Trial by Parcel: Costing Product Sampling as a UK Direct Mail Channel

Product sampling by post is frequently presented as a neat extension of a direct mail campaign: add a sachet, a miniature or a trial-size unit to the envelope and let the doormat do the selling. The appeal is understandable. A physical sample can create a sensory encounter that print alone cannot, and it can shorten the distance between awareness and trial. But the commercial logic is not the same as a letter. A sampling mailpack carries unit cost, weight-band postage, kitting labour, damage risk and a conversion burden that must be met by actual trial and repeat purchase. UK brands should treat it as a cost-per-trial channel, not as a free add-on to a direct mail plan.

The doormat is a trial channel, not a bonus line

The first discipline is to separate the sample from the message. A mailpack may contain a letter, a voucher, a QR code and a product unit, but only the unit creates a trial. The letter and envelope create attention and context. The sample creates the experience. Those are different jobs, and they should carry different cost assumptions.

Too many sampling plans are signed off on a unit-cost comparison: this miniature costs 80p, that sachet costs 45p, so the cheaper unit must be more efficient. That logic ignores the postal consequences. A sample changes the weight, thickness and rigidity of the pack. It may push the mailing from a letter format into a large letter or parcel format, alter the sortation route, require additional cushioning and increase the risk of damage in the postal system. The cheapest unit can therefore produce the most expensive delivered trial.

The weight band is the first commercial decision

In the UK, postage is not a single rate applied to a campaign. It responds to format, weight and dimension, and the step changes between bands can be severe. A sampling pack that sits just above a weight threshold may pay a disproportionately higher rate for every unit. The same effect appears when a rigid sample makes the pack non-machineable or requires a box rather than an envelope.

Marketers should ask for postage modelled at the pack level, not at the sample level. The question is not what the unit weighs in isolation. It is what the complete mailpiece weighs once the sample, outer, reply device, cushioning and any protective film are assembled. The answer should be tested with the actual mailhouse and the actual postal contract. A few grams can matter more than a few pence of unit cost.

Breakage, leakage and the units that never trial

A dispatched sample is not a delivered trial. Units can be crushed, split, leak, expire, be misdelivered or arrive unusable. Some will be discarded before the recipient understands what they are. The commercial model must use a denominator of usable trials, not units posted.

This is where sampling differs most sharply from a standard direct mail campaign. A damaged letter still carries a message, albeit a poor one. A damaged sample may carry nothing but a mess. For liquids, powders, fragile items and anything with a scent or texture, the breakage rate should be measured in a pre-test rather than assumed. A 3% to 7% unusable rate is not unusual in physical sampling, but the figure must be evidenced for the specific unit and pack. If the unit is expensive, the replacement allowance and the reputational cost of a poor arrival both belong in the budget.

Fulfilment is where the model gets lumpy

Sampling introduces a fulfilment operation that a standard mailpack may not require. Units must be received, checked, stored, kitted and inserted. Some will need hand assembly. Some will need batch codes, expiry checks or regulatory labelling. The mailhouse may be able to insert a flat sample, but it may not be set up to handle a fragile or oddly shaped one. That can mean a separate co-packing supplier, a split delivery and an additional quality-control step.

These costs are real, and they rarely scale down neatly. Minimum run lengths, setup time and storage can make a small test disproportionately expensive. That does not make the test wrong. It means the test budget must include the fulfilment overhead rather than pretending it is part of the mailhouse's standard insertion rate.

A cost-per-trial model UK marketers can defend

An illustrative model is useful, provided it is treated as a framework rather than a benchmark. Suppose a brand plans 10,000 sampling mailpacks. The unit costs £1.80, the primary pack £0.55, the outer mailer £0.45, postage £1.35, fulfilment and labour £0.70, data and suppression £0.12, and a damage or replacement allowance £0.20. The total is £5.17 per unit dispatched. If 6% of units are undeliverable or unusable, the cost per usable trial is about £5.50.

That figure is the starting point, not the verdict. The brand must then compare it with the gross margin it expects from a first purchase and from repeat purchases. If the first-order gross margin is £15, the trial-to-purchase conversion needed to recover the sampling cost on the first order is roughly 37%. That is a demanding threshold for many categories. If repeat purchase is central to the business case, the calculation should use a defensible lifetime margin rather than an optimistic first-order figure.

The model should also separate the cost of the sample from the cost of the response mechanism. A QR code, a unique voucher code or a matched control group adds cost, but without one of them the brand cannot tell whether the sample caused the purchase or merely coincided with it.

The thresholds that should stop a sampling plan

A sampling plan should be paused or redesigned when the cost per usable trial exceeds the gross margin available from a realistic conversion rate; when breakage is high enough to undermine the recipient experience; when the pack cannot be produced within the postal format assumed in the budget; or when the fulfilment operation requires a minimum volume that the test cannot support. These are commercial thresholds, not creative objections.

For 2026, the brief should ask for landed cost per usable trial, not cost per unit. It should require a damage allowance, a replacement policy, a measured conversion route and a control group. It should also ask what happens to the units that never reach a doormat. Sampling can be a powerful direct marketing channel, but only when it is costed as one. The sample is the product experience. The trial is the commercial event. The budget should follow the trial.