Email Newsletter QA Checklist Before You Hit Send

Before sending an email newsletter, check five things: every fact the AI wrote (prices, dates, codes, claims), every link and merge tag with a test send, how it renders on a phone and in dark mode, whether it's small enough to avoid Gmail clipping, and the deliverability basics: an authenticated sending domain, a visible unsubscribe link and the right audience.

AI changes where the risk sits. Layouts and links break the way they always did, but AI drafting adds a new class of error: confident facts that nobody typed, such as a launch time, a discount percentage or a product claim that sounds right. Those pass a quick skim, which is why the first section of this checklist is about facts, not formatting.

The worked example is an illustrative handmade jewellery seller sending a collection launch to 4,200 subscribers, drafted with an AI assistant from a short brief. The checklist caught nine problems in 25 minutes. Examples from a print shop, a software reseller, an events company, a clothing shop and a sign maker show items the jewellery seller didn't trip over.

Each item below has the reason it matters and how to verify it. Work through the sections in order, because later checks (a test send, for instance) are wasted if the facts are still wrong.

Section 1: facts the AI may have invented

1. Every date and time matches the plan. Why: AI fills gaps with plausible times, and a wrong launch time sends customers to a page that isn't live. How: list every date and time in the email and compare each with the brief, not with your memory.
2. Every price, percentage and discount code is exact. Why: "15% off" and "20% off" are equally plausible to an AI, and codes get "tidied" (AUTUMN15 becoming AUTUMN-15). How: copy each code from the email and apply it at checkout on your own shop.
3. Product names, materials and claims are true. Why: descriptions drift towards generic claims such as "hypoallergenic" or "handcrafted from sustainably sourced materials". How: underline every claim about a product and ask whether you could prove it if a customer or regulator asked. The tutorial on fact-checking AI marketing copy has a method for this.
4. Stock and availability are current. Why: the AI drafts from what you gave it days ago. How: open each featured product and check it's in stock and at the price stated.
5. Nothing personal is assumed. Why: AI likes lines such as "since you loved our autumn range last year", which are false for new subscribers. How: read the email as someone who joined yesterday.

Section 2: merge tags, segments and personal details

1. Merge tags have fallbacks. Why: a first-name tag with no default produces "Hi ," for everyone without a first name on file. How: send a test to a contact record with no first name and read the greeting.
2. The segment is the one you meant. Why: the wrong segment is the most expensive newsletter mistake, because you can't unsend it. An illustrative clothing shop sent a children's clothing promotion to its whole list instead of the parents' segment, and a batch of customers unsubscribed that afternoon. How: check the recipient count against what you expect before scheduling.
3. Suppressions are applied. Why: people who unsubscribed or bounced must not receive it, and recent buyers of a product may not want a discount on the thing they just paid full price for. How: check the platform's suppression settings and exclude recent buyers where it makes sense.
4. Dynamic blocks show the right version. Why: content that changes by segment or location can show a blank block if the condition fails. How: preview as a contact from each segment.

Section 3: links, codes and tracking

1. Every link opens the right page. Why: AI-drafted buttons often link to a homepage, an old product or a placeholder. How: click every link in the test send, on a phone, not in the editor.
2. Discount links apply the code. Why: a link that promises "15% off applied at checkout" and doesn't apply it produces angry replies. How: open it in a private browser window and go as far as checkout.
3. Tracking parameters are consistent. Why: analytics groups visits by the utm_source, utm_medium and utm_campaign tags on your links; inconsistent tags split one campaign into several. How: check that every link carries the same campaign name.
4. No link shorteners or odd domains. Why: they look like phishing, especially in security-related emails. An illustrative software reseller sent a "please update your password policy" newsletter with shortened links, and three customers reported it as suspicious. How: use full links on your own domain or your email platform's tracked links.
5. The unsubscribe and preference links work. Why: a broken unsubscribe link turns into spam complaints. How: click it in the test send and complete the process with a test address.

