The modern email guide —
the highest-ROI channel, rebuilt.

01Why email still outperforms every channel people claim is replacing it

Three reasons email keeps winning, year after year:

• You own the relationship. Every other channel rents attention from a platform that can change the rules overnight (Facebook organic reach, Google algorithm, iOS tracking). Email is yours.
• The inbox is high-intent. People who open a marketing email have actively chosen to be reachable by your brand. Compare that to paid display where your ad interrupts someone mid-scroll.
• It compounds. Every new subscriber adds permanent reach. Every new automation added to the system runs 24/7 without ongoing cost. Paid media resets to zero when you stop paying; email keeps sending.

Email is the only owned channel that compounds. Every other “owned” channel is actually rented from a platform that eventually changes the terms.

The chart below shows typical open rates by industry. The gap between categories is less about medium and more about list quality — non-profit lists tend to be recently-opted-in and mission-aligned; e-commerce lists often include legacy coupon-hunters who don't engage any more. Getting the list to the right shape is half the battle.

02The lifecycle framework: welcome, nurture, convert, retain, win-back

Every subscriber moves through five stages. Your programme should have touchpoints at each.

1. Welcome (day 0–10)

First impression is the entire relationship. A strong welcome series delivers the promised incentive, introduces the brand, and lands the first purchase (e-commerce) or the first qualified action (SaaS / B2B). 4–5 messages over 10 days is the standard cadence. Welcome sequences are consistently the highest revenue-per-send of any automation.

2. Nurture (active subscribers not converting yet)

Value-forward content: product education, customer stories, how-to material. Mixed with relevant offers. Weekly cadence works for most brands; bi-weekly for more technical audiences. The mistake here is treating nurture as “newsletter” — a good nurture programme is segmented by interest and lifecycle signals, not a single bulk send.

3. Convert (triggered by behaviour)

Abandoned cart, browse abandonment, post-purchase cross-sell, price-drop alerts, back-in-stock notifications. These are the automations that print money. If your programme is missing any of them, fix that before tuning anything else.

4. Retain (existing customers)

Replenishment reminders, loyalty programme touchpoints, VIP tier communications, review requests. Retention emails drive 2–3× higher AOV than acquisition emails because you're not paying the conversion tax — these people already trust you.

5. Win-back (dormant, 90+ days since engagement)

A focused sequence offering something meaningful (a real discount, a survey, a product update). After 2–3 attempts without response, move to a sunset list. Keeping unengaged subscribers hurts deliverability for the rest of your list — trimming them aggressively is a deliverability move, not a “we lost them” concession.

03Segmentation that actually moves revenue

“Segmentation” is one of the most abused terms in email. A list of 50 segments nobody uses is just a list. A handful of segments you actually send to with different copy + offer is what earns revenue.

The four segmentation axes that matter

• Lifecycle stage. New subscriber, active, dormant, churned. Drives everything.
• Behavioural signal. Recent purchase, browse behaviour, email engagement in the last 30 days, cart activity.
• Predicted value. High-value vs. low-value based on purchase history, LTV model, or firmographic data for B2B.
• Interest or category. What they've shown interest in — product category, use case, industry, persona.

Combine two or three of these per send and you have a targetable audience. For B2B: “Fintech prospects who downloaded the whitepaper but haven't booked a demo” is a useful segment. “Everyone who ever gave us an email” is not.

04Automation flows — the backbone of a modern programme

Automations are the leverage of your programme. Built once, they run forever. The core flows worth investing in, ranked by typical revenue contribution:

• Abandoned cart — highest revenue-per-send. 2–3 messages across 48 hours: reminder, objection handling / social proof, final offer. Don't discount in email 1 — wait until email 3 if at all.
• Welcome series — 4–5 emails across 10 days. Deliver incentive, introduce brand, drive first purchase.
• Browse abandonment — lighter-touch cousin of abandoned cart. Triggered when someone viewed a product but didn't add to cart.
• Post-purchase — receipt, setup help, cross-sell, review request. Spread across 30 days. Review requests at day 14.
• Win-back — 3-touch sequence for dormant subscribers. Last email includes a survey or sunset notice.
• Back-in-stock / price-drop — triggered notifications. Huge revenue-per-send when the product matches interest.
• Birthday / anniversary — nice-to-have. Works for lifestyle brands; skip for B2B.

05Deliverability: the invisible tax on sloppy senders

You can have the best copy and the perfect offer — if your emails land in spam, none of it matters. Deliverability is the single underrated discipline in email marketing.

