All articles
Every guide and tool page, grouped by topic, newest first within each.
AI in retention
How AI is used in retention work: churn prediction, segmentation, email copy, agents, and how to check what a model gives back.
26 articlesWhat is prompt injection, and how can it affect an AI marketing agent?
See how instructions hidden in source material can redirect an AI marketing agent. Use a fictional campaign example to check permissions, approvals and tool actions.
How do you catch AI hallucinations in marketing emails?
Check AI-written email claims against current product and offer records. Use a worked audit to catch invented benefits, missing exclusions and false promises.
Will AI replace email marketers?
AI can replace email-writing tasks without replacing the whole role. Examine the evidence, audit your responsibilities and choose useful skills to build.
Can AI find retention problems in customer feedback?
AI can surface patterns in reviews, surveys and support messages. Learn how to verify a theme, check whether it relates to repeat purchasing, and test a specific fix.
How do AI spam filters decide what reaches the inbox?
See which spam-filter signals Gmail and Microsoft disclose, why authentication is not an inbox pass, and how to diagnose a real placement problem.
Can AI explain why a customer might churn?
A churn model can explain its score, but not prove a customer's motive or the effect of an offer. Learn how to read feature attributions and turn them into a testable retention decision.
Do loyal customers react differently to AI customer service?
Loyal customers may read AI service as a change in how the company values them. See what the research measures and how to choose between AI, human and assisted service.
What customer data should an AI email assistant be allowed to access?
Choose customer data and permissions for an AI email assistant by the job it must do. Check read, write and retention access before connecting your store or ESP.
When should AI customer service hand off to a human?
Route customer service issues to a person before an AI's limits turn into a failed conversation. Use a clear transfer rule, context packet and service checks.
Can you trust an AI-generated email performance report?
Audit an AI email report against its source, metric definition and order records before using its conclusions. Includes a conversion-rate check and a decision record.
How do you check an AI-created email segment before sending?
Check AI-generated audience rules against your campaign brief, subscription requirements and real profiles. Use an inclusion-and-exclusion test before sending.
Do AI-generated email subject lines perform better?
AI subject lines can win a specific email test, but the evidence does not show a universal advantage. Compare matched sends and judge clicks or orders, not opens alone.
How to make AI email copy sound like your brand
Build a reusable AI email voice brief, separate style examples from current campaign facts, and review the draft with a worked rewrite and checklist.
Does AI actually save time on email marketing?
Measure AI email production from brief to approval, including editing and rework. Use a time-budget worksheet to decide which tasks are worth handing over.
Churn prediction vs uplift modelling: which customers should you target?
Churn prediction ranks risk. Uplift estimates what an offer changes. Use a worked renewal example to choose recipients after incentive and contact costs.
How does AI personalise loyalty programme rewards?
AI can choose a loyalty reward and the requirement to earn it. Learn how the model uses member data, learns from response and accounts for incentive costs.
What is agentic AI in retention marketing?
Agentic AI selects and revises actions within a retention workflow. Learn what distinguishes it from a preset flow or AI-written copy, and what control to check.
What is AI customer retention?
AI can predict which customers may leave, choose an intervention, or help deliver it. The retention question is whether that decision makes more customers return or renew.
Do AI agents improve customer retention?
Trials show that AI agents can change engagement and service speed. Neither result alone proves customers stay longer. Here is what the evidence supports and how to test retention lift.
Does AI-written email copy improve retention?
Research finds useful short-term results for AI-written email copy, but not a general retention lift. Here is how to test the copy method against repeat purchasing.
How is AI customer segmentation different from RFM segmentation?
RFM groups customers by purchase recency, frequency, and spend. AI segmentation may group by similarity across chosen data or target a predicted outcome. Here's how to compare them.
Next best action vs next best offer: what is the difference?
Next best offer picks a sales offer or promotion to show. Next best action can also choose a message, service support, or no contact.
What is next-best-action in retention marketing?
Next-best-action chooses what to do for one customer now—not just who is likely to respond. See the action set, rules, ranking, and holdout a retention policy needs.
What is predictive churn, and how do the models work?
Predicted churn is a per-customer estimate of churn risk, and three different quantities travel under the name. How buy-till-you-die models and churn classifiers work, why the definition of churn is something you choose, and what to check before you act on a score.
How is predicted customer lifetime value calculated?
Two ways to calculate predicted CLV: models that read only each customer's order history, and machine learning trained on past customers. How each one works, why customers with no repeat purchase are the hard part, and how to check the number.
What does AI actually do in email marketing?
AI gets applied at six points in an email programme, and the vendor documentation is precise about the mechanical ones and silent about the decisions. What the help pages promise, what they refuse, and where the automation stops.
Email flows
The automated emails a store sends, what starts each one, and how to set them up.
9 articlesWhich email flows should an online store set up first?
Start with welcome, checkout recovery and post-purchase emails. Use your audience, customer gaps and working event data to decide which flow to build first.
