🗞️ AI Magic for E-Commerce: From Custom GPT to Canva Creativity

Welcome to the 5th edition of our new AI for E-Commerce newsletter, your trusted resource for actionable AI strategies in eCommerce, especially on Amazon!

In each issue, we'll spotlight practical AI use cases, showcase our favorite cool tools that you can start using today to improve your eCommerce business. Our focus is on tangible benefits. No technical jargon, no pie-in-the-sky concepts—just straightforward, actionable advice to help you stay competitive and grow your online business. 

Let's go...

For this first section on AI Practical Use Cases, we’ll talk about a new Custom GPT by Canva, that can considerably shorten your creative cycle.

We already know that ChatGPT can generate fantastic images with some descriptive text. Where it gets challenging though, is when you want to edit certain parts of those images or composition and can’t.

Enter Canva GPT!

This custom GPT allows you to describe a composition in GPT. It produces 2 sample options that you can click on and edit directly in Canva.

Here’s an example of a prompt for the Instagram channel of a Dog toys brand:

“Make an Instagram post showcasing two dogs running towards the beach. Room for captions”.

Canva GPT generates two composition options, both clickable:

These open directly in the Canva app, and then I can edit them freely as needed - the colors, the text, the font, even replace the image!

This saves me so many steps:

• Thinking of a layout

• Planning suitable colors

• Picking a font

• Selecting multiple text lines

• Picking an appropriate image

• Putting it all together

With the Canva GPT, I get a pretty good starting point and then I can make my design even better. Imagine being able to do this for your social media at scale. Your VA can save so much time!

Try it out! It’s FREE!

For the Cool Tools section today, we are talking about TinyWow, a collection of 200+ tools that help you perform super useful mini-tasks. Today I will focus on these two:

1. URL to PDF

2. Extract images from PDF

Let’s say you wanted to extract the images of all of your competitors on Amazon’s search results page 1 for a particular keyword.

You would first want to download the search results page using the URL to PDF tool. Simply perform the search (or ask your VA to) as normal. Then feed the generated URL, which can look something like this (https://www.amazon.com/s?k=candles&crid=H6N8EYM5EGYA&sprefix=b003v4tv8o%2Caps%2C152&ref=nb_sb_noss_2) into the TinyWow’s URL to PDF generator. In a minute, you will be able to download the result as a PDF.

Next, use TinyWow’s Extract Images from PDF tool. It will give you a zip file of all the images. You are now free to browse through these, select some for inspiration, or go on to performing additional processes.

How much you ask? It’s FREE! Check it out.

For this week’s Nerd Bytes section, I’d like to explain what RAG is. The term Retrieval Augmented Generation (RAG) gets thrown around quite a bit, but what does it actually mean and how is different from regular GPT?

GPT is a pre-trained model that has been fed billions of articles, books, and more. However, sometimes the knowledge of GPTs may be too generic for your needs. Sometimes, you need answers that are more specific to a particular knowledge base, such as your company’s internal data.

When you create a “Custom GPT”, you are allowed to upload your own documents for training purposes. This layer of documents acts as a “filter” above regular GPT and is essentially using RAG.

Your prompts will first go through this retrieval filter, and then through the general purpose knowledge contained in ChatGPT, to give you responses that are more catered to your special use cases.

This is great for building internal training tools. But next time, we will cover vector embeddings, that are even better. Stay tuned!

For the Thought leader section this week we asked our friend, Izabella Ritz from Ritz Momentum to share her thoughts on the state of AI today and this is what she had to say:

“Many experts are discussing Rufus and Cosmo. To make it simpler to understand, think of Rufus as the front-end of your listing and Cosmo as the backend. Cosmo ensures your listing is shown to potential buyers, while Rufus scrapes all possible information from your listing to provide buyers with valuable insights about the product they want to purchase.

As long as you communicate properly on your listing about the benefits, features, and experience your client will get, Rufus will help you sell. Cosmo will show your product to the audience based on their interests and recent searches. COSMO generates commonsense knowledge that aligns with user needs. Using refined large language models and incorporating human feedback, COSMO has greatly enhanced search relevance, session-based recommendations, and search navigation on Amazon. This leads to better user experiences and significant revenue improvements.

TIP: When creating your listing, add thoughtful information based on customer insights you can scrape using VOC.AI. Cover all FAQs based on your competitors' listings with products similar to yours. By including all the information that will help your client make the right choice, your listing will convert visitors into revenue for your bank account.”

Izabella runs Ecom Product Finders: The World's Most Useful Newsletter for Finding Products to Sell on Amazon (10k+ subscribers). Check it out here.

We hope you liked this edition of the AI for E-Commerce Newsletter! Hit reply and let us know what you think! Thank you for being a subscriber!