Kickstarter Shipping Advice for Creators: Updates, Delays and Refunds

Why shipping communication is a product decision

Most creators treat campaign updates as an obligation and shipping as a logistics task. They are the same activity seen from two directions. A campaign that ships on time and communicates badly produces more complaints than a campaign that ships four weeks late and communicates well, because the second one has managed expectation while the first has merely met it and left backers guessing.

The reason is psychological rather than logistical. A backer who pledged in March and hears nothing until September is not measuring the delay against a schedule. They are measuring it against silence, and silence is read as concealment. The same eight week slip, narrated in four dated updates with reasons, reads as a project being managed.

There is a measurable commercial consequence too. Campaigns that maintain an update cadence through fulfilment see materially higher repeat purchase rates on the post-campaign retail launch, which is where most of the lifetime value of a bag campaign actually sits. The backer who watched the process is the customer who buys the second bag at full price.

A useful way to think about it is as information inventory. Backers are waiting on a product, and while they wait, information is the only thing you can actually deliver. Running out of information is the same failure as running out of stock: the customer experiences it as unavailability, and in the crowdfunding context unavailability of news is interpreted as bad news. Keeping a small buffer of prepared updates is the equivalent of holding safety stock.

So plan communication like production. Give it an owner, a cadence, a set of pre-drafted messages and a place in the same calendar as sampling and shipping. Our programme coordinators treat the update schedule as a deliverable of the production programme rather than as an afterthought, and that framing is what makes it happen under pressure.

Setting expectations on the campaign page

Every communication problem during fulfilment traces back to an expectation set on the campaign page. Four statements do almost all the work, and all four belong above the fold of the shipping section: which countries you ship to, whether shipping is included, whether duty and tax are included, and the estimated delivery month.

State delivery as a month, never a date, and never a number of weeks from close. A month survives a two week slip without being technically wrong, whereas "ships in August" becomes a broken promise on 1 September. If you want to be more precise, give a quarter, which is what most experienced creators now do for first campaigns.

Name the countries you exclude, explicitly. Backers in excluded countries who discover the exclusion after pledging are the most vocal complaint source in crowdfunding, and the cost of preventing it is one sentence before launch. If you are unsure whether you can ship somewhere, exclude it on the page and add it later rather than the reverse.

Add one sentence committing to a communication cadence, such as an update every two weeks during fulfilment. This costs nothing at launch and it is the single most effective sentence in the page, because it converts a future silence into a visible broken promise that you will then feel obliged to keep.

Add a line on what happens if you cannot deliver at all. It feels pessimistic to write before launch and it is the sentence that protects you most if the worst occurs, because backers who were told in advance what the fallback is respond to a failure very differently from backers who were not. One sentence naming the fallback, usually an alternative tier or a refund, is enough.

The update cadence that survives a delay

The cadence that works is boring and fixed: one update every 14 days during production and fulfilment, regardless of whether anything has happened. The word regardless is the important one. A fortnightly update with nothing to report is not filler; it is the mechanism that makes the update with something to report credible.

Each update should carry four elements in the same order: what happened since the last one, what happens next, the current delivery estimate, and the date of the next update. Four elements, fixed order, roughly 200 to 350 words. Backers stop reading long updates, and they search short ones for the delivery date, so put it in the same place every time.

Increase frequency rather than length when something goes wrong. Two short updates in a week during a crisis outperform one long explanatory post, because the first update buys goodwill and the second proves the first was not an isolated apology. The campaigns that recover well from delays are almost always the ones that over-communicated in the bad week.

Two mechanical habits make the cadence survive a busy month. First, write the next update immediately after publishing the current one, while the facts are fresh, and leave it as a draft with the date in the title. Second, keep a standing list of three things you can always report honestly, such as the current phase, the next milestone and one photograph, so an update is never blocked by a lack of news.

Note the incident row. The commitment is a response within 48 hours of you knowing, not within 48 hours of backers noticing. The difference is usually three or four days, and it is the difference between an update that leads the story and an update that answers an angry comment thread.

Writing the delay update: structure and wording

A delay update has five moves, in this order. State the new date. State the reason in one sentence. State what you have already done about it. State what backers need to do, which is usually nothing. State when the next update comes. Anything else, including apology, explanation of context and reassurance, belongs after those five and never before them.

The reason sentence is where most updates fail. "Production delays" is not a reason, it is a category. "The lining fabric failed the colour fastness check and the replacement lot arrives on the 14th, which moves completion by 12 days" is a reason. Specificity is what makes the delay credible, and vagueness is what makes backers assume the worst.

Avoid three constructions. Do not say "we are working around the clock", which reads as panic. Do not say "these things happen", which reads as indifference. Do not say "we appreciate your patience", which reads as a template. Say what you did, what happens next, and when you will write again.

