The frame
Most consumable brands on Amazon treat Subscribe and Save as a setting they flip on once and forget. Eligible products get a 5% checkbox, the listing gets a small subscribe widget, and the brand assumes the program is doing its job. It is not. Subscribe and Save is the single largest LTV lever available on Amazon for consumables, and the gap between brands that work it and brands that ignore it shows up directly in customer acquisition cost economics inside twelve months.
This is the framework I use when a brand asks why their repeat rate looks flat despite shipping a consumable product. It applies to skincare, haircare, fragrance refills, supplements that sit under beauty or health, pet food, household consumables (cleaning, laundry, paper), grocery staples, and any SKU where the unit gets used up inside ninety days. It does not apply to colour cosmetics with long usage cycles, tools, accessories, or one-time-purchase categories. The economics break if the product is not actually consumed at the cadence the subscription assumes.
The framing matters. Subscribe and Save is not a discount programme. It is a retention mechanism that Amazon happens to subsidise on the discount side. If you think about it as a price reduction, you will optimise the wrong variables and your P&L will tell you the programme does not work. Think about it as a retention contract with the buyer, and the numbers start making sense.
Where this framework applies (and where it doesn't)
The mechanics are the same across categories - the economics differ.
- Beauty consumables (skincare, haircare, fragrance refills): the framework as written. 30 to 90 day usage cycle, mid-to-high contribution margin, listing quality matters most.
- Supplements: similar to beauty consumables. Tighter cadence (often 30 day), more variable retention because efficacy beliefs change.
- Pet food and treats: the cleanest fit. Predictable consumption, high subscription affinity, high price-sensitive baseline so the discount tier matters more.
- Household consumables (cleaning, laundry, paper goods, dishwasher tabs): high subscription affinity, lower contribution margin so the maximum discount tier needs careful sizing.
- Grocery / food staples: works for shelf-stable repeat-purchase items (coffee, tea, snacks, cereal). Fresh and frozen sit outside the standard programme.
- Where it doesn't work: colour cosmetics, fashion accessories, electronics, tools, anything with a 6+ month replacement cycle, anything where the buyer wants the next purchase to be a different SKU.
The five filters in the next section work the same across all eligible categories. The contribution margin floors and the breakeven math shift by category - that's covered in the P&L section below.
Eligibility is more selective than the policy document suggests
Amazon's documentation says any consumable in eligible categories can enrol in Subscribe and Save once the seller meets account-level criteria. That is technically correct and operationally misleading. The question is not whether your SKU can be enrolled. The question is whether enrolling it will produce a positive return given your specific cost structure.
The five filters I run on every SKU before enrolment:
First, the consumption cadence has to match a real subscription interval. Amazon offers one, two, three, four, five, and six month intervals. If your product gets consumed in five to seven weeks, the cadence is wrong and customers will either over-receive and cancel, or stretch the interval and treat the discount as a one-time coupon. Look at unit-per-customer-per-year data from Amazon Brand Analytics (specifically the Repeat Purchase Behavior report). If it does not cluster around a clean monthly multiple, the programme will leak.
Second, the contribution margin at the maximum discount tier has to be positive. The maximum tier is 15% off, triggered when the customer receives five or more subscribed items in one auto-delivery to one address (across any participating sellers, not just yours). Prime members enrolled in Amazon Family can stack an additional 5% on top of the seller-set tier. Most brands cannot absorb 15% off list and remain margin-positive after FBA fees, returns, advertising, and refunds. If you cannot, you need to design the tier structure with a hard ceiling at 10% or even 5%.
Third, the SKU has to have stable supply. Subscribe and Save cancellations driven by out-of-stock are punitive. The customer gets an email, often switches to a competitor, and the brand never sees them again. If your inventory planning is not airtight, do not enrol the SKU. Fix the supply chain first.
Fourth, the listing has to be at a conversion rate that survives the slight friction of the subscribe-versus-one-time choice on the detail page. Brands typically see a small conversion rate drop on Subscribe and Save eligible pages, because the choice itself slows the buyer. The size of the drop varies by category - more sensitive in price-led categories, less sensitive in trusted-brand categories. If your baseline CVR is already weak, the addition will hurt before it helps.
Fifth, the SKU has to be one that customers actually re-order. This sounds obvious. It is not. Brand Analytics will show you the repeat purchase rate for the SKU specifically, not the brand. A hero product with a 25% twelve-month repeat rate is a Subscribe and Save candidate. A best-selling but novelty SKU with a 6% repeat rate is not, no matter how high the volume.
Discount tier selection is where most brands set fire to margin
Amazon offers two discount tiers under Subscribe and Save. The base tier is 5%, applied to any single subscribed unit. The bundle tier (10% or 15%, set by the seller) triggers when the customer receives 5 or more subscribed items in one auto-delivery to one address - the items can be from any participating sellers, not just yours. Sellers opt in to the bundle tier and choose the discount level. Prime members enrolled in Amazon Family stack an additional 5%, so a customer hitting the maximum tier on an Amazon Family account effectively pays 20% less.
