Self-insuring means skipping per-shipment insurance and covering losses out of your own pocket — and for high-volume shops shipping low-value parcels, the math usually favors it: when your real loss rate times your average claim value is smaller than what premiums would cost across the same shipments, insurance is a losing bet you're allowed to decline. The exceptions are sharp, though: high-value orders, fragile categories, and shipments where one loss would genuinely hurt should carry insurance even in a shop that self-insures everything else. Here's how to run the numbers instead of guessing.

What does self-insuring actually mean for a shop?

Insurance is a trade: you pay a small certain cost (the premium) to escape a large uncertain one (the loss). An insurer can offer that trade profitably because it pools thousands of shippers — and prices premiums above expected losses, plus overhead, plus margin. That last part is the pivotal fact: on average, across many shipments, insurance costs more than the losses it covers. It must, or the insurer wouldn't exist.

A shop that ships in volume is, in a small way, its own pool. Two hundred parcels a month is enough repetition for losses to behave statistically rather than catastrophically: you won't lose zero parcels or fifty, you'll lose roughly your loss rate's worth, month after month. At that point you can keep the insurer's margin for yourself — decline the premium, absorb the occasional loss, and come out ahead over time.

Self-insuring is not the same as ignoring risk. It's a deliberate policy with three parts: knowing your actual loss rate, insuring the exceptions where pooling breaks down, and setting aside the money you're saving so a bad month is a bookkeeping event rather than a crisis. Skip any of the three and you're not self-insuring — you're just uninsured.

One scope note: whichever way you decide, GoatLabels makes the choice per shipment — an insurance toggle at quote time, with the fee folded into the all-in price you compare across carriers. GoatLabels isn't the insurer and doesn't decide claims; coverage and claim adjudication sit with the carrier or insurance provider behind the toggle, and the claims process itself is covered in lost or damaged package claims.

How do you compute your break-even loss rate?

The decision reduces to one comparison per category of shipment:

Expected loss per shipment = loss rate × average cost of a loss. Insure when the premium is below that number; self-insure when it's above.

To use it, you need two inputs from your own records — not industry folklore:

Your loss rate. Out of your last few hundred shipments, how many ended in a paid-out loss — lost in transit, damaged on arrival, or a delivered-but-not-received case you refunded? Count real money events, not tracking scares that resolved.

Your average cost of a loss. What a loss actually costs you: replacement product at your cost (not retail), a second label, and the time to handle it. If you can replace a $60-retail item for $22 of materials and a $6 label, your loss cost is closer to $30 than $60.

Worked example — illustrative numbers, not quotes: suppose insuring a $75 order costs about $1.50 in premium. Break-even loss rate = $1.50 ÷ $30 true loss cost = 5%. If your measured loss rate is 0.5%, insurance at that price costs ten times your expected loss — self-insure. Flip the inputs — a $600 collectible you can't replace at a discount, same $1.50-per-$75-of-value premium structure — and one loss wipes out the savings from hundreds of uninsured shipments. Insure it.

The pattern generalizes: premiums scale with declared value, but your loss rate doesn't. That's why the answer differs by order value band rather than being one policy for the whole shop.

Which orders should always carry insurance anyway?

Self-insurance works because losses are small relative to the pool absorbing them. It breaks exactly where that stops being true:

Outlier-value orders. Any single shipment whose loss would erase weeks of self-insurance savings — the $500 order in a shop whose average is $40. The pool logic doesn't cover the order that is the pool.
Irreplaceable or full-cost items. One-of-one art, sold-out editions, consignment goods — anything where "replacement at your cost" isn't available and a loss means refunding full price.
Categories with elevated damage rates. If your records show ceramics or framed pieces claiming at several times your base rate, that category's expected loss justifies premiums even when the shop-wide math says skip.
Shipments where the dispute profile is bad. High-value parcels to porch-piracy-prone addresses, or buyers whose order pattern smells off. For delivered-but-stolen scenarios specifically, note that insurance and signature requirements solve different problems — a "delivered" scan usually ends a loss claim, which is when a signature requirement is the tool that would have helped. See delivered but not received for how those cases actually resolve.
When one loss threatens cash flow. Early-stage shops without a cushion should insure more aggressively — the mathematically optimal policy is worthless if a single bad week breaks you before the law of large numbers arrives.

