One brand, two plants: keeping an own-brand product identical across sources

Splitting the volume weakens your position at both plants

Whatever quantity earned your current price gets divided the moment a second source is added, and both halves may land in a less favourable band. You also hold two production lots of the same item instead of one, which raises the cash tied up and the storage consumed without raising what you can sell. Neither plant now treats you as an account worth protecting when their capacity is short. Brands frequently discover that the arrangement built to reduce fragility has instead moved them to the back of two queues. Volume, not sentiment, buys priority on a production schedule.

A specification written for people who already know the answer will not travel

The document that governs your product at its original plant is usually incomplete, because the operators there fill the gaps from memory: the mixing sequence, the tolerance nobody wrote down, the raw material they always use even though the spec permits alternatives. Handing that document to a second site produces something adjacent to your product rather than your product. Portability means writing down what is currently assumed — grade and origin of each input, which process settings genuinely affect the outcome, the limits at which a batch is accepted, the retained reference sample — and testing whether a stranger reading only the paperwork arrives at the same result.

Customers notice the drift before your inspection does

Colour, scent, texture, thickness, fit, the sound a closure makes: consumers register these comparatively, against the unit they bought last time, and they do it far more sensitively than an incoming check with a pass-or-fail limit. Two plants working legitimately within the same tolerance band can sit at opposite ends of it, and the shopper experiences that as a product that changed. Hold a golden sample, narrow the acceptance range beyond what either plant would offer unprompted, and get first production from the new site compared side by side with current stock before any of it reaches a customer.

Two sources mean two evidence trails and two sets of records

Test reports, assessments and supporting documentation are generally tied to a specific formulation made in a specific place, so a second site frequently means rebuilding part of the evidence pack rather than photocopying it. Requirements differ by product and by market and this is orientation rather than legal guidance, but plan on the assumption that duplication is not free. Batch coding matters just as much: when a complaint arrives, you need to identify which plant made that unit within minutes, not after a search through email. Design the code before the first run, not after the first problem.

Often a single qualified plant plus a rehearsed exit beats permanent duplication

The real question is what a stoppage would do to you. If losing supply for a period is survivable, the cheaper structure is one plant you know well, a fully documented and portable specification, a modest cushion of finished stock, and a shortlist of alternative sites that have already seen the paperwork and quoted against it. If a stoppage would cost a retail listing you cannot regain, duplication earns its cost. Decide from the consequence of interruption in your specific channel rather than from a general preference for having a spare.

Frequently asked questions

Do I need a backup factory before my first product has proved itself?

Rarely. Qualification consumes cash and attention that an unproven product cannot justify, and splitting a small order across two sites weakens both relationships. What is worth doing early is cheap: keep the specification written well enough that another plant could quote from it, and know who else in the category could make the item. That preserves the option without paying for it until demand shows the product is worth protecting.

How do I stop two plants drifting apart on the same product?

Control the reference rather than the recipe alone. Keep retained samples from an agreed benchmark batch, send them to both sites, and make acceptance a comparison against that benchmark rather than a reading against a wide tolerance. Sample from every run at both plants and compare them to each other, not only to the paperwork. Where an input is supplier-specific, name the supplier in the specification rather than describing the material generically and hoping.

What usually breaks first when a second supplier is introduced?

Packaging and artwork, more often than the product. Each site has different print partners, tolerances and pack sources, so the same artwork file emerges in a slightly different colour, on a slightly different substrate, on a bottle whose shoulder sits marginally differently. Consumers compare packs on a shelf far more readily than they compare contents, so lock the pack specification and print standard across both sites before you worry about the formulation.