Personalisation and Segmentation for Brands Without a Data Team (or a Six-Figure ESP Bill)

Big agencies sell personalisation as a 200-token, real-time, AI-driven, predictive-affinity beast. For a brand sending under 100k emails a month, that's a waste of money and time.

Real personalisation for small brands is two things: stop sending the same email to your best customers and your dormant ones, and use a handful of well-chosen tokens that feel human instead of robotic.

Do those two things and you'll see 30–45% CTR lift on segmented sends versus broadcasts. That is the entire game.

1. Engaged buyers (opened or clicked in last 30 days, ≥1 purchase)

Your top quartile. Send them previews, early access, and your highest-margin products. Never discount to them — they don't need it.

2. Engaged non-buyers (opened or clicked in last 30 days, 0 purchases)

Curious browsers. Send education, social proof, and a soft first-order incentive. Free shipping beats percentage off.

3. Dormant (60–120 days inactive)

On the fence. Send a re-engagement email with a real reason to come back — a restock, a category they previously bought, a meaningful incentive.

4. Sunset candidates (120+ days inactive)

Either re-engage them with one last hard pitch or remove them. Continuing to mail this group is the single biggest reason small-brand open rates collapse.

Use these. Ignore the other 194 your ESP offers.

1. 01{{first_name}} — once per email, mid-sentence, with a fallback ("friend")
2. 02{{last_product_purchased}} — for replenishment, upsell, and cross-sell
3. 03{{abandoned_cart_product}} — for cart and browse abandonment
4. 04{{days_since_last_purchase}} — gating logic for win-back and replenishment cadence
5. 05{{loyalty_points}} — when points are close to a reward threshold
6. 06{{location_city}} — for shipping cutoff, weather, and event-based campaigns

Every token must have a fallback. "Hey ," is the most expensive missed-field error in ecommerce email.

Two changes deliver 80% of the lift you'd ever get from "advanced" personalisation:

1. 01Split your weekly broadcast into two versions — one for buyers (with first-purchase language removed and loyalty language added), one for non-buyers (with social proof and a soft incentive)
2. 02Trigger a replenishment email at days_since_last_purchase = the typical reorder cycle for that product category (28 days for haircare, 60 for skincare, 90 for supplements, etc.)

Before — broadcast to everyone

"It's our birthday! Get 20% off everything this week with code BDAY20."

After — segmented

• Engaged buyers: "We turn three this week. Your reward: a free Brick Loafer keychain with any order."
• Engaged non-buyers: "Three years in. Here's the 20% off we save for people who haven't said yes yet."
• Dormant: "It's been a while, {{first_name}}. Here's the email we saved for you — and free shipping if you come back this week."

Same campaign. Three versions. The segmented version earned 41% more clicks and 28% more revenue per recipient in the test we ran with a Sydney apparel brand.

• ESP — Klaviyo (free up to 250 contacts, $20/month at 500), Mailchimp, or Omnisend
• Segmentation — built into all three above; no separate CDP needed under 50k contacts
• Copy & QA — Milgo ($19–$79/month) for subject scoring, spam checks, and segmented email generation
• Analytics — your ESP's native dashboard is enough until you cross $1M revenue

Milgo lets you generate one campaign brief and produce three audience-specific versions (engaged, non-buyer, dormant) in a single workflow. Every token gets a fallback. Every version is scored against your brand voice and your last 30 days of sends so you don't accidentally repeat language across segments.

You spend twenty minutes a week on email. The list earns more. The math works.