The default mistake is to enable both tiers at the maximum discount. The brand assumes more discount equals more subscriptions. That is half-true. What actually happens is the customer cohort that would have subscribed at 5% subscribes anyway and the brand absorbs an additional 10 points of margin loss on those buyers. The incremental subscriber acquired by going from 5% to 15% is usually a smaller share of total subscribers than the discount cost implies.
The framework I run is tiered conservatively. Single unit gets 5%, which is the floor and the right starting point. Bundle tier of 10% is the right next step if the brand has multiple consumable SKUs that naturally pair. Bundle tier of 15% is reserved for situations where the brand is actively trying to win share against a category leader who is at 10% and the contribution margin can absorb the additional five points.
How to check what competitors are at
Subscribe and Save discount tiers are visible on the public detail page. Open the top three competitor SKUs in your category and look for the subscription discount displayed near the price. Note the percentage. Build a one-line per-competitor record: SKU, list price, discount tier, estimated review volume. Repeat quarterly. If competitors shift from 5% to 10% in the same window, that is a category-level signal - either the leader is making a retention push or one of them found that the higher tier pays back.
Test the tiers separately
Run the 5% single-unit subscription for ninety days, look at the actual subscriber count, then layer in a 10% bundle tier and watch what happens to average order value. The bundle tier moves AOV materially when it works - that is where the programme starts paying for itself rather than just acquiring subscribers cheaply.
Listing copy and image treatment that earns the subscription
The subscription decision happens in two places on a detail page. The radio button toggle near the price, and the in-cart upsell. If the listing copy does nothing to support the toggle, the buyer defaults to one-time purchase and you lose the LTV opportunity in the first session.
Three listing treatments earn the subscription decision.
First, the bullet copy should reference usage cadence directly. "Lasts approximately 30 days per bottle for daily users" or "One unit covers a six-week skincare routine." Amazon does not surface this naturally and the buyer is doing the math anyway. Help them.
Second, the main image and the secondary scale shot should make the consumption rate visible. A skincare bottle with a pump that shows fill level, a haircare unit with the recommended dosage shown in the lifestyle shot. The framework for image sequence on consumable SKUs lives in the 7-image grid post.
Third, the A+ Content module should include a "How often to reorder" panel. Not a discount call-out. Not a "subscribe and save" call-out. A factual panel that anchors the customer expectation around when they will run out. The subscription becomes the obvious solution to a problem the customer now understands.
Avoid the temptation to scream "Subscribe and Save 15%" in the main image. It looks promotional, Amazon's image guidelines may flag it, and it shifts the buyer mindset from purchase to discount-hunting. The discount is the close, not the lead.
Churn signals to watch from week one
Subscribe and Save subscribers churn for predictable reasons, and most of those reasons are upstream of the programme itself. Watching the right signals in the first ninety days of enrolment determines whether the LTV uplift compounds or stalls.
The four signals to track weekly. The specific thresholds below are starting benchmarks - they vary by category, price tier and brand maturity. Use them as a "look closer" trigger, not a hard rule.
Skip rate inside thirty days of first delivery. If skip rate climbs above roughly 15%, the cadence is wrong. Either the interval is too tight or the customer is not actually consuming the product at the assumed rate. Adjust the default interval before the cohort churns out completely.
Cancellation rate inside ninety days. Healthy programmes in mature subscription categories typically sit in the 10 to 20% range for 90-day cancellation. Materially above that means the product is not retaining, and no amount of discount tuning will fix it. The fix is upstream in the product or the targeting of who is subscribing.
Quantity reduction without cancellation. This is the early warning signal most brands miss. A customer drops from three units per delivery to one unit per delivery. They have not cancelled, but the LTV is degrading. Brand Analytics surfaces this and most teams ignore it.
Refund rate on subscribed deliveries versus one-time orders. If subscribed deliveries are getting refunded at a higher rate, something in the fulfilment or packaging experience is breaking on the second and third delivery. Customers expect the first order to be perfect. They expect the third to be just as perfect, and any drop in experience is read as a reason to cancel.
P&L math you should run before enabling, not after
The economic question on Subscribe and Save is not whether subscribed deliveries themselves earn margin. They do, until you push the discount tier too high. The real question is whether S&S genuinely lifts customer LTV - or whether it just discounts purchases that would have happened anyway.
Most brands enable S&S and then look at the wrong number. They check whether subscribed deliveries are CM1-positive (they almost always are) and conclude the programme is working. It might not be. The discount cost is the same whether S&S converted a new repeat customer or just gave a permanent discount to a customer who was already loyal.
