The parcel is the one channel every customer opens
Most retention channels need permission. Email and text reach only the customers who agreed to them, and the e-commerce email and SMS flows guide builds within that limit. Ads reach only the customers the platforms can still match. The parcel reaches everyone who bought, including the customers who ticked no boxes and the gift recipients who never visited the store. It arrives when the customer has chosen to pay attention, often opened by hand and examined piece by piece.
That makes the parcel the one owned channel without a list. It also makes it a channel with no second chance. An email that misses can be followed by another; a card that misses has been read once and thrown away. The repeat purchase and loyalty guide names packaging that does not disappoint as part of the first experience that decides whether a second order follows. The post-purchase email flow tutorial builds the messages that follow an order for customers who can be emailed; the parcel is where the store speaks to every customer, emailable or not.
| Channel | Reaches | Needs permission | Can follow up | Main job after an order |
|---|---|---|---|---|
| The parcel and its insert | Every buyer and gift recipient | No | No | First impression, first use, one next step |
| Service emails | Every buyer with an email address | No, while they stay free of promotion | Yes, within the order's progress | Order, dispatch, delivery and help |
| Marketing emails and texts | Customers with a recorded basis for marketing | Yes | Yes | Next purchase, loyalty, reviews |
| Retargeting ads | Customers the platforms can match | Depends on tracking consent | Yes | Reminders and the next purchase |
Packaging as part of the product
Packaging has three jobs, in order: get the product there undamaged, make opening it easy, and make it recognisably the brand's. A product that arrives broken, crushed or leaking is the worst first impression a store can make, and it comes back as a return or a complaint; the return rate entry explains how returns are counted and why they belong in every channel's numbers. Damage is the first thing to read in the store's own returns and complaints data before spending anything on how a box looks.
Opening comes next. A parcel that needs a knife, sheds filler across the floor or hides the product under layers of wrapping turns the moment into a chore. Then identity: the colours, type, tone and imagery that the brand identity and voice guide sets out apply to the box, the tape, the tissue and the card as much as to the website, and that guide notes that packaging has its own shape and constraints. Consistency matters more than expense. A plain box with a well-designed card and a note in the brand's voice can feel more considered than an elaborate one that looks like nothing else the brand makes.
Some customers share what they find in the box, and that user-generated contentContent about a brand made by customers or creators rather than by the brand itself. can become social proofEvidence that other people have chosen, used or approved something, used to reduce a buyer's doubt. for the next shopper. The user-generated content guide covers asking for it and getting permission to reuse it. Design the opening for the customer, though, not for a camera: a box that is pleasant to open is shared more often than one that was designed to be filmed.
What an insert should ask for
An insert that asks for everything gets nothing. A card carrying a discount code, a social handle, a review request, a newsletter offer and a referral link gives the customer five decisions to make and no reason to make any of them. Choose one job for each insert, and let the job follow from what the customer most needs at that point:
- First-use help. Care instructions, a quick start, a recipe, the mistake new owners make. For products that need setting up or learning, this is often the most useful insert, because a customer who succeeds with the first product is more likely to buy again.
- The next purchase. A route back to reorder the same product, or to the item customers usually buy next. For consumables, a link straight to reordering the same product is often more useful than a general discount.
- A referral. An invitation to pass something on to a friend, built on the mechanics the referral programme tutorial sets out.
- The loyalty programme. How to join and what it gives, for stores that run one; the loyalty programmes guide covers whether a programme is worth having at all.
- An honest review. A request to every buyer, under the rules in the next section.
The job can change with the order. A first-time buyer may get first-use help; a repeat buyer may get the loyalty card; a customer whose order contained a consumable may get the reorder route. Where the packing process allows it, choose the insert from what is known about the order rather than sending one card to everyone.
A sample of another product is a different kind of insert, a product rather than a message, and it should come with a card that explains what it is and where to buy the full size. The guide to samples and trial sizes covers when a sample in another order is worth its cost and how to follow it up.
Sign-ups and permission
An insert can invite the customer to join the email list. It cannot create the permission itself. The list building and consent guide describes the UK regulator's direct marketing guidance as allowing a business to market similar products to existing customers only in a limited case, where they were given a chance to refuse at the point of sale and in every message; beyond that, electronic marketing needs consent. The order itself is not a subscription. So the card points to a sign-up page, and the page asks clearly, explains what the customer will receive and records the answer. The customer's own details then become first-party dataData a business collects directly from its own customers, visitors and users. with a known basis, rather than an address added to a list because it happened to be in the order system.
Review requests that stay within the rules
An insert is a natural place to ask for a review, since it arrives with the product. The rules on how to ask are the same as for email. The reviews and social proof guide sets out the general principles from the FTC's guidance: a store may offer something for leaving a review but not for leaving a positive one, the incentive has to be offered regardless of what the customer writes, and incentivised reviews must be identified as such wherever they appear. The FTC's Endorsement Guides questions and answers cover material connections such as payment or free product in the same terms. The service recovery and complaints guide records that the CMA has made clear that faking reviews, or suppressing genuine negative ones, is unacceptable.
