plan | Fill weight check frequency, limits, and corrective action defined; first-off approved vs golden pack |
| Package compatibility | Full system tested as one unit, not component by component | Compatibility report: formula in final pack, filled at production dose | No leak, no deformation, no dose drift, no label lift after the defined test duration |
| Final packing | Packing hall on the same site, under the same QMS | Batch record packing section; artwork version log; retained coded sample | Artwork version and lot code match the approved master; reconciliation within tolerance |
| Traceability | One record links bulk lot → components → filled lot → shipper | Mock recall or traceability exercise report | Full upstream and downstream trace achievable within an agreed window (commonly 2–4 hours) |
| Micro and stability | Testing linked to the specific batch and pack, not a generic dossier | Stability protocol and results; preservative efficacy data; micro results per batch | Protocol duration and conditions stated; results traceable to your formula version |
A common and expensive mistake is approving a container on appearance and a formula on feel, then discovering in the field that the two do not work together. A skincare pack only counts as validated when the final formula, the primary container, the closure or pump, the seals, the decoration, the filling process, and the transport environment have been tested as one combination.
So when a factory says “in-house”, confirm that the compatibility protocol was run in their own filling hall, with their own filled units, at the production dose — not on a bench sample from a lab.
If you can visit, ask for this route. The purpose is not inspection theatre; it is to see whether the paperwork you were sent matches the floor.
Expect the evidence review to add roughly one to three weeks to project setup, and on-site verification one to three days of travel and scheduling. Set against the alternative — a stability failure found after 5,000 units are filled, or a label version error discovered at the retailer — this is usually the cheapest line in the budget. Sequence it early: lock the formula version and the pack system first, then run compatibility and stability, then approve artwork, then pilot, then production. Compressing stability is possible only if you accept the residual risk in writing.
No. A certificate tells you a site has a documented quality system within a stated scope. It does not tell you which unit operations happen on that site for your SKU, nor whether your specific formula and pack were validated. Pair the certificate with an executed batch record, a compatibility report, and a clear written statement of which steps are subcontracted.
Ask for the in-process control plan and two or three completed control sheets from comparable runs. You want to see the check frequency, the specification limits, the recorded values, and what happened when a value drifted. A factory that can only send a specification, and not a completed record, has not shown you control.
Not at all. Most factories buy containers and pumps from specialist suppliers, and that is normal. What matters is that the compatibility and filling validation were performed in the factory’s own hall with the final components, and that the component supplier’s technical data and version control are part of the record.
It can work, but the risk shifts. You then need the bulk manufacturer’s batch record, micro results, and specification alongside the filler’s records, plus a documented bulk transport and hold-time validation. Many brands accept this only for simple, well-characterised formulas; for emulsions and preservative-sensitive products it is harder to defend.
Ask for the golden pack to be signed and dated by both parties, with the formula version and component version recorded on it. Then confirm that each production batch retains units coded with the same lot number as the batch record, and that the first-off sample was approved against that golden pack before the run continued.
If you would like to run this verification against a live project, send your target markets, SKU list and formats, positioning, packaging direction, and target launch date. With those five inputs we can tell you which documentation already exists, which tests still need scheduling, and where the realistic critical path sits. Request a Quote.
This article is informational only and is not legal, regulatory, or metrology advice. Requirements differ by market and by product; confirm obligations with your Responsible Person or qualified counsel.