Retention
The three retention levers: frequency, recency, breadth
Repeat revenue grows exactly three ways: customers buy more often (frequency), come back sooner (recency), or buy across more of your catalog (breadth). Pick the lever that fits your product.
Retention feels vague until you name its moving parts. There are only three: frequency (more orders per customer), recency (shorter gap between orders), and breadth (more categories per customer). Each fits a different product type, and pulling the wrong one wastes effort. AI can match the levers to your catalog and reorder data; you run the play.
The three retention levers: frequency, recency, breadth
Repeat revenue grows in exactly three ways, and naming them turns a vague goal into a plan. Customers buy more often (frequency), come back sooner (recency), or buy across more of your catalog (breadth).
AI can look at your product type and reorder pattern and tell you which lever actually fits.
Fastest path: one prompt, end to end
AI prompt — paste into ChatGPT / Claude
You are a retention strategist. Use MY numbers only (invent nothing). What I sell: [product type + how many distinct categories] Avg orders per customer per year: [#] Typical gap between orders: [# days] Share of customers who buy from >1 category: [%] Do this: 1. Tell me which of the 3 levers (frequency, recency, breadth) has the most headroom for MY product, and why. 2. Give ONE concrete play for that lever (frequency = bundles/subscriptions; recency = timed reminders; breadth = cross-category recommendation). 3. Say which lever does NOT fit me and why, so I don't waste effort. 4. Name the one metric that proves the lever is working. If a number is missing, ask; do not guess. Output: best lever + why + one play + lever to skip + metric.
Or do it in 4 steps
- Frequency: get each customer to buy more times. Best for consumables and low-cost repeatables. The play is bundles, multi-packs, or a subscription that makes buying again the default. Measure orders per customer per year.
- Recency: shrink the gap between orders. Best when people already reorder but slowly. The play is a reminder timed to the natural cycle, or a limited-time reason to come back now. Measure average days between orders.
- Breadth: get customers into more of your catalog. Best for multi-category stores where a buyer only knows one shelf. The play is a cross-category recommendation right after a first purchase. Measure share of customers buying from more than one category.
- Pick one lever, not all three. The wrong lever for your product is wasted work: pushing frequency on a once-a-year durable annoys people, while pushing breadth on a single-product store has nowhere to go. See repeat-purchase-flywheel for how any one lever compounds into LTV.
Worked example (labeled): A coffee brand sells one category, buyers reorder every ~40 days. Frequency and recency have headroom (a timed replenishment nudge at day 35, a subscribe-and-save); breadth does not, because there is only one shelf.
A home-goods store with 12 categories but 80% single-category buyers is the opposite: breadth is the big lever. Match to your own data.
Name your lever, run one play, and watch its single metric move before adding another.
Do
- ✓Pull frequency (more orders per customer) when you sell consumables or low-cost repeatables; the play is bundles, multi-packs or a subscription.
- ✓Pull recency (shorter gap between orders) when people already reorder but slowly; the play is a reminder timed to the natural cycle.
- ✓Pull breadth (more categories per customer) in multi-category stores where buyers only know one shelf; the play is a cross-category recommendation after the first purchase.
- ✓Pick one lever, run its single metric, and let it move before adding another.
Avoid
- ✕Don't push frequency on a once-a-year durable; nudging people to rebuy what they can't use yet just annoys them.
- ✕Don't push breadth in a single-product store; there's nowhere for the customer to go.
- ✕Don't pull all three levers at once; you won't know which one moved the number.
- ✕Don't leave retention as a vague goal; name the lever or you can't build a plan around it.
Quick tips
- Let AI look at your product type and reorder pattern and tell you which lever has the most headroom, plus which one to skip.
- Each lever has its own scoreboard: orders per customer per year (frequency), average days between orders (recency), share buying from more than one category (breadth).
- Once a lever is working, remember any single one compounds into LTV; see repeat-purchase-flywheel.
Brand in focus
Amazon — one lever per product, all three across the catalog
Amazon.comSubscribe & Save (frequency) and “customers also bought” (breadth) are two of its most-cited retention mechanics
Amazon fits because it visibly runs a different lever for each product type instead of one blanket tactic: Subscribe & Save pulls frequency on consumables, “time to reorder” reminders pull recency, and “customers also bought” pulls breadth across its huge catalog. What it does well is match the lever to the item rather than pushing all three everywhere. Watch-out: a small store can't do all three at once; the transferable lesson is picking the single lever that fits your product, which is exactly what Amazon does item by item.