The non-negotiables

• Authenticate. SPF, DKIM, and DMARC correctly set on your sending domain. Since February 2024, Gmail and Yahoo require all three for bulk senders, plus one-click unsubscribe. Non-compliance means your messages silently go to spam or get rejected.
• Warm sending domains and IPs. New domains need a 4-week ramp, starting with engaged subscribers, to build reputation.
• Keep the list clean. Remove hard bounces immediately. Suppress anyone who hasn't engaged in 180 days. Never mail addresses that complain or unsubscribe.
• Match subject to content. Mismatch is the fastest way to trigger filter heuristics.
• Use a reputable ESP. Shared IP pools vary wildly in reputation. A dedicated IP is worth it for senders over 100K/month.

The funnel chart below shows where messages disappear. Only 0.5% of 10,000 sends typically converts — and at every stage, deliverability leaks are compounding losses.

06AI in email: personalisation, optimisation, and the end of templates

“AI in email” five years ago meant a subject-line optimiser. Today it means something fundamentally different: every send can be reshaped per subscriber without the marketing team writing 5,000 variants.

Where AI is actually doing work in email programmes

• Dynamic content blocks. Product recommendations, featured content, and offers selected per subscriber at send time based on behaviour + predicted interest.
• Send-time optimisation. Learning each subscriber's individual open-most-likely window and scheduling to hit it.
• Subject-line generation + testing. LLMs draft 5–10 variants to a brief; split-test in real time; winner is promoted to the full send. Eliminates the 40-minute subject-line committee.
• Churn-risk prediction. Models score every subscriber for predicted engagement decay. Win-back triggers fire before the subscriber fully goes dark.
• Frequency calibration. Some subscribers want daily; others want monthly. AI calibrates per-subscriber frequency to minimise unsubscribes while maximising revenue.

The chart below shows typical lift across the four core metrics when a baseline programme adds AI personalisation. The unsubscribe reduction is as important as the revenue lift — each unsubscribe erodes the long-term asset value of the list.

07Testing, iteration, and the cadence of improvement

Every mature email programme has a testing rhythm. The programmes that stagnate are the ones running the same template every week.

What to test, in order of impact

• Offer. Biggest lever. “10% off” vs “free shipping” vs “bundle deal” produce dramatically different responses even on the same segment.
• Subject line + preview text. Still the entry gate. A/B test 2 variants on 20% of the list, send winner to remaining 80%.
• Send time. Tuesday 10am isn't universal. Test your audience's actual peak engagement windows.
• Call-to-action placement + copy. Above-fold vs mid vs end. “Shop now” vs more specific CTAs. Small lifts compound.
• Visual vs text-heavy design. Text emails often out-click image-heavy emails for B2B; reverse for e-commerce lifestyle brands.

Don't A/B test more than one variable at a time on a single send, and don't call a winner before 2–3 hours for a mid-size list. Stat-sig matters — a 1% lift on 500 subscribers is noise.

08Measurement — beyond opens and clicks

Apple Mail Privacy Protection broke open-rate measurement in 2021 — opens are now partially bot-generated and can't be used for fine-grained decisions. What matters instead:

• Revenue-per-send (RPS). Total revenue attributed to the email ÷ total sends. The honest metric.
• Revenue-per-subscriber per month. Programme-level ROI. Target varies by industry — $5–15/subscriber/month is a reasonable e-commerce benchmark.
• List growth net of decay. New subscribers minus unsubscribes/sunsets. Flat or negative net growth is an emergency.
• CTR and conversion rate. Click-through tells you whether the offer resonated; conversion rate tells you whether the landing page delivered.
• Unsubscribe rate + spam complaint rate. Leading indicators of deliverability decay. Any send above 0.1% spam complaint needs investigation.

The ESP reports all of these but too many programmes only look at opens and clicks. Revenue-per-send is where the real tuning decisions happen.

09Choosing an ESP and building the stack around it

Your ESP is the centre of the email stack. The right choice depends on what you're sending and the other tools around it.

• Klaviyo — Shopify-native e-commerce. Strong segmentation, predictive models, deep integrations. Most of our e-com clients are here.
• Customer.io / Braze — event-driven SaaS. Best when you're sending based on product usage patterns, not commerce events.
• HubSpot / Marketo / Pardot — CRM-native B2B. Good when email is one channel inside a broader marketing automation stack; weaker deliverability at scale than specialist ESPs.
• Mailchimp / ActiveCampaign — SMB workhorses. Fine for simple programmes; automation features lag behind Klaviyo.
• Campaign Monitor / Moosend — newsletter-first. Strong for content businesses.

Around the ESP: a CDP or data warehouse feeds enriched subscriber signals in. Your CRM tracks the downstream lead/customer state. InsightAX or your attribution tool ties every email send back to revenue. The email tool is the sender; the surrounding stack is what makes it smart.