The two cases that matter
Same $30 SKU. Standard CM1 (after COGS, FBA fees, returns) is $12, or 40% margin.
Case 1: S&S converted a one-time buyer into a 3-delivery subscriber. This is the win case. A customer who would have bought once and disappeared is now receiving multiple deliveries because the subscription mechanic removed the re-purchase friction.
- Discount cost per unit: $30 × 5% = $1.50
- New CM1 per delivery: $12 - $1.50 = $10.50
- Without S&S enabled, LTV: 1 × $12 = $12
- With S&S at 5%, LTV: 3 × $10.50 = $31.50
- LTV uplift per converted customer: $31.50 - $12 = +$19.50. Big win.
Case 2: S&S cannibalized a 3-purchase loyal buyer. This is the loss case. A customer who would have bought 3 times anyway is now getting a permanent discount on every purchase.
- Discount cost per unit: $30 × 5% = $1.50
- New CM1 per delivery: $12 - $1.50 = $10.50
- Without S&S enabled, LTV: 3 × $12 = $36
- With S&S at 5%, LTV: 3 × $10.50 = $31.50
- LTV loss per cannibalized customer: $31.50 - $36 = -$4.50. Net negative.
The same SKU, the same discount, opposite outcomes
Notice what changes between the two cases: nothing about your product, your price, or your discount cost. The discount per delivery is identical. The cumulative CM1 from the subscribed deliveries is identical.
The only thing that differs is the counterfactual - what would that specific customer have done if S&S were not enabled. And the LTV swing is significant: +$19.50 vs -$4.50 on the same SKU.
For Subscribe and Save to pay out at the brand level, the LTV gained from converted one-time buyers has to exceed the LTV lost from cannibalized loyal buyers. That is the real breakeven.
How to know which side you are on
The metric to look at is incremental subscriber rate - of the customers who subscribed, how many came from a cohort that historically did not repeat, versus a cohort that already repeated naturally?
Most brands cannot segment that cleanly. The practical proxy: compare your SKU's organic 90-day repeat-purchase rate (Amazon Brand Analytics surfaces this) against the category baseline for the same product type. Absolute thresholds are misleading because category baselines vary enormously.
Category baselines to anchor against (publicly available data):
- Pet food: subscription rates of 58% for dog food, 45.9% for cat food. Total category repeat rate sits in the 35-50% range.
- Grocery staples: ~55% of category shoppers have at least one S&S subscription. Repeat rate 40-55%.
- Beauty / personal care: ~42% S&S adoption in the category. Repeat purchase rate ~25-30%.
- Supplements: repeat purchase rate ~29% on average for Amazon nutrition brands.
- Pet supplies (broader, beyond food): ~35% S&S adoption. Repeat rate 30-45%.
- Consumables overall: 30-45% repeat purchase rate due to natural replenishment cycles.
The actionable comparison:
- If your SKU's repeat rate is below your category baseline: you have conversion headroom. S&S is more likely to convert one-time buyers (net positive).
- If your SKU's repeat rate is at or above your category baseline: most subscribers will be customers who would have repeated anyway. Cannibalization risk is higher. Be conservative on the 10%/15% bundle tier.
For context: profitable Amazon brands typically run 30%+ overall repeat purchase rates. Below 25% is a warning signal that the product or listing has a deeper problem before S&S can help.
Per-delivery margin math (the simpler check)
Once you have made the brand-level call to enable, the per-delivery math is the second-level check. At what tier do subscribed deliveries stop being CM1-positive?
On the $30 SKU with $12 baseline CM1:
- 5% (single unit): discount cost $30 × 5% = $1.50 · new CM1 per delivery $12 - $1.50 = $10.50 (35% margin)
- 10% (bundle): discount cost $30 × 10% = $3.00 · new CM1 per delivery $12 - $3.00 = $9.00 (30% margin)
- 15% (bundle): discount cost $30 × 15% = $4.50 · new CM1 per delivery $12 - $4.50 = $7.50 (25% margin)
All three tiers are CM1-positive on this SKU. The question is which tier optimises LTV after accounting for the conversion-vs-cannibalization split.
Why higher tiers amplify the risk
The cannibalization cost on a loyal buyer scales linearly with the discount.
- At 5%, cannibalized loyalist (3 purchases): 3 × $1.50 = $4.50 LTV lost per cannibalized customer
- At 10%, cannibalized loyalist (3 purchases): 3 × $3.00 = $9.00 LTV lost per cannibalized customer
- At 15%, cannibalized loyalist (3 purchases): 3 × $4.50 = $13.50 LTV lost per cannibalized customer
The conversion win stays roughly the same. The cannibalization loss triples. Which is why the standard advice "test the 10%/15% bundle tier separately, not bundled with the 5% single-unit tier" matters. You are not just measuring whether more people subscribe. You are measuring whether the additional subscribers are conversions or cannibalisations.
