Reduce packaging costs: the five levers and what they actually deliver

Why packaging costs are hard to pin down

In many ranges packaging sits between five and fifteen per cent of product cost — high enough to matter, low enough to slip through. That middle position explains most of it.

Three patterns recur. Demand is fragmented: boxes here, film there, labels through a third order, often separately per site. Together a relevant category; on any single invoice, unremarkable. Price is checked against last year rather than against the market — which makes a three per cent increase look moderate even when the raw-material index fell over the same period. And ownership is split: procurement owns the price, logistics the handling, marketing the appearance. Whoever would have to think about all three is rarely named.

On top of that comes an effect you hear repeatedly from purchasing managers: packaging is the line item people feel safe about, because they have known the supplier for years. Familiarity, though, is not a price argument.

How much is realistic — and what it depends on

There is no sensible flat answer. Anyone giving one is either selling something or has not looked. Two axes determine the span, and they act independently.

First axis: the industry

Where the specification is tight, there is little room in the material — the lever shifts to price and volume. Where volume and freight dominate, the specification itself is the biggest lever.

Second axis: the role packaging plays for you

This distinction is missing from almost every guide on the subject, and it explains more than the industry does.

Packaging is strategic when it sells alongside the product — brand, unboxing, shelf presence. Then price is not the only criterion, a supplier change is risky, and the design constrains. The room in unit price is smaller. What gets overlooked instead, regularly, is another one: separating brand-relevant from non-brand-relevant items. The outer carton the customer sees has to carry the brand. The shipping box around it, containing that carton and the void fill, usually does not — and is nonetheless often procured in the same breath and from the same supplier.

Packaging is tail spend when it runs along as a C-item. Rarely tendered, incumbent supplier for years, prices drifting unnoticed. This is typically where the widest spans sit — not because it was negotiated badly, but because it was never negotiated.

The tail-spend case is also the harder one. The potential is large, but mandate, data quality and ownership are missing. The blocker is not the saving potential but the internal effort to realise it.

Which is precisely why an external, data-driven approach pays off most in this case: it substitutes the capacity that will not be freed up internally for a side category.

What the numbers show

Across all PAXLY customers, the 2025 saving came to a median of 23.4 per cent. More important than the median is the spread behind it: individual results range from 17 per cent at s.Oliver (fashion) to 35 per cent at Flaconi (e-commerce), with DEVELON at 30 per cent (industry), Burda at 25, frischli at over 20 (food) and SellerX at 20 in between.

That spread is not random — it follows the two axes above. The median is therefore an average across a bandwidth, not a commitment for any single case. Anyone quoting you a figure without first asking about your industry and the role of packaging is guessing.

The five levers at a glance

The levers work at different speeds and to different depths. Four concern procurement, one the product itself.

Lever 1: create competition

The single biggest lever is almost always to create competition at all. Not a harder negotiation with the incumbent, but the question of who they are actually competing against.

A direct enquiry to the familiar supplier is not a tender. It produces a price but no benchmark. It only becomes meaningful once several suppliers cost the same specification and the quotes come back comparable line by line. The effort involved is why it does not happen — not a lack of insight.

On rhythm: every twelve to eighteen months per item, regardless of trigger. Tendering only on complaint or price increase means measuring yourself against your own previous year. Bringing it forward makes sense at an index gap of roughly seven per cent, at conspicuous delivery reliability across two quarters, or at marked growth in demand. How that runs operationally is described on the page about structured tenders.

Lever 2: check against the market, not against last year

Comparing two quotes tells you which is cheaper — not whether both are too expensive. That needs a reference point outside your own supplier relationship.

Recognised raw-material indices provide it. In practice this bites in two places. In quote assessment, it shows whether a price matches the market level or whether you are paying a list price — the gap between list and market price is typically between 5 and 20 per cent and forms the core of the negotiating margin. On announced price increases, the demand can be set against the actual index movement instead of arguing over percentages.

For composite products such as printed folding cartons or labels, one step comes first: separate the material share from the finishing premium. Otherwise you are comparing two numbers made of different components. Which indices are useful for what is set out under price index.

Lever 3: put the price adjustment in the contract

The tender sets the price level once. The clause determines what happens over the two to three years that follow. Across the term, the second part is often the larger — and is usually decided in passing.

An index clause links the unit price to a raw-material index, with a threshold (typically around five per cent) and a fixed periodicity. The adjustment is then calculated rather than negotiated, and the annual price round disappears. What matters is symmetry: the clause has to work in both directions. That is where most fail in practice — a clause built too simply can be exploited systematically to your disadvantage through the asymmetry between rising and falling raw-material prices.

The order of magnitude of that effect is around 2.5 per cent per adjustment, with a negative compounding effect over the term. The mechanics behind it — which components belong in a sound clause, and how to spot an unsound one — are covered in the article on price indexing.

Lever 4: bundle volumes

The same box, ordered three times by three sites, is three small orders. Bundled, it is one medium-sized one — and gets a different price.

Typical consolidation effects sit at 8 to 15 per cent better terms, accompanied by a reduction in supplier numbers of 30 to 50 per cent. The second figure is the underrated one: fewer suppliers means fewer contacts, fewer framework agreements, fewer places where prices drift unnoticed.

Bundling runs in two steps, and the first is a data step, not an organisational one: first a shared view of demand instead of ten spreadsheets, then joint procurement. Start with step two and you end up debating responsibilities without knowing what over. More on this under supplier management.

Lever 5: review the specification

The only lever that applies to the product rather than to procurement — and the only one that works twice: on material and on freight.

In short: box ranges grow historically and are rarely calculated against actual order history. Oversized boxes cost material, void fill and — because many carriers bill by volumetric weight — freight. How to approach that systematically is set out in detail under right-size packaging; here it stays at its place as the fifth lever.

What comes first?

The order depends on the same question as the size: strategic or tail spend.

If packaging runs along as a C-item, start with inventory and competition. First a complete list of every item with part number, material class and supplier — without it you can neither prioritise nor quantify. Then tender the largest items by volume. Market comparison and clause follow in the same move, because both belong in the award anyway. Specification comes last: it is the slowest lever and needs data that the first step generates in the first place.

If packaging is strategic, reverse the order. It starts with the separation: which items carry the brand, and which do not? The non-brand-relevant part can then be treated like tail spend — tendered, bundled, indexed. The brand-relevant part is optimised through specification and total cost, not through a change of supplier. Throw both into one pot and you risk either the brand presentation or leaving money on the table for the rest.

In both cases: the fastest measurable effect comes from levers 1 and 4, the most durable from lever 3. Lever 3 does not work immediately — it works across the entire contract term.

Where it breaks down in practice

Three patterns, in order of frequency.

The incumbent as the default. Not out of convenience, but because a change means work and a mistake would be visible. The remedy is not a change at any price but a tender the incumbent takes part in — they frequently stay and become cheaper anyway.

Unit price instead of total cost. Minimum order quantity, tooling, lead time, payment terms, delivery and print costs all shape what an item really costs. A quote leading on unit price can end up behind on the full calculation.

Confusing the purchasing moment with the term. The tender is a one-off event; indexing carries the following two to three years. Treat them as the same and you optimise the wrong lever — then wonder why a well-negotiated price sits above the market again after eighteen months.

A fourth point concerns the choice of sources: a substantial share of the advice on this topic comes from companies that sell packaging. That does not make the advice wrong, but it bounds it — no distributor will recommend halving your demand, or tendering it, as the first lever. The test question is simple: does the author earn money on the material they are writing about?