Email went from 9% to 34% of total DTC revenue for a mid-market beauty brand.

Context

Mid-market DTC beauty brand at $26M run-rate. They had a reasonable paid-social business and a growing retail wholesale footprint. Email was treated as "free money" — one weekly batch send to the whole list, two underfed abandoned-cart flows, and a sign-up form offering 10% off.

The CMO came to us saying: "Everyone tells me email should be 25-35% of DTC revenue. Mine is 9%. Fix that without burning my list."

The audit

• One weekly batch to the whole list. No segmentation. No send-time optimization. Open rates had drifted from 24% to 11% over 18 months.
• Only 2 flows live. Abandoned cart (missing a step and a wait condition misconfigured) and a 2-email welcome.
• Sign-up form was a popup giving 10% off to everyone. No tiering, no exit intent variant, no quiz / value-prop collection.
• No cohort retention view. They couldn't see repeat-purchase behavior at a cohort level. Impossible to tune LTV.
• SMS was a separate tool with no overlap awareness. Customers were getting the same offer across both channels on the same day.

"The 9% email share wasn't because our list was bad. It was because we weren't talking to our list like real customers."— CMO

Strategy

Pillar 1 — Rebuild the sign-up experience

Replaced the one-size-fits-all popup with a 2-step quiz: skin type → concerns. Zero-party data captured on signup, used to segment and personalize the first 90 days of lifecycle. Sign-up rate +74%, and new subscribers now carried useful profile data.

Pillar 2 — Build out 11 lifecycle flows

Welcome (segmented by quiz output, 5 variants). Abandoned browse. Abandoned cart (3-touch). Post-purchase (routed by product). Replenishment (dynamic timing by product consumption rate). Winback (4 segments based on recency). VIP (top 10% by LTV). Review request. UGC request. Birthday. Sunset (list hygiene).

Pillar 3 — Rebuild the campaign cadence

Moved from 1 batch send/week to 3–4 segmented campaigns/week. Segmentation by (a) purchase recency, (b) quiz-derived profile, (c) engagement tier. Build a merchandising calendar aligned with product launches, holiday calendar, and content-partnership moments.

Pillar 4 — SMS as a complement, not a competitor

Integrated SMS with email in Klaviyo. Suppressed email send 48 hours after an SMS send on the same offer. SMS used for time-sensitive moments (restocks, flash sales, abandoned-checkout last-chance) only. Not a parallel channel for every campaign.

Pillar 5 — Install predictive LTV

Built a predictive LTV score (first-order propensity model trained on 24 months of order history) and routed first-purchase cohort into higher- or lower-touch lifecycle paths based on the score. High-score cohorts got the VIP track from order 1; low-score got the discount-driven winback focus.

Timeline

Tech stack

What actually drove the number

• The quiz-driven signup → segmented welcome. Day-30 revenue per new subscriber lifted 3.2× because people got relevant product routing from the first email.
• Replenishment flow. Highest-ROI single flow. Dynamic timing based on actual product consumption rate (we modeled this per-SKU). Second-order rate on hero SKU jumped from 21% to 54%.
• VIP flow + predictive LTV routing. Top-10% LTV customers now get a differentiated experience (early access, limited drops, personal notes). AOV in this cohort up 44%.
• Segmented campaigns over batch. Open rate +118%. Unsubscribe -37%. Deliverability metrics improved alongside send volume increasing.

What we got wrong

• First version of the abandoned-browse flow was too aggressive. 3 emails in 72 hours felt stalker-y based on reply volume. Dialed to 2 emails over 4 days. Complaints down 71%, revenue flat.
• Predictive LTV model trained on insufficient data on first pass. We used 12 months. 24 months was needed for seasonality. Retrained in month 7 — accuracy up 18%.

"Email went from a nice-to-have to the single biggest driver of our DTC business. It also made our paid acquisition more efficient because we could afford a higher first-order CAC knowing the lifecycle would compound."— CMO

What's next

Engagement ongoing. Current focus: a dedicated subscription product (replenishment-as-a-subscription), a loyalty program integrated with the email program, and testing a GPT-based personalization layer for product-copy variants per segment.