I Want a Professional Candle Formula Change-Control System — How Should I Handle Supplier Changes, Material Substitutions, Formula Adjustments, Retesting, Approval, Batch Records and Release into Commercial Production?

The system on one page

Every page in this series answers one narrow question: what a new wax invalidates, why a replacement lavender needs its own load test, why an “equivalent” wick is not evidence, why 3 mm of jar diameter matters. This page is the frame they hang on. It covers the four ways a change enters a candle business — a supplier change (the same material by name from a different source), a material substitution (a different material), a formula adjustment (load, blend ratio, wick size, dye dose) and a process change (melter, mixing, batch size) — and the seven stages every one of them passes through before it is finished.

Two design choices make it work in a small business. First, the stages are the same for every change, but the depth is set by a class: a new delivery of the same listing barely touches stages 3–4, while a new wax goes through all seven in full. Second, the system runs on documents the business already needs — a specification per SKU, an approved-supplier list, batch records, a complaint log — plus one new one, the change-control form. Nothing on this page needs software; a folder per SKU and a numbered form are enough to start.

Stages 1–2: four ways in, one front door

Changes rarely announce themselves as changes. A buyer finds a cheaper wax; a supplier runs out of a wick; a jar arrives from a new batch slightly wider; a production assistant starts using a drill mixer because it is faster. The system's first job is to make every one of these come through the same front door: a change request with an exact From and To, filed before the new material or method touches a production batch. The table below maps each entry type to its most common trigger and the posts that own its detail.

The impact assessment in stage 2 is short but it is where most systems fail silently. List every SKU that uses the old material, not the one that prompted the change; mark bestsellers and SKUs in open orders; list the documents that name the material. A wax change requested for one candle often reaches every candle on that wax. The procurement-freedom post explains why a purchasing team's idea of “equivalent” is the commonest unassessed change, and the approved-supplier list is the document that tells them what is already allowed without a request.

Stage 3: the master retest matrix

The master matrix below collects the default status for each type of change, using the series vocabulary — KEEP (still valid), CONFIRM (short side-by-side), REPEAT (full retest), NEW (never needed before). It is a starting point for box 4 of the form, not a substitute for thinking: each hub explains when a cell should move, and the reasons behind each cell are general candle-making practice, never CSI measurements. CSI publishes no melt points, hardness, burn data, wick-size charts, jar dimensions or lot information for any listing, which is why so many cells say REPEAT.

* NEW where the replacement wax has no published fragrance ceiling — in CSI's range, ceilings are stated only for Luxury Soy Wax Chunks (up to 13%) and Luxury Soy Wax CSI 464 (up to 10%). Two rules sit on top of the matrix. One variable per arm: if two things must change together, test them in separate arms against one control, as the wax-and-fragrance-together post shows, and if a candle already failed after two changes, the two-changes post shows how to separate them. Checklists per material: the hub checklists — wax, fragrance, wick and jars, dyes, additives and process — list every line to score.

The standard retest, for any single change

Whatever the change, the full retest follows one protocol, scaled by the matrix. It compares the changed candle with a control — a candle from the approved materials, ideally retained from an approved batch — changing one variable at a time. Every figure in it is a labelled working practice, not a standard; change them and write down what you used.

Acceptance on safety lines is always comparative, across the whole system. CSI publishes no numeric flame-height or jar-temperature limit, so the rule is: no worse than the approved control on every safety line, within your jar supplier's guidance and the recognised candle fire-safety standard you work to — ASTM F2417 is a commonly cited one. A hotter jar or a taller flame is never traded for more throw, lower cost or a deadline; the better-throw-hotter-jar post and the hotter-jar wax post explain why.

Stages 4–6: approval, batch record, release

Approval is a decision against a rule written in advance, by someone other than the requester where the business is big enough to allow it. In a two-person business, the maker requests and the owner approves; in a one-person business, write the rule down before testing and hold yourself to it. The outcomes are pass, fail or conditional — released on partial evidence because a deadline forced it, with the missing tests listed as dated watch items. The 5,000-candle order shows a conditional release used honestly; the 1,000-candle wax emergency and the wick emergency set the minimum evidence.

The batch record for the first production batch after a change is the most important one the business will write that month: exact listings, packs, order dates, your own lot labels, measured jar values, the process actually run. CSI publishes no lot numbers, so your labels are the only trace. Release is not the approval repeated; it checks that production made the approved candle. Burn one or two gate candles from the first batch beside the control, keep retained samples, issue the new specification version, update the approved-supplier list and costing, and mark the cut-over batch. The discontinuation file shows the full documentation for a forced change.

Stage 7: post-release monitoring closes the loop

Monitoring has three parts. Complaints by version: log complaints per 100 sold for the new version against the old one, over a comparable number of units, by category — a new category is a new change request. Watch items: close each dated item from a conditional release with its result. Re-confirmation: approved backups and long-unchanged SKUs are re-checked on a schedule — every six months is a sensible working practice — because materials drift and change notices are not promised by CSI or, usually, anyone else. The backup post covers re-confirmation in detail.

The priced kits: once, and per change

The system needs two kinds of spending. Once: the bench tools that make every number on every record measured rather than remembered. Per change: the standard six-candle retest — three variant and three control — at 200 g and 8%, which needs 96 g of oil (100.8 g with a 5% pour-loss allowance, a labelled working assumption) and 1,159.2 g of wax with the same allowance.

For a year's budget, use production materials rather than test packs. At listed prices this candle costs ₹264.39 in materials (Chunks at the 5 kg rate, the oil at 1 kg, Eco Thin C1, White Matte jar). Take the illustrative year from the form post — 6 class 1 changes at one candle each, 4 class 2 at six candles, 2 class 3 at six candles on each of three bracket SKUs — and the year's validation uses 66 candles, about ₹17,449.61 of materials. That is the whole material cost of running change control for a small range for a year, before time.

The rules that hold the system together

Where to start depends on where you are. A maker with one bestseller should write its specification and batch record this week. A maker with a range should triage it with the 20-SKU post and adopt the form. A business already buying from several suppliers should build the approved-supplier list and the incoming gate. And anyone in the middle of a forced change should start with the hub for the material: wax, fragrance, wick or jar.

Frequently asked questions