Section 4: how it looks on a phone, in dark mode and with images off

1. It reads well on a phone. Why: most subscribers will open it on a small screen first. How: open the test send on your own phone and read it all the way down.
2. Dark mode doesn't hide anything. Why: dark logos on transparent backgrounds vanish on a dark background, and some apps invert colours. How: switch your phone to dark mode and open the test again.
3. Images have alt text, and text in images is repeated. Why: some email apps don't load images until the reader allows it, so the alt text is what they see first. A sign maker's newsletter built as one big image showed nothing at all in those apps. How: turn images off in a test inbox, or read the alt text in the editor. For how to write it, see generating and checking alt text with AI.
4. The email isn't clipped in Gmail. Why: Gmail clips emails whose message size is over 102KB and hides the rest behind a "View entire message" link. Mailchimp's help page on Gmail clipping explains that the size counts your text, link tracking code and HTML styling rather than the images themselves. A clipped newsletter can hide your footer, including the unsubscribe link. How: open the test send in Gmail and scroll to the bottom; if you see the clipping notice, cut content or code.
5. The preview text is set. Why: without it, many inboxes show the first line of the email, often "View this email in your browser". How: look at the test send in the inbox list, not just opened.

Section 5: deliverability rules Gmail and Outlook now enforce

These have become requirements rather than good practice, and AI tools can't check them for you. Google's email sender guidelines set out two tiers.

1. Your sending domain is authenticated. Why: Google requires every sender to set up SPF or DKIM, and senders of more than 5,000 messages a day to Gmail accounts to set up SPF, DKIM and DMARC, with the From domain aligned. Microsoft's Outlook.com announced similar authentication requirements for senders of more than 5,000 emails a day, enforced from May 2025. How: your email platform's domain settings show whether your domain is authenticated; if it says "not verified", fix that before anything else.
2. Unsubscribing is easy. Why: Google requires marketing messages from senders over 5,000 a day to support one-click unsubscribe and to include a clearly visible unsubscribe link in the body. How: most email platforms add the one-click headers for you; check the visible link is in the footer and not clipped.
3. Spam complaints stay low. Why: Google asks senders to keep the spam rate reported in its Postmaster Tools below 0.10% and never to reach 0.30%. How: if you send to many Gmail addresses, set up Postmaster Tools and look at the trend after each send.
4. Everyone on the list opted in. Why: people who never asked for your emails are the ones who press "report spam". How: don't import contacts from invoices or business cards into a marketing list without their consent.
5. The footer identifies you. Why: many anti-spam laws require you to identify the sender and offer an unsubscribe, and some require a postal address, which is why email platforms put your address in the footer. How: check the footer shows your business name and the address your platform requires.

The jewellery seller's launch email, checked item by item

The brief given to the AI assistant was five lines: collection name, six pieces, launch at 7pm on Thursday, 15% off for subscribers with code AUTUMN15 until Sunday midnight, and a link to the collection page. Here is what the checklist found in the draft, with illustrative detail:

None of these would have been caught by reading the draft once for tone. Four were facts the AI had supplied itself (the time, the code format, the hypoallergenic claim and the "since you loved" line). The rest were ordinary email problems. The quick sum on whether 25 minutes is worth it: at 4,200 subscribers, a wrong launch time that sends even 2% of them to a page that isn't live is 84 disappointed customers on launch night, and a code that fails at checkout costs sales outright.

If the newsletter was drafted inside your email platform, the same checks apply. Klaviyo's marketing agent, now called Composer, drafts across email, SMS and other channels, and it too needs reviewing before anything goes live. For choosing and setting up the drafting side, see writing a customer newsletter with AI in 30 minutes and the round-up of AI email marketing tools for small businesses.

Running the checklist in 25 minutes with two people

The checklist works best split between two people, because the person who wrote or prompted the draft reads what they meant rather than what's there.

• Person A (the drafter), 10 minutes: sections 1 and 2. They know the brief, so they're fastest at spotting wrong facts and segments.
• Person B (a fresh pair of eyes), 15 minutes: sends the test, then works through sections 3 to 5 on a phone and in Gmail.
• Both: a final read of the subject line and preview text together, out loud.

That final read of the subject line deserves its own few checks, because AI subject-line suggestions tend towards urgency and tricks. Is it true (no "Last chance" when the offer runs for a week)? Does it match the email (no "Your order update" on a marketing email)? Would it make sense to someone who doesn't know you? Many anti-spam rules forbid misleading subject lines, and subscribers punish them with the spam button anyway. Five subject lines an assistant suggested for the jewellery launch, with the verdicts (illustrative):

The chosen subject line had a second benefit: it repeated the launch time, so people who only glanced at the inbox list still got the key fact.

A sole trader can do both halves, with a break between them: draft in the morning, check after lunch. Keep a copy of the checklist in the same folder as your newsletter drafts, and tick it off in a copy for each send so you can see later what was checked. An illustrative print shop that sends a monthly trade newsletter found that the ticked copies were also the quickest way to answer "did we check that price?" when a customer queried one.