Because the right answer changes order by order, per-shipment toggling matters more than any blanket policy. A blanket "always insure" wastes money on the $20 orders; a blanket "never insure" bets the shop on the $600 one. On GoatLabels the toggle is on every quote, so your rule — "insure above my threshold, skip below" — takes one glance per order to apply. Signature requirements work the same per-shipment way, with the fee folded into the quote.

How does default declared-value coverage change the math?

Many carrier services include a base level of declared-value coverage automatically — a modest amount of protection bundled into the label price on certain services, with the specifics varying by carrier, service, and channel. Check the current terms for the services you actually use, because the details shift; the strategic effect is what matters here:

Included coverage moves your insurance threshold up. If a service already covers the first slice of value at no extra cost, then orders whose true loss cost falls inside that slice are effectively insured for free — buying additional cover there is pure waste. Your decision rule becomes: self-insure up to the included coverage, buy the top-up only where order value meaningfully exceeds it and the break-even math says so.

This also quietly favors comparing carriers on more than price: two similar quotes can carry different included coverage, which changes the effective cost of protecting a mid-value order across USPS, UPS, FedEx, and DHL.

One caution: included coverage still requires a successful claim — documentation, packaging standards, deadlines. It's real protection, not automatic reimbursement. Keep photos of high-value packouts and file promptly.

How do you build a self-insurance reserve that actually works?

The reserve is what separates self-insurance from wishful thinking. The mechanism is simple: charge yourself the premium you declined, and bank it.

Set your internal premium. A flat amount per shipment or a small percentage of order value — roughly what insurance would have cost is a fine starting point.
Move it somewhere visible. A separate savings account, or at minimum a tracked line in your books. The reserve must be countable, or it will silently get spent.
Pay losses only from the reserve. Every lost or damaged order draws from it, never from operating cash. This keeps your true loss economics honest and your bad weeks boring.
Review quarterly. Reserve growing steadily? Your loss rate is below your internal premium — you're winning the bet, and you can lower the internal premium or keep building the buffer. Reserve shrinking? Your loss rate is higher than assumed — raise the internal premium, fix the packaging or carrier mix causing losses, or move more orders back to real insurance.

Here's the shape of the bet over 1,000 shipments — illustrative arithmetic to show the structure; substitute your own premiums and loss costs:

Order value band	Loss-rate scenario	Insure everything	Self-insure with reserve
Low (avg $25)	0.5% (5 losses)	~$750 in premiums, losses covered	~$60 in losses paid — self-insure wins big
Mid (avg $75)	0.5% (5 losses)	~$1,500 in premiums	~$150 in losses — still wins clearly
Mid (avg $75)	2% (20 losses, fragile goods)	~$1,500 in premiums	~$600 in losses — wins, but the gap narrows; fix the damage rate
High (avg $400)	0.5% (5 losses)	Premiums buy real protection	One bad cluster can exceed years of savings — insure these

The low- and mid-value rows are where most small shops live, and they're where self-insurance earns its keep. The bottom row is why the policy is per-shipment, not per-shop.

A final point on the money itself: self-insuring pairs naturally with pay-as-you-go shipping costs. There's no monthly software minimum on GoatLabels — you fund a wallet, pay per label, and the insurance line appears only on the shipments where you chose it — so a slow month costs you nothing in either premiums or subscriptions (see pricing).

FAQ

Is self-insuring risky for a very small shop? It's riskier the fewer parcels you ship, because losses arrive in clusters rather than smooth averages. Under a few dozen shipments a month, lean toward insuring mid- and high-value orders while your loss data accumulates; the statistics that make self-insurance work need volume to show up.

Does GoatLabels provide the insurance when I toggle it on? No — GoatLabels surfaces the option and folds the fee into your all-in quote, but the coverage sits with the carrier or insurance provider, and claims are decided by them, not by GoatLabels.

What loss rate should I assume if I have no data yet? Don't assume — insure conservatively for your first few hundred tracked shipments and count every real loss. Well-packed parcels on tracked services typically lose at a fraction of a percent, but your products, packaging, and destinations set your actual number.

Does insurance cover a package marked delivered but never received? Generally a delivery scan ends a carrier loss claim — insurance protects against loss and damage in transit, not porch theft. For high-risk deliveries, a signature requirement (also a per-shipment toggle) is the tool aimed at that failure mode.

Should the reserve be a real bank account? Ideally yes — friction is a feature. A reserve that exists only in your head gets spent. A separate account that only fulfills claims makes the self-insurance bet auditable, and its balance tells you at a glance whether the policy is working.