🗞️ Make AI do the hard work: Coding for non-coders 💥💪

Welcome to the 7th edition of the AI for E-Commerce newsletter, your trusted resource for actionable AI strategies in eCommerce, especially on Amazon! We have now crossed 1000 subscribers💥!

In each issue, we spotlight:

• Practical AI Use Cases

• Cool tools that you can start using today to improve your eCommerce business

• Nerd Bytes for all you nerds out there

• Voices from AI Thought Leaders in the E-Commerce space

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 creating a simple automation using AI and Google Sheets. I want to create a simple file concatenator app that has the following workflow:

1) Drop a bunch of csv files into a Google Drive folder

2) Run a script to stitches these all together, retaining only 1 common header row at the top

You can do this easily with ChatGPT and Google AppScript.

You probably did not know that LLMs are great at translating words → Code. I can vouch for it, having created dozens of micro apps now using this method. You don’t have to be a coder to generate code. Simply describe what you want your AppScript to do:

Then copy the code it generates into memory.

Next, go to a Google Drive folder where you will be dropping your files and also running the code. Select More →Google AppScript as shown.

Dump a copy of the code that ChatGPT generated into the AppScript area. Save your project and test it out by clicking Run. If you get errors, no problem, throw those back to ChatGPT and ask it to fix. You will be surprised at how good LLMs are at this!

As you can see below, I uploaded a bunch of monthly Amazon reports and it correctly concatenated the files into a sheet with the date in the file name.

Finally, an update! In last week’s newsletter I had outlined a process for cleaning up your bullet points to be TOS compliant by August 15th. I got my CustomGPT working thanks to a key input from Brett Bohannon on my LinkedIn post about using AMZ instead of Amazon to bypass a trademark related block from OpenAI. So here is my CustomGPT. Feel free to use and share!

For the Cool Tools section today, we are talking about FluxPromptAI (also referred to as Enhanced AI), designed by our friend and Ex-Amazonian Brad Moss. This is not just a tool, it is a whole AI workflow builder💥!

Here’s how it works:

1. Design a workflow you want

2. Pick from the building blocks (Text/Image/Audio etc.):

3. Select multiple blocks in your sequence and connect them up using a simple click and drag between handles:

   

4. When you are all done, click on Run Prompt and it will run the entire workflow for you! 🔥 How cool is that?!

The use case are unlimited! We are currently building and testing these workflows for PPC Ninja. Stay tuned for our results in one of our upcoming newsletters.

For this week’s Nerd Bytes section, I’d like to touch on the top 3 AI models out there today (August 2024):

1. GPT-4 (OpenAI)

• Description: GPT-4 and all it’s flavors (including 4o and 4o mini) is a powerful AI that understands and creates text. It's used for making chatbots, writing content, and even helping with coding.

• Applications: Chatbots, summarizing text, translating languages, and assisting with programming.

• Strengths: It creates high-quality text, can be customized for specific tasks, and works with both text and images.

2. Gemini 1.5 (Google DeepMind)

• Description: Gemini 1.5 is an advanced AI model from Google. It's built for processing language and can handle many tasks efficiently.

• Applications: Online search, translating languages, chatbots, and other AI tools in Google products.

• Strengths: It understands language very well, works smoothly with Google services, and keeps getting better with updates.

3. Claude (Anthropic)

• Description: Claude is an AI model focused on being safe and reliable. It’s designed to perform well in language tasks while being easy to understand.

• Applications: Safe chatbots, ethical AI tasks, and research on making AI safe.

• Strengths: It’s built to be safe, works well with language, and focuses on ethical use.

These models are at the forefront of AI research and application, each bringing unique strengths and capabilities to the table. I tend to use OpenAI’s GPT4 the most, but I am totally keeping my eyes on parallel developments in open source AIs, for example, the Llama model by Facebook.

For the Thought leader section this week we asked our friend Brian R Johnson from DeepM Marketplace Intelligence about the future of AI in E-Commerce and this is what he had to say:

STICKING THE LANDING

TLDR: To keep high search positions, focus on the search rank factors considered by A9 to be “essential” to your specific product niche.

In the Amazon marketplace, achieving and maintaining a high search rank is essential for long-term organic search sales success.  Products often experience a temporary boost in search rank due to increased sales velocity during promotions or events - such as Prime Days. A common observation is that once the event ends, the product fails to “stick the landing” (to use an Olympic gymnastics metaphor), meaning its rank often falls down.  This reveals weaknesses in the listing's key search rank factors.

If your product’s rank declined after Prime Day (or any temporary promotion), the good news is that it signals that the listing has the potential to rank higher in search.  The bad news is that it also means the product is dependent on that increased sales velocity to hold its elevated position. But, why?  Sales velocity is 1 example of over 30 Search Rank Factors (SRF) used by the algorithm to sort products in Amazon search results.   

The advancement of machine learning and artificial intelligence in the past few years led to our discovery of Amazon’s Search Rank Factors (SRF).  SRFs include elements such as relative conversion rate, unit sales velocity, inventory location, keyword relevance, and category alignment — to name a handful of highly impactful factors considered by A9.  Not holding the search position when only one of those thirty search rank factors fails to outperform competition underscores the need for optimization of the remaining search rank factors — starting with your product listings and promotional methods to ensure they meet both shopper expectations and the ranking criteria unique to their product niche.  Which “key” search rank factors matter to A9 varies from subcategory to subcategory — meaning — the popular elements that work well to rank a Vegetable Steamer may do nothing to help a Heart Rate Monitor rank well.

To correct this rank volatility, either
1. Focus each listing's content and PPC advertising on the search terms that produce organic search sales at a conversion rate above that of most competitors [see Ritu’s article on how to read an SQP report].
2. Leverage advanced competitive intelligence analytics (#deepm) to identify the combination of search rank factors that are both highly impactful to your product subcategory AND that you have control over (such as those factors listed above).  

Reach out to me on LinkedIn [https://www.linkedin.com/in/brianrjohnson-amazon/]”