On a card, that means:
- Ask every buyer. A card that goes only into the parcels of customers expected to be pleased selects the reviews as surely as deleting the bad ones would.
- Ask for an honest review, in those words. Not a five-star review, and no picture of five stars.
- Keep any incentive unconditional and disclosed. If there is a reward, it is for reviewing, whatever the review says, and the reviews that earned it are labelled.
- Do not combine the request with a discount for the next order. The post-purchase flow tutorial keeps the two apart in its emails, and a card that mixes them invites the reading that the discount is a reward for praise.
- Ask at the right time. The card arrives before the customer has used the product, so a card that says when to review, after first use, usually works better than one that asks for a review on opening.
This is general direction drawn from regulators' guidance, not legal advice, and the rules in each market where the store sells should be checked.
Inserts in marketplace orders
An insert in a marketplace order is a way to introduce the brand's own store to a buyer who belongs, for now, to the marketplace. The marketplaces versus own store comparison treats packaging inserts, product registration, warranty activation and care content as ways to give marketplace buyers a reason to visit the store, within the marketplace's rules on contact. Those rules belong to each marketplace, they change, and the site does not document them, so this guide states no specific rule. Read the current policies of each marketplace before designing an insert for its orders, and keep the card's job to what the policy permits. A card that helps the customer use the product is the least likely to conflict with anything.
Tracking inserts and QR codes
A card that nobody can trace back cannot be judged. Give every insert a destination that identifies it: a short, memorable address on the store's own domain that redirects to the right page with campaign parameters attached, or a QR code that carries the tagged address. The naming conventions and UTM guide sets the rules for the UTMURL tags identifying traffic source. tags, so that each insert version reports under its own name. The guide to measuring offline media explains why tagged visits from printed material are a floor, not the whole response: many customers will type the main address or search for the brand instead.
Two more signals fill some of the gap. A question at checkout or after the next order, asking where the customer heard about the offer or what brought them back, is the method the self-reported attribution guide sets out, with its warnings about how people remember. And a unique code printed on one insert version shows how many customers used it, though only those who used it.
Testing an insert
The question an insert test answers is whether customers who received it bought again more often than customers who did not, or who received a different insert. Scans and code redemptions cannot answer it, because they count only the customers who acted on the card, and many of them would have reordered anyway.
Two designs work for most stores. The first alternates inserts by order week: every order packed in one week gets card A, the next week card B, and so on, so that seasonal swings fall evenly on both. The second holds out a random share of parcels that get no insert at all, which shows what the insert adds rather than which of two is better. Either way, read the repeat purchase rateShare of customers who placed more than one order in the period. of each group over the same window, matched to the product's natural cycle, from the store's own order data. Small groups give unstable results, so run the test long enough to collect enough customers in each group, and read differences that are small with suspicion.
A coffee seller rethinks its insert card (illustrative)
Tampwhirl Coffee swaps a busy card for a brew guide
Tampwhirl Coffee is an invented online coffee seller. Its insert card carries a 10% discount code, its social handle, a request for a review, a newsletter offer and a referral link. In a quarter, 2.1% of orders use the code and the team cannot say whether anything else on the card did anything.
It designs a second card with one job: a brew guide for the beans in the parcel, ending with a QR code that opens a tagged link to reorder the same beans. The review request moves to the post-purchase email for customers who can be emailed, and to a short line on the brew guide asking for an honest review after the first week of brewing, with no reward attached.
For twelve weeks, orders alternate by week between the old card and the new one: 3,100 first orders get the old card and 3,040 the new. Within eight weeks of their first order, 21.4% of the old-card group order again, against 24.9% of the brew-guide group. The QR code itself is scanned by about one customer in nine, and the team treats the difference in second orders, not the scans, as the result. It keeps the brew guide and tests a version for decaf beans next.
If cards and emails both bring in few reviews, the diagnosis page for customers not leaving product reviews measures reviews per delivered order by source and finds where the request breaks down.
Common mistakes
- Putting everything on one card. Five asks get no response; one ask, chosen for the moment, gets a better one.
- Spending on presentation before fixing damage. A beautiful box that arrives crushed is still a crushed box.
- Asking only happy customers for reviews. Selective requests misrepresent the picture as surely as deleting bad reviews would.
- Rewarding positive reviews. Any incentive is for an honest review, whatever it says, and is disclosed.
- Treating the order as consent to market. The card invites; the sign-up page asks and records the answer.
- Ignoring marketplace rules. Each marketplace sets its own rules on contacting its buyers; read them before designing the card.
- Measuring the card by its scans. Compare repeat purchase between groups that did and did not receive it.
- Untagged addresses. A card that sends people to the homepage with no tag can never be judged.