How do you stop browse, cart and checkout emails overlapping?
Set stage-specific stop rules, recheck eligibility before sending, and trace event delays or duplicate senders. Includes a handoff table and an acceptance-test record.
Why is my email flow not sending?
Trace an absent flow email through entry, waiting, manual review, skips and delivery. Fix future sends and decide what happens to missed messages separately.
What is a replenishment email?
A replenishment email reminds a customer to reorder a product shortly before it runs out. Setting one up is easy. The hard parts are picking the delay and telling whether the email adds reorders or just takes credit for ones already on the way.
What is a browse abandonment email?
One platform sends it when a shopper views a product. Another sends it only to people who never reached a product, and sells the first version under a different name. The delays run from one hour to twenty-four.
What is a win-back email?
A win-back email is an automated message or sequence sent to people who previously bought or engaged and then went quiet. The hard part is not the phrase. It is the inactive threshold: published defaults run from 45 days to 360 days, while several guides say the right clock depends on the buying cycle.
What is a post-purchase email flow?
A post-purchase email flow is the automated sequence a store sends after someone buys. The published guides disagree about which emails are in it, how many it holds, and whether the clock starts at the order or at delivery — and three of them disagree with themselves. What each version means, and what to write down beside your own number.
What is a welcome email flow?
A welcome email flow is the automated sequence a store sends after someone joins its list. The published guides disagree about how many emails it holds, how fast the first one goes out, whether a first purchase belongs in it, and what the open rates everyone quotes are counting. What each version means, and what to write down beside your own number.
What is an abandoned cart email?
One platform sends it when a shopper adds to their cart. Another sends it only after checkout starts. A third starts a ten-minute clock once an email address is typed. The 70% figure everyone quotes counts a fourth group again.
Deliverability
Getting email into the inbox: authentication, consent, bounces and list hygiene.
11 articlesWhat is sender reputation, and how do you check it?
Understand domain and IP reputation, choose the right email reputation check, and interpret provider dashboards, sender scores and blocklist results.
Transactional vs marketing email: what is the difference?
Classify receipts, abandoned-checkout reminders and stock alerts by what they say. Compare US and UK mixed-content rules, permission checks and unsubscribe duties.
Is Gmail's Promotions tab spam?
Separate Gmail Promotions from Spam, understand why authentication does not guarantee Primary, and record placement before trying to change it.
Hard bounce vs soft bounce: should you resend the email?
Learn the difference between hard and soft email bounces, read the full failure reason, and decide when to retry, suppress an address or investigate your sending setup.
SPF vs DKIM vs DMARC: what does each one do?
Understand SPF, DKIM and DMARC, why passing checks can still fail alignment, and how to audit authentication on an actual marketing email.
Should you use a subdomain for email marketing?
Separate marketing, transactional and staff email with a practical sending-domain plan. Check authentication, actual routing and the limits of reputation isolation.
What is one-click unsubscribe, and does a footer link count?
Understand Gmail and Yahoo's one-click unsubscribe rules, then audit message headers, request handling and the contacts your marketing sends exclude.
Single vs double opt-in: which should you use for email?
Choose an email opt-in method, audit confirmation and welcome-message failures, and compare usable subscribers without mistaking a higher rate for more customers.
Can you email people who never opted in?
Compare US, UK and Canadian email consent rules, then check whether customer records, bought lists or abandoned checkouts belong in a marketing send.
How do you improve email deliverability? A guide to finding and fixing the problem
Diagnose rejected messages, spam placement and weaker engagement, then fix authentication, audience quality and sending patterns using current mailbox-provider guidance.
Is email list cleaning worth it? When to suppress contacts or pay for validation
Separate routine list cleaning from paid email validation. Decide which contacts to exclude, suppress or keep, and check whether cleaning actually lowers your bill.
Campaigns and content
Writing, timing and sending campaigns, and how emails render.
7 articlesHow do you write marketing email subject lines that get opened?
Write subject lines from the offer and content, keep important conditions, and check the inbox preview. Includes original examples and a practical review record.
Why does Gmail clip your emails, and how do you fix it?
Find why Gmail hides part of your marketing email, reduce the right code, and check the delivered message before calling the clipping problem fixed.
Why does dark mode change your email colours?
Find out why dark mode changes email colours, which rendering problems need fixing, and how to test logos, buttons and text in an actual delivered message.
How often should you send marketing emails?
Choose an email frequency by counting campaigns and flows per customer, resolving overlaps, and testing whether another send adds value without unacceptable opt-outs.
When is the best time to send marketing emails?
Choose an email send time for your audience. Compare fixed schedules with AI optimization, test the outcome you need, and keep time zones and offer deadlines clear.
Is email marketing still worth it in 2026?
Decide whether email marketing earns its budget. Count the full cost, calculate the extra orders needed to break even, and separate credited revenue from profit.
What is email list segmentation?
