Flo

What are the most common mistakes in AI-designed emails?

By the Flo team · Updated

Short answer

The most common mistakes are made-up or wrong products, invented claims or prices, text placed over the product or a face, generic layouts that ignore the brand, the same image reused across a series, missing or wrong footer details, and copy that sounds like AI. A quick checklist review catches almost all of them before the client does.

The checklist

MistakeWhat to look for
Wrong productGenerated or outdated packaging, wrong variant
Invented claimsBenefits, stats, reviews or prices the brand never stated
Text over the subjectHeadline or button covering the product, a hand or a face
Generic lookDefault fonts, colours that are close but not the brand's
Logo problemsWrong file, unreadable on the background, missing
Repeated imagesSame photo in back to back emails of a series
Footer gapsMissing unsubscribe link or address, wrong social links
AI sounding copyLong dashes everywhere, "elevate", "unlock", groups of three
Stretched or cut photosDistorted products, jagged cutouts

Why they happen

Most of these come from missing inputs, not bad models. A tool without the real product photos generates them. A tool without the brand's rules guesses. A tool that never looks at its own render cannot see that the headline covers the product.

Copy that sounds like AI

Readers have learned the tells: long dashes, "in today's fast-paced world", "elevate your routine", neat lists of three. Ask for short, specific copy in the brand's own voice, and cut anything a person at the brand would not say.

Claims are the serious one

An invented "clinically proven" or a made-up review is not a style issue. It is a trust and compliance risk for the client. Every claim in an email should come from the brand's own site or the brief.

Build the check into review

Do not rely on spotting problems by eye. Run the checklist above at internal review, before the client sees the email. See how to review AI-designed emails.

How Flo avoids them

Flo works from each brand's real product photos and wording, does not invent claims or prices, and reviews its own desktop and mobile renders against the brief before an email reaches your board, including checks that text sits clear of the main subject. It writes without long dashes, and every email gets a footer with the unsubscribe link and the brand's address as live text.

Questions agencies also ask

Do AI email tools make up product features?
Some can, if they are not grounded in the brand's own content. Check every claim against the brand's website or the brief.
How do I make AI email copy sound human?
Keep it short and specific, use the brand's own words, and remove filler phrases and long dashes. Read it aloud as the brand would say it.
What is the single most important check?
That the product shown is the real, current product. It is the mistake clients spot first.