Before voice memos and Notion docs, people used to take dictation at 200+ words per minute, with symbols no one else could read. It was called ‘shorthand’. And it was the original productivity flex.

One of the best things I’ve done for speed and efficiency is to train ChatGPT to understand my shorthand.

I’ve taught it the meaning behind the acronyms, the abbreviations, the mental shortcuts that live inside my head but never make it into documentation. And now, when I use them, it doesn’t pause or ask. It just acts.

Take "BMM" for example. That's short for Broad Match Modifier, a way of structuring broad match keywords so that Amazon ads match accurately with the right search terms and not to their synonyms.

By now I've had enough conversations with ChatGPT that when I say "Convert to BMM," it doesn't just throw + signs everywhere. It applies the logic I trained it on: each important word gets marked with a + in front of it (so Amazon can match it exactly with search terms), those words can appear in any order, and duplicates get collapsed intelligently.

So it knows that +nike +running +shoes and +running +shoes +nike mean the same thing, so no need for both.

That's not prompting. That's training.

Here's a messy keyword list. The kind you'd normally spend 15 minutes cleaning, collapsing, and de-duping by hand. Instead, I fed it to ChatGPT using nothing but the word "BMM.

Because I trained it on what my version of “BMM” means, it knew exactly what to do.

You’re collapsing intelligently, not compressing blindly. And ChatGPT handles it like an assistant who already knows how you think.

This only works because ChatGPT didn’t learn “BMM” from a single prompt.
It learned it from many conversations—watching how I collapse keywords, what I preserve, what I cut, and what I never let it touch. Over time, it picked up my logic like a teammate would.

And at the right moments, I used these magic words: “Remember this”

That’s when the shorthand became second nature.

So now when I say “Convert to BMM,” it doesn’t ask questions.
It executes—my way.

Why It Matters

I don’t waste 10 minutes re-explaining keyword logic every time.
I don’t need spreadsheets for copy-paste-cleanup.
I just say “Convert to BMM,” and it gets it—my way.

This is the future of PPC workflows:
You → Teach AI once
AI → Handles it like a trained assistant, forever

The right shorthand can feel like magic—just a few words, and your AI gets it. What will yours look like?

Did you miss our AI for Automation Workshop this month? You can still purchase the replay here!

We’ve been exploring OpenArt lately—a creative platform that lets you generate images, edit just parts of them, and even turn them into short videos with consistent characters.

It solves a problem that’s quietly frustrating with most AI tools: character consistency.

We recently used OpenArt to create this short beauty clip (below). No camera crew, no actors—just a few well-placed prompts and some light editing. What stood out wasn’t the generation itself (which many tools can do), but how reusable and editable everything was.

The model was generated using OpenArt with Flux and the video was generated with VEO2. OpenArt is an AI creator platform that includes Veo 2 as well as other models such as Flux, Kling 2.1 and more.

Here's what worked well:
🔹 You can change just one part of the image without starting over.
🔹 It keeps the same face across scenes—even in video.
🔹 There’s a built-in “Consistent Character” mode that’s surprisingly reliable.

If you’ve ever tried building a content series, ad variations, or a UGC-style video and watched your AI character slowly morph into someone else—you’ll get the value here right away.

💰 Pricing:
Free plan with 100 credits/month
Paid plans start at $10/month for more volume and advanced tools

This one’s worth bookmarking if you do any kind of visual storytelling where consistency matters. 👉 Try OpenArt

PPC Ninja is helping brands future proof their listings for AI, helping you build stunning images and videos with AI. Reach out to [email protected] to explore how we can scale your content production across Social media, Amazon ads, Amazon Posts efficiently and affordably.

Search used to mean typing in a few words, scanning SEO-packed pages, dodging ads, and hoping something useful showed up.

Perplexity takes a different approach.

It gives you a clear, direct answer, cites its sources, and keeps the experience focused. No clutter. No guessing.

It’s fast, relevant, and built for how people actually want to search now - conversationally.

What sets it apart is the blend of real-time web data with large language models. You ask something, it goes out, gathers the best information, and responds with a clean summary and clickable references such as images, videos and more. It’s thoughtful and transparent by design.

Perplexity is model agnostic. You can choose which LLM model powers your answer: GPT-4, Claude, Mistral, or others, depending on what you need.

And it keeps getting better.

You can now choose where your answers come from—whether that’s the open web, academic papers, SEC filings, or social conversations. It’s all in one place and just a toggle away.

They’ve also added voice and conversational search that is pretty good.

Last week, Perplexity partnered with Nvidia to bring localized models to Europe. That means answers in 24 languages, tuned to local context and running on region-specific infrastructure. It’s a meaningful step toward search that adapts to the way people actually speak, think, and live—wherever they are.

Perplexity isn’t just catching up. It’s quietly redefining what a search engine can be.

We know it's been tough for Sellers recently, between ever-rising Amazon fees and tariffs. That's why my good friend Max Sinclair at Ecomtent (www.ecomtent.ai) and his team have decided to give back to our community with their new FREE plan.

“Ecomtent Freemium provides you with our market-leading analytics dashboard - including metrics on how optimized your catalog is for Amazon's A9 and COSMO algorithms. It is the most comprehensive audit on the visibility of your products available on the market.”

They also provide you with the "lowest hanging fruit", the SKUs that could most benefit from content optimization to drive incremental sales based on their extensive experience and data. Their paid plans only come into play should you want to action this data through their tool - i.e. generate visual or written content. However, you can of course opt to just take action on these insights via your existing processes, and reap the benefits without paying a penny. Check it out www.ecomtent.ai

Excited to be co-hosting Week 4 of Seller Sessions each month with Danny McMillan! We’ll be diving into all things AI—what’s real, what’s useful, and what’s next. New episodes drop on the fourth Tuesday of every month!

When: June 26, Where: Irvine, CA
Tickets here

Want to follow me on TikTok for more updates?

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! Know anyone who might be interested to receive this newsletter? Ask them to subscribe here: www.ppc-ninja.com/subscribe. They will thank you for it 💥💪!!

~Ritu