Product descriptions have all the time been a significant ecommerce promoting software. However crafting distinctive and persuasive copy is difficult.
So nearly a decade in the past, I wrote what has develop into one among my favourite articles, “Methods to ‘Manufacture’ Product Descriptions for Ecommerce.” The piece describes a composition course of and, finally, produces this description for a soup spoon:
Hungry for some hearty rooster noodle or creamy clam chowder? This soup spoon has a big bowl meant to haul bisques and broths to your mouth. In truth, this soup spoon can maintain about thrice as a lot soupy goodness as your commonplace desk spoon. You possibly can take thrice as many bites, or you would purchase this soup spoon and slurp giant.
Descriptions at Scale
I had been scuffling with product descriptions on the web site of my employer, a regional brick-and-click farm-and-ranch retailer. We had hundreds of things needing descriptions for our ecommerce web site. I attempted three choices: Amazon’s Mechanical Turk, a Romanian product description company, and a template-based software.

Synthetic intelligence has made it simple to provide product descriptions in a mechanical or algorithmic means, nearly like manufacturing.
Writers employed by way of the Turk or in Romania tended to strategy every description like a novel with a plot. Some descriptions had been glorious, however not all. The unhealthy ones missed the purpose.
I then devised the product description “manufacturing” course of.
- Concentrate on a single factor. The merchandise’s prime function or aggressive benefit.
- Connect with a profit. Clarify how the function helps the patron.
- Make clear and specify. Make the feature-and-benefit bridge exact and comprehensible.
- Take a look at your phrases. Take away redundancy, add motion verbs, and spark curiosity.
- Spell and grammar examine. Sloppy textual content erodes belief and harms conversions.
- Add an introduction. Polish, humanize, and insert a hook.
- Add a name to motion. Encourage the sale.
The method was a framework for writers and advertising groups.
Mad Libs
Sadly, the method took quite a lot of time. So we doubled down on a tactic that I likened to enjoying “Mad Libs” — a template-based phrase sport — and wrote one other article, “Generate Product Descriptions with Pure Language Software program.”
Pure language templates had been forerunners to right this moment’s generative synthetic intelligence platforms. The templates used product-category variables to mass-produce copy.
A template for LED televisions would possibly seem like this:
{} {{ display screen measurement >= 55 ? {} “viewing”: “the exhibits and films you’re keen on: }} from the consolation of your {} with this {{ display screen measurement }} {{ model }} {{ display screen kind }} tv.
That instance got here from Greatest Purchase:
Expertise cinema-like viewing from the consolation of your lounge with this 55-inch Insignia LED tv. It shows Blu-ray and high-definition motion pictures in full 1080p decision with gorgeous HD element. Use the three HDMI inputs to create a house theater expertise with this Insignia LED TV and your different audio and video units.
Generative AI
I’ve used the “manufacturing” and natural-language processes on dozens of ecommerce websites.
However generative AI makes the duty simpler with higher outcomes — morphing from templates and pure language into immediate engineering.
Begin with Google’s five-step immediate engineering framework: Activity, Context, References, Consider, and Iterate.
Activity
Instruct the AI on what to do. Right here the intention is to provide a product description specializing in a single function.
First, produce a template-like immediate:
Compose a compelling product description for [Product Name], specializing in the [Detailed Key Feature]. Describe how the [Short Key Feature] advantages clients, offering particular examples and comparisons.
The variables could possibly be:
- [Product Name] – “a soup spoon.”
- [Detailed Key Feature] – “giant spoon bowl that holds soup.”
- [Short Key Feature] – “giant bowl.”
The immediate addresses two parts of my authentic manufacturing course of: “concentrate on a single factor” and “hook up with a profit.”
Context
Generative AI is healthier with context. So to engineer our immediate, we as soon as once more “hook up with a profit” whereas including the “make clear and specify” step to incorporate our goal buyer.
Think about you might be an skilled copywriter engaged on product descriptions. Compose a compelling product description for [Product Name], specializing in the [Detailed Key Feature]. Describe how the [Short Key Feature] advantages clients, for instance, [Benefit Bridge], offering particular examples and comparisons.
Embody a direct deal with name to motion, too. Be aware [Product Name] is for [Target Audience]. Use a [Brand Voice] tone. Guarantee the outline is partaking, grammatically appropriate, and avoids repetition. Use a various, related, and significant vocabulary.
Our variables could possibly be:
- [Benefit Bridge] – “so clients get extra soup with every chew.”
- [Target Audience] – “anybody who likes soups and doesn’t wish to take tiny bites.”
- [Brand Voice] – “informal and enjoyable.”
I’ve built-in three steps from the unique course of: “check your phrases,” “spell and grammar examine,” and “add a call-to-action.”
References
Subsequent, we are able to present the AI with examples of profitable, category-specific product descriptions.
The product description could also be related in type to those examples:
“[First Example]”
“[Optional Second Example]”
Craft three glorious product descriptions.
Filling within the instance variables, we’ve got a accomplished immediate.
Think about you might be an skilled copywriter engaged on product descriptions.
Compose a compelling product description for a soup spoon, specializing in the big bowl or head that enables it to carry soup. Describe how the big bowl advantages clients — comparable to “get extra soup with every chew” — whereas offering examples and comparisons.
Embody a direct name to motion.
Be aware {that a} soup spoon is for anybody who likes soups and doesn’t wish to take tiny bites. Whenever you write, use an informal and enjoyable tone. Guarantee the outline is partaking, grammatically appropriate, and avoids repetition. Use a various, related, and significant vocabulary.
The product description could also be related in type to this instance:
Hungry for some hearty rooster noodle or creamy clam chowder? This soup spoon has a big bowl meant to haul bisques and broths to your mouth. In truth, this soup spoon can maintain about thrice as a lot soupy goodness as your commonplace desk spoon. You possibly can take thrice as many bites, or you would purchase this soup spoon and slurp giant.
Craft three glorious product descriptions.
Consider
Even the most effective AI prompts won’t all the time produce the specified outcomes, so consider the output. The instance immediate generated this response from ChatGPT 4o.
Uninterested in chasing soup round your bowl with a tiny spoon? Meet the soup spoon designed for actual soup lovers. Its beneficiant, deep bowl enables you to scoop up wealthy, velvety bisques, hearty stews, and soul-warming broths in satisfying mouthfuls — no extra dainty sips. In comparison with your common tablespoon, this unhealthy boy holds practically thrice as a lot, so you’ll be able to get pleasure from extra taste with each chew. Whether or not you’re slurping ramen or savoring home made rooster noodles, this spoon will get the job performed. Go huge or go hungry — seize yours right this moment.
Iterate
The ultimate step in Google’s immediate engineering framework is to iterate or repeat the method with minor adjustments till you get the appropriate response.
Product descriptions have by no means been extra essential for ecommerce success. How we produce and check them displays their significance.