Prompts that reduce what the checklist has to catch

The checklist catches errors; the prompt can prevent some of them. The jewellery seller now drafts with these rules at the top of the brief:

Write a launch email from the brief below.
- Use only the facts in the brief. If you need a fact that isn't there
  (a time, price, material, stock level), write [CHECK: what's needed].
- Copy discount codes, times and prices exactly as written in the brief.
- No claims about materials, allergies, sustainability or origin unless
  they appear in the brief word for word.
- Don't refer to anything the reader did before ("since you loved...").
- Write a subject line under 50 characters and preview text under
  90 characters, and repeat any offer shown in an image in the body text.
Brief:
[collection name, pieces, launch time, offer, code, end date, links]

An illustrative result: the next draft came back with "[CHECK: which pieces have matching earrings?]" instead of inventing sets, and the code copied exactly. It still got the end date wrong once, writing "Sunday" where the brief said "Sunday midnight", which matters when people try to use a code at 11pm. The prompt reduced the list of problems from nine to three; the checklist still found the three. An events company that sends open-evening invitations uses the same approach, and its most common catch is a date the AI moved to the wrong day of the week, which is why section 1 compares every date with the brief. If the copy itself reads awkwardly, run it through a plain English and readability check before the test send.

Newsletter QA: more questions

How many test inboxes do I need?

Three covers most small lists: a Gmail account opened in the Gmail app on a phone, an Outlook account on a computer, and the Mail app on an iPhone if many of your subscribers use iPhones. Your email platform's reports show which email apps your subscribers actually use, so match the test inboxes to the top three.

Should I still look at open rates?

Treat them as a rough signal at best. Privacy features in some email apps load images automatically, which inflates opens, and Google says it doesn't track open rates at all. Clicks, replies, sales and unsubscribes tell you more about whether a newsletter worked.

Does every newsletter need a plain-text version?

Most email platforms create one automatically from your HTML. Open it once in the platform's preview, because AI-drafted emails with lots of buttons and images sometimes produce a plain-text version full of bare links and missing words.

What if I find a mistake after sending?

If it's a wrong fact that affects customers, such as a price, date or discount code, send a short correction to the same segment quickly and honour the original where it's reasonable. For typos or a broken image, a correction email usually does more harm than good; fix it in the web version and move on.

Further reads

• Email Marketing Tools With AI: What a Small List Costs per Month — What email tools with AI cost for a small list.
• What Can AI Do for Your Email Marketing? 10 Practical Uses — Practical ways to use AI in email marketing beyond drafting.
• Social Media Pre-Publish Checklist for AI-Drafted Posts — The equivalent checklist for social posts.
• AI Content Approval Workflow: Draft, Check, Sign Off — Put the checklist inside a draft, check and sign-off workflow.
• Tone of Voice Examples: AI Copy Before and After Editing — Before-and-after edits to make AI newsletter copy sound like you.
• How to Catch Outdated Information in AI Answers — Catch out-of-date facts before they reach subscribers.
• How to Set Up AI Abandoned Cart Emails for a Small Online Store — Turn on the free built-in checkout reminder, add two follow-ups, and let AI write the copy while plain rules decide who is emailed and when.
• Winning Back Lapsed Dental Patients With AI Messages — How to find lapsed dental patients, split them by why they stopped coming, and use AI to write reactivation messages that don't guilt or scare.
• Turning One Online Course Into Emails and Posts With AI — Transcribe the lessons, pull out one idea, mistake and quick win per lesson, and turn each into an email or post. Twenty lessons can fill two months.
• Marketing a Physio Clinic With AI: Posts, Tips, and Newsletters — A physio clinic's marketing routine built on AI: caseload-led topics, voice notes turned into posts and newsletters, claims checks and booking tracking.
• How to Write Cold Emails With AI That Stay Out of Spam — The sending set-up, DNS records, warm-up pace and AI prompt that keep cold emails in the inbox, with a before-and-after rewrite and warning signs to watch.
• How to Build an Upsell and Cross-Sell Email Sequence With AI — Find what customers really buy next, plan a short triggered sequence, draft it with AI from a proper brief, and stop it the moment someone buys.
• How to Check AI Translations Before Customers See Them — A three-layer check for AI-translated marketing: checks anyone can run, a second-AI review, a native reviewer's brief and an error score that decides go-live.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Google Workspace Admin Help (Email sender guidelines: authentication, spam rate, one-click unsubscribe); Mailchimp help (Gmail clipping at 102KB); Microsoft Tech Community announcement of Outlook requirements for high-volume senders (2025); series fact sheet for Klaviyo Composer (checked September 2026).