Category margin profiles
The exact margin profile shifts by category, which is why the same playbook produces different recommended ceilings:
- High-margin consumables (beauty serums, premium supplements): 40-55% CM1. Even 15% S&S leaves comfortable per-delivery margin. Cannibalization risk is the binding constraint, not per-unit profitability.
- Mid-margin consumables (mass beauty, household goods): 25-40% CM1. The 15% bundle tier starts to cut close to breakeven on per-delivery margin. Most brands cap at 10%.
- Low-margin commodities (pet food, paper goods): 15-25% CM1. The 15% tier often turns subscribed deliveries CM1-negative. These categories usually cap at 5% single-unit, no bundle.
How this interacts with TACoS/CM1 ceiling math
The discount comes out of CM1. If you are running the TACoS spend ceiling framework, your CM1 input needs to reflect the subscription-weighted discount, not the gross-of-discount number. For a SKU bucket where 30% of orders are subscribed at 5%, the effective CM1 is roughly 1.5 points lower than the headline CM1. Update your spend ceiling math accordingly or your weekly TACoS check will be miscalibrated.
Geographic and seasonal patterns
Geographic shape
US is the mature market - high enrolment rates, competitive discount tiers, faster churn cycles, steeper listing-quality requirement. Brands launching in the US Amazon ecosystem assume the programme is contested.
UAE / Gulf markets, India, and emerging Amazon marketplaces show different dynamics. Subscribe and Save adoption is lower, which means lower competitive pressure on tiers but also slower brand-level enrolment growth. Brands operating across these markets need different shape on the programme - lower discount tiers usually pay back faster because the marginal subscriber is acquired more easily.
UK and EU sit between US maturity and emerging-market patterns, with category-by-category variation.
Prime Subscribe and Save Day
Amazon runs a category-specific seasonal event around Subscribe and Save adoption (timing varies year to year, typically once a year). During the event, eligible SKUs get incremental visibility and the cost of acquiring a new subscriber drops materially. Verify timing and category eligibility with your Amazon AM each year. Plan a coordinated push - listing copy adjustments, inventory pre-position, ad spend lift on hero SKUs - the four weeks before the event.
Vendor Central vs Seller Central
The Subscribe and Save mechanics work in both 1P (Vendor Central) and 3P (Seller Central) environments, but with different control levels:
- Seller Central: the brand controls enrolment, tier selection, and listing copy directly. Faster to iterate.
- Vendor Central: Amazon's category teams have more influence on tier defaults and program participation. Slower to iterate but sometimes better positioning during seasonal events.
If you operate hybrid (some SKUs 1P, some 3P), reconcile the discount tier strategy across the two channels - mixed tiers across the same brand confuse customers and can leak the 10%/15% tier choice if the bundle calculation goes wrong.
When NOT to enable Subscribe and Save
The five filters at the top catch most of the obvious "not now" cases. A few additional cases worth flagging:
- Brand in pre-PMF phase. If you do not yet know which SKU will be the hero, enrolling everything is process for the sake of process. Wait until you have stable repeat winners.
- Cash-constrained brands. S&S delivery on auto-renewal commits the brand to ship at a specific cadence regardless of inventory cost trajectory. If your inventory financing is fragile, S&S amplifies the stockout risk.
- Brands with active suspension or compliance review. Programme participation can pause unexpectedly during reviews. Resolve the compliance issue first.
- Categories with regulated active ingredients (cosmetics with prescription-level claims, supplements with regulated ingredients). Verify the auto-renewal mechanic complies with your category's local regulations - subscription auto-renewal is treated differently across markets.
- SKUs in active price testing. S&S locks in the discount mechanic on top of whatever pricing strategy you are running. If you are still discovering the right list price, do not layer the discount complexity on top.
Bottom line
Subscribe and Save is not a promotional setting. It is a retention contract with the buyer, subsidised by Amazon's discount mechanic, and it only pays out for brands that have the consumption fundamentals, the contribution margin, and the listing quality to support the contract. If those three are in place, the LTV uplift is one of the largest single levers available to a consumable brand on Amazon. If they are not, enrolling the SKU before fixing the fundamentals just compounds the leak.
Run the five filters before enrolling. Run the P&L breakeven math at each tier you are considering. Watch the four churn signals from week one. Adjust the discount tier conservatively. Update your TACoS spend ceiling for the subscription-weighted CM1. Plan around Prime Subscribe and Save Day. Re-verify the multi-unit trigger rule with your AM at least once a year because the specifics shift.
If you want me to walk through your specific SKU mix and tell you which ones belong in the programme and which ones should not, message me on WhatsApp. If you want a written diagnostic on the broader account, the Amazon Account Audit covers Subscribe and Save eligibility and tier sizing as part of the scope.