Decide in advance who is allowed to publish. On a two person team, the person who is not negotiating with the production line should write the update, because the negotiator is usually too close to the problem to describe it plainly. Agree that no update goes out with an unverified date in it, even when the pressure to say something is high, because an unverified date is the one error that cannot be walked back.

One more element earns disproportionate goodwill: name the cost you absorbed. If a re-make cost 4,000 USD and you paid it rather than shipping a compromised product, say so in one sentence. It is not boasting, it is evidence, and it converts an abstract delay into a visible decision that backers can respect.

What to do in the first 48 hours of a known delay

The first 48 hours after you learn about a delay determine whether it becomes a story or a footnote. Work in this sequence. First, establish the new date with the people who control it, and do not publish until you have it, because a revised estimate that slips again does more damage than a late but accurate first estimate.

Second, establish the scope. How many backers are affected, which tiers, which countries. A delay affecting 200 backers in one lane is a different communication from one affecting all 1,200, and the update must say which it is. Publishing a global message for a partial delay unnecessarily alarms the unaffected majority.

Third, write the update using the five moves and publish it. Fourth, prepare the answers to the three questions you will definitely be asked: can I change my address, can I get a refund, and will this affect the quality. Answer all three inside the update rather than waiting for them in comments.

Fifth, and this is the step most creators skip, tell your support channel and any retail or press contacts on the same day. Backers who hear about a delay from a journalist or from a retail partner before they hear it from you have a legitimate grievance, and it is entirely preventable with one hour of emails.

Prepare the internal version of the update as well as the public one. The internal note names the supplier, the carrier or the process that failed and what is being done about it; the public note names the effect and the new date. Keeping the two separate stops the public update from turning into a blame statement, which backers read as deflection rather than as information.

Tracking: what to send, when, and what to avoid

Send tracking at dispatch, not at order. A tracking number issued when the label is created but before the parcel moves produces a fortnight of "no information available" screens and a wave of support tickets. Wait until the carrier or hub confirms the first scan, then send the number with the carrier name and a note on expected transit.

Tell backers which carrier will deliver the final leg. On consolidated lanes the tracking number often changes at the hub, and a backer who receives a domestic tracking number they did not expect will assume something is wrong. One sentence in the dispatch update explaining that tracking is re-issued on arrival removes that entire category of enquiry.

Avoid sending tracking in bulk without a check. Upload the manifest, wait 24 hours, then verify a sample of 20 numbers yourself before the mass send. A single formatting error in a tracking upload can invalidate hundreds of numbers, and discovering that after backers have tried them is far worse than a one day delay in sending.

Handle the lost parcel case with a defined threshold rather than case by case. Agree a number of days past the stated transit window at which you treat a parcel as lost, commonly 21 days, and publish it. Below the threshold the answer is that the parcel is in transit; above it, you reship immediately and chase the carrier separately. A published threshold removes the argument and stops the support queue being run on emotion.

Set the expectation on transit time explicitly, and use the worst case in the range. If the lane runs 9 to 16 days, say up to three weeks. A parcel that arrives inside the stated worst case generates no complaint; a parcel that arrives on day 12 against a stated 10 generates one, even though nothing was actually late.

Handling refund requests during fulfilment

Refund requests arrive in three waves: during the campaign, during production, and during dispatch. The correct response differs by wave. During the campaign, refund through the platform and move on; it is cheap and it is not your money yet in any meaningful sense.

During production, the honest answer is that funds are committed. Production is paid under T/T 30/70 terms, so a deposit has gone out and materials are bought. Decline the refund politely, explain that the money is in the product, and offer a transfer instead: many backers accept having their reward reassigned to a friend or to a waiting list, which costs nothing.

During dispatch, the calculation changes. By then you have a landed cost per parcel and a replacement pool. A refund returns the full pledge out of funds already reduced by roughly 8 to 12 percent in platform and processing fees, while a replacement costs 18 to 34 USD. Replace where you can and refund only what you genuinely cannot fulfil.

Publish the refund position once, in the campaign FAQ, and never negotiate it publicly in the comment section. A public negotiation about one backer's refund sets an expectation for the next 40. Answer the specific case privately and keep the public statement consistent with the returns and replacements policy you published before launch.

There is a fourth wave that arrives after dispatch and behaves differently: buyer's remorse dressed as a complaint. These backers have received the product and describe a minor issue as a major one in the hope of a refund. The test is whether they will accept a replacement. Offer it once; if the offer is refused, refund and move on rather than arguing, because the support time costs more than the refund.

The comment section: triage without feeding it

The comment section is not a support channel and treating it as one makes it worse. The rule that works is acknowledge publicly, resolve privately. Reply to a complaint in one sentence saying you have messaged them directly, then actually message them. Other backers see responsiveness without being given a template to copy.

