Most retention strategies start with tactics — abandoned cart flows, loyalty programs, win-back emails. That's the wrong order. The audit comes first, because half the time the leak isn't where you think.
Search "customer retention strategy" and you'll get fifty articles that all start the same way: a list of tactics. Add an abandoned cart flow. Launch a loyalty program. Send a win-back. They're not wrong tactics — but they're the second thing to do, not the first.
The first thing is the audit. Half the time the leak isn't where teams assume. The Sunday email is fine, the cart flow is fine, the LTV is actually growing — but you're losing 40% of post-purchase customers in week three because the second-purchase nudge fires when stock is out. Tactics built without that audit ship work that doesn't move anything.
This piece is the audit you should run before you write a single line of new copy.
A customer retention strategy is the operating plan for keeping the customers you already acquired — the sequence of measurement, lifecycle stages, channels, and incentives that turn one-time buyers into repeat buyers. It's not the loyalty program. The loyalty program is one tactic inside it.
The strategy answers three things, in this order:
Most teams skip steps 1 and 2 and start at step 3 with whatever tactic the agency is best at selling.
If you run a sub-£10M / AUD $15M ecommerce business, the retention leak is almost always at one of these five points. Audit them in order before adding tactics.
Median repeat-purchase rates across 200+ DTC stores in the £1M–£10M revenue band, 2025. Your numbers may be ±15% but the shape repeats.
Welcome series ends at email 3. Customer hears nothing for the next three weeks. The most engaged moment in the whole lifecycle wasted. Audit: check whether the welcome series links to genuinely useful content (how to use the product, what to expect next) or whether it dumps "20% off your next order" three times in a row.
The single biggest retention number is whether someone makes a second purchase. If first-to-second conversion is under 25%, nothing else you do matters — you're filling a bucket with a hole in it. Audit: when does your category's repeat-purchase happen naturally? Skincare = 35 days, supplements = 28 days, apparel = 90 days. Is your trigger firing at the right window?
Email frequency too high, list goes quiet, opens drop, deliverability follows. Or frequency too low and the customer forgets you exist. Audit: open rate by send day-of-week, complaint rate, list churn vs growth.
Sometimes the product genuinely isn't right for them. No retention strategy fixes that. Audit: review survey + post-purchase NPS by acquisition channel — the leak might be specific to one channel that over-promises.
Most win-back flows are a discount sledgehammer. Customer either takes the discount and trains your audience to wait, or ignores it. Audit: segment by RFM (Recency, Frequency, Monetary) — high-monetary lapsed customers usually need a reactivation reason, not a discount.
Forget loyalty programs for a week. Pull these numbers first:
| Metric | What it tells you | Benchmark (DTC) |
|---|---|---|
| First-to-second purchase rate | Whether retention is broken at the trigger point. | 25%+ healthy · 15%–25% borderline · <15% leak |
| Median time between purchases | The natural cadence — your nudges should align. | Category-specific (skincare 35d, food 21d, etc.) |
| Repeat customer revenue % | How dependent the business is on new acquisition. | 40%+ resilient · 20%–40% growing · <20% rented |
| Email LTV / SMS LTV | Owned-channel contribution to retention revenue. | 25%–40% of total LTV from email is normal |
| Churn rate (monthly) | List health + lifecycle saturation signal. | 1%–2% organic, 5%+ means saturated |
| NPS by acquisition channel | Whether one channel is bringing bad-fit customers. | Variance >20 points across channels = issue |
You can pull every one of these out of Shopify + Klaviyo without any extra tools. They take a competent analyst two hours to assemble. That two hours saves you six months of retention "tactics" that don't move anything.
Only after the audit. Tactics matched to leak point:
Replenishment reminder timed to category cadence. Bundle/cross-sell offer based on first product. Post-purchase NPS that routes to support if score <7.
Reduce frequency, segment by engagement, suppress the 30%+ inactive bucket from your daily sends. Open rates recover, deliverability follows.
Survey-then-segment. Build a content series for the segment that's struggling. Loyalty program won't fix this — better matching at acquisition will.
RFM segmentation. Reactivation reason (new product, new use-case), not discount. Loyalty program belongs here, not at acquisition.
Loyalty programs work for businesses with a sustained repeat customer base where transaction frequency is the lever — coffee, supplements, beauty, pet food. They don't work as a fix when first-to-second purchase is broken. The customer hasn't earned a tier, hasn't accumulated points, hasn't reached any reward worth caring about — they were already going to leave.
If first-to-second is healthy (25%+), then a well-designed loyalty program can lift LTV by 15–25% over 12 months. Gartner's loyalty research consistently shows the highest-performing programs are tier-based with status visibility, not points-based with discount unlocks. The status matters more than the maths.
| Phase | Days | What gets shipped |
|---|---|---|
| 01 · Audit | 1–14 | Six benchmark metrics pulled, leak point identified, written diagnosis with priority leak. |
| 02 · Design | 14–35 | Tactical fix specced for the priority leak — flow logic, copy direction, segmentation rules, success metric. |
| 03 · Implement | 35–60 | Flow built in Klaviyo, segments created in Shopify, A/B test against control if traffic permits. |
| 04 · Measure + iterate | 60–90 | 30-day window for signal. Either ship a v2 or move to the next leak point. Don't add tactics on top until the first one has a number against it. |
Two related pieces in this series go deeper into specific leaks:
First-to-second purchase rate. Nothing else matters until that one is healthy (25%+). Lifetime value, repeat revenue %, and churn rate are all lagging indicators of whether this one number is moving.
The honest answer depends on category. Mature DTC categories (skincare, supplements) where repeat is the business model: 30–40% of marketing budget into retention. Subscription-native businesses: closer to 50%. Early-stage businesses still finding fit: 80–90% acquisition. The mistake is assuming a fixed split — the right ratio is whatever fixes the binding leak first.
Yes, when they're the default. Customers trained to expect 20% off learn to wait for 20% off. The damage is to acquisition cost on the second purchase — you've effectively reduced AOV without any incremental volume. Discounts work when they're earned (tier-based loyalty), timed (one annual sale event), or segmented (high-RFM win-back). They don't work as a blanket weekly promotion.
Worth it when first-to-second purchase rate is already healthy (above 25%) AND median time between purchases is under 90 days. Below that, the program never gets enough cycles to compound. Above 90-day intervals, the customer forgets the program exists before they earn anything.
Retention is the outcome metric. Lifecycle is the channel and timing strategy that produces retention. Email + SMS + paid retargeting are the typical lifecycle channels; retention is what happens when they're working.
Qwrki runs retention as one slice of an operating layer that also covers acquisition, analytics, lifecycle creative, and the platform underneath. We start every retention engagement with the six-number audit above — usually two hours of work — and tell you which leak point is priority before we suggest a single tactic.
That sounds slow. It saves three months of building flows that won't move the number.
If you want the audit run honestly, book a one-hour call. We'll either tell you what to ship — or that you don't need us yet.
Most ecommerce email is run as a campaign calendar — every Tuesday, every Thursday, every payday. Automation flips that. The brand sends because the customer did something, not because it's Tuesday.
Lifetime value gets quoted in every pitch deck and used in roughly zero operating decisions. Here's the version that actually moves the business — and the four ways it gets calculated wrong.
Anyone can spend on Facebook ads. The difference between spending and running the channel comes down to four signals — and most agencies fail at least two of them.