Email list segmentation splits one list into groups that each get a different email. Building segments is easy. Knowing which ones change what people do is harder, and field experiments show the obvious picks are not always the ones that respond.
Email metrics
What open rates, clicks and email-attributed sales can and cannot tell you.
4 articlesWhat is a good email click rate?
Judge your email click rate with a dated ecommerce benchmark, the right denominator and consistent bot filtering. Includes a worked example and diagnosis table.
What are bot clicks in email, and should you filter them?
Understand email bot clicks, check suspicious engagement, and audit filtering across reports, attribution, segments and click-triggered automations.
Can you trust email open rates?
Understand how privacy loads, blocked images and bots affect email open rates. Check report settings before judging a campaign or resending to non-openers.
How do you track which emails make sales?
Track email sales with consistent campaign tags, verified purchase data and documented attribution rules. Use an order example and checklist to reconcile reports.
Store metrics
Retention rate, churn, lifetime value and the other numbers a store runs on, with the formulas side by side.
6 articlesWhat is average order value?
Twelve of fifteen publishers divide revenue by orders. They do not mean the same thing by revenue — three strip taxes and shipping, one keeps them, and a store platform's own column is a third thing again. What AOV counts, and why yours may not compare.
What is purchase frequency?
Nine publishers give the same ratio: total orders divided by unique customers. Two of them then print a repeat-buyer benchmark further down the same page without saying it is a different quantity. What the number counts, and why yours may not compare.
What is repeat purchase rate?
Repeat purchase rate is the share of customers who bought more than once. The published formulas disagree about which customers go in the denominator, about whether the window is a month or a lifetime, and about whether one of the metric's own names points at a completely different number. What each version measures, and what to write down beside it.
What is customer churn rate?
Churn rate is the share of customers a store loses over a period. The published formulas disagree about who is in the denominator, about what counts as leaving, and about whether the number is monthly or annual. What each version measures, and what to write down beside it.
What is customer lifetime value?
Your platform shows what a customer has already spent, plus a forecast of the next year. The published formulas run on a lifespan of several years, and some multiply by profit margin. What each version answers, and where they disagree.
How is customer retention rate calculated?
Klaviyo's own example gives 64%, dividing repeat buyers by one-time buyers. Divide the same repeat buyers by all 92,000 customers and it is 39%. Seven formulas answer the question.
Retention strategy
Getting customers to buy again, and why they leave.
5 articlesWhich customer retention strategies work for an online store?
Choose ecommerce retention strategies from the customer problem: product support, replenishment, win-back or rewards. Check repeat purchases and contribution.
How do you turn Black Friday buyers into repeat customers?
Build a first-time BFCM buyer cohort, check the first order before follow-up, and judge second purchases after discounts and returns rather than assuming a sale buyer needs another coupon.
Why do shoppers abandon checkout?
Use the funnel, payment events and a checkout test to separate browsing from payment failures and purchase barriers before changing recovery email.
What is RFM segmentation?
RFM segmentation scores every customer on recency, frequency and monetary value. The three labels are settled. The scale is not: published models run 1-3, 1-4, 1-5 and 1-10, the direction flips between vendors, and the same code 111 names the best customer in one system and the worst in another.
What is retention marketing?
Retention marketing is the work of keeping the customers you already have. The two numbers most often used to justify it trace to 2000 and — by one guide's account — 1990, and one of them is routinely credited to a document that does not contain it.
Klaviyo
7 articlesKlaviyo reporting in 2026: which reports to use and how to read them
Find the right Klaviyo report, build campaign and flow reports, schedule updates, and check why revenue totals or reporting dates differ.
Klaviyo AI in 2026: Composer, predictions and MCP explained
What Klaviyo AI can do for your marketing, how Composer pricing works, and what to check before using predictions or connecting an external AI assistant.
Klaviyo integrations in 2026: what connects and how to check the sync
Connect your store and other apps to Klaviyo. Understand what data syncs, set up Shopify, check customer events and troubleshoot missing information.
Klaviyo features in 2026: what you can do and what costs extra
Explore Klaviyo's campaigns, flows, segments, forms, integrations, SMS, reporting and AI. See what each does, what needs setup and which extras need a separate plan.
Who is Klaviyo for? Fit, learning curve and support (2026)
Does Klaviyo fit your business type, size and team? See what it takes to set up and learn, and what support each plan includes.
What is Klaviyo? A plain review for store owners (2026)
Explore Klaviyo’s email builder, automated flows, segmentation, SMS, integrations and AI, with product screenshots, pricing and advice on who it suits.
Klaviyo pricing in 2026: what you pay at your list size
What Klaviyo's Email and Marketing plans cost at 2,500, 10,000, 25,000 and 100,000 contacts, who Klaviyo counts as a contact, where the bill jumps, and how to keep it down.
Best-of lists
Email marketing platforms compared for a specific need, with sources and dates alongside the recommendations.
1 article