Triage into three buckets. Answerable now: address changes, tracking questions, sizing. Answerable later: anything requiring a supplier or carrier response, where you post the date you expect the answer. Not answerable: speculation, requests you will not grant, and repeated posts from the same person, which get one reply and no further engagement.

Never delete a critical comment unless it is abusive or contains personal information. Deleting criticism is the single fastest way to convert a question into a campaign, and the screenshot will outlive the campaign page. Leave it, answer it once, and let the next dated update do the work.

One pattern deserves naming because it destroys more campaigns than any single delay: the escalation spiral, where a creator answers the same complaint four times in public, each time with slightly more detail, and the fourth reply contradicts the first. Cap public engagement on any one thread at two replies, then move to private message. Consistency matters more than completeness in a forum everyone can read.

Pin one comment with the current delivery estimate and update it whenever the estimate changes. Most backers read the pinned comment rather than scrolling 60 updates, and a single maintained source of truth removes the need to answer the same question in five places.

Communicating partial shipment and wave dispatch

Wave dispatch needs explaining before it happens, because a backer who sees somebody else receive a parcel while theirs has not shipped will assume they were skipped. One paragraph in the pre-dispatch update, stating the wave order, the wave size and the reason for the order, prevents most of that.

Give a reason for the order that is not arbitrary. Ordering waves by destination, by tier, or by pledge date are all defensible and all explainable. Ordering them by warehouse convenience is not, and backers will detect it. The most common and least contentious rule is destination lane order, which is also usually the cheapest.

Publish the wave boundary dates and hold them. If wave one ships on the 12th and wave two on the 26th, saying so up front means the backers in wave two are waiting for a known date rather than for an unknown event. Known waits are tolerable; unknown waits generate comments.

When a wave ships, post the same four elements as any other update, plus the count and the destination group. Backers in later waves use those posts to estimate their own arrival, so include the transit range again rather than assuming they remember it from three updates ago.

One mechanical detail prevents a surprising amount of friction: do not post photographs of one wave's parcels before that wave has actually shipped. A backer in a later wave who sees a picture of 300 parcels leaving will assume theirs was in the pile. Post the photograph after the wave is confirmed dispatched, and caption it with the wave number and destination group.

When a reward genuinely cannot be delivered

Sometimes a tier cannot be produced, a country becomes unshippable, or a component fails compliance. This is the hardest communication in crowdfunding and it has a reliable structure. State what cannot be delivered and why, state the alternative, state what it costs the backer, and state how to choose something else.

Offer at least one alternative that costs you more than the original, because it signals good faith and it is usually cheaper than the reputational cost of a bare refund. An upgrade to a higher tier at no charge, or the original reward plus an accessory, both convert a failure into a story backers repeat positively.

Set a decision deadline for backers, typically 21 days, and state what happens to non-responders. Silence is not a decision, and an unstated default produces a second wave of complaints three months later. Name the default, usually the closest equivalent tier, and apply it on the stated date.

Keep a dated record of every communication in this category. If a dispute escalates to the platform or to a payment provider, the campaign that can show a dated sequence of offers, deadlines and defaults resolves it in days. The same record-keeping habit applies across the rewards fulfilment plan, where dispatch, returns and communication share one timeline. The campaign with an undocumented goodwill effort resolves it in weeks, or loses.

Tell the affected group directly rather than in a general update, if the platform allows it. A message to 200 affected backers is more work than one post to 1,200 and it produces a dramatically better response, because the unaffected majority is not alarmed and the affected minority knows they have been identified rather than left to work it out from a general announcement.

Closing the campaign: the final update and what comes next

The final update is read more than any other and it is where most campaigns undersell themselves. It should do four things: confirm completion with numbers, thank backers without sentimentality, state what happens next for support and warranty, and invite the next step, which is usually the retail launch or the waitlist.

Include the real numbers. Parcels shipped, countries reached, replacement rate, and the date the warranty window closes. Specificity here does more for trust than any amount of warmth, because it demonstrates that the campaign was tracked rather than improvised. Campaigns we support typically close with a replacement rate between 2 and 4 percent, and stating it reassures rather than alarms.

State the support and warranty path plainly: which inbox, what response time, what the warranty covers and for how long. A final update that leaves backers with a monitored address converts a completed transaction into the start of a customer relationship, which is the actual asset the campaign created.

Before you publish that final update, reconcile it against the fulfilment checklist and note which items went unticked. Doing this in public is optional, but doing it privately is what converts one campaign's experience into a tighter plan for the next one, and it is the cheapest form of research available to a creator.

Then stop updating. A campaign that continues posting monthly with nothing to say trains its audience to ignore the channel, and if the campaign runs again, that trained inattention is expensive. Close cleanly, keep the warranty commitment, and save the update channel for something worth saying.

Frequently Asked Questions

People Also Ask