I use AI for email marketing. It creates personalized emails & videos for abandoned carts! The app I use is CartSaver. Highly recommend using AI to increase your email revenue.
definitely. thank you for your insight
I’m focusing on expanding the video element beyond abandoned carts.
Videos have great potential to drive results for many different stores. Will also focus on the timing and capturing intent effectively.
Sure, that’d be nice. Where can we chat?
We use tools like ChatGPT and Notion AI to draft and localize content faster, especially for platform updates like TikTok Shop or Temu changes, and then refine it manually. We also use AI for basic data analysis and idea validation, which saves time but still keeps humans in control of the final messaging. Or software like 4Seller—I’m not sure if that counts as AI. There are some software programs that can predict which products you’re out of stock on and create restocking plans. You can find many of them on the Shopify app store, and they’re all quite intelligent.
AI tools are being used to enhance marketing not replace it by refreshing product content, generating visuals, and supporting customer interactions. The best results come from pairing AI’s speed and scale with human creativity and strategy.
Brands use Pikes AI to create bulk product visuals, automatically placing text, layouts, and design elements to produce ready-to-use, on-brand creatives without manual design. Lightroom helps build consistent moodboards and visual identity, while Gemini supports research and prompt development to steer strategy in the right direction. Ahrefs uncovers search demand, optimizes content, and guides SEO using AI to accelerate execution while maintaining creative and strategic control.
At The Genie Lab, we use AI to help Shopify merchants create better content faster. Our AI ReWrite App can refresh product descriptions, blog posts, and pages—improving clarity, SEO, and engagement—all directly from Shopify. Our partners really like to use the app.
hey there..
great question. the noise level with “ai marketing” is incredibly high right now, so cutting through the hype is half the battle.
internally, and for the merchants we advise, we try to split ai tools into two buckets: creation (making stuff) and agents (doing stuff). most people get stuck in the first bucket, but the real roi is shifting to the second.
here is the stack i usually recommend to entrepreneurs to keep it lean but powerful:
-
the “brain” & strategy: claude or chatgpt (specifically the ‘o1’ or ‘4o’ models). don’t just ask them to “write a blog post”. feed them your raw customer support transcripts or reviews and ask them to extract “pain points” and “hooks” for your ads. it’s way better at strategy than creative writing.
-
visuals: midjourney is still king for high-fidelity creative concepts, though canva’s built-in ai tools are getting surprisingly good for quick social edits.
-
segmentation (the technical part): if you are on shopify, tools like klaviyo are adding predictive ai that tells you when a customer is likely to buy next. this is crucial because sending a mediocre email at the perfect time beats a perfect email at the wrong time.
for bitbybit, we focus heavily on that second bucket: agentic ai.
instead of just using ai to write a marketing message, we use ai agents to handle the traffic that comes back. for example, if we blast a whatsapp campaign, we have an ai agent sitting there ready to answer “do you have this in size m?” or “how much is shipping?” instantly. we found that marketing falls flat if the immediate response isn’t there, so we automated the action (checking inventory, creating draft orders) rather than just the copy.
my advice? start with a solid “brain” tool like claude for strategy, and don’t overcomplicate the rest until you have traffic bottlenecks.
full disclosure: i’m part of the team at bitbybit.. we build those agentic tools for whatsapp, so i’m definitely biased towards ai that does work rather than just writes work..
hope that gives you some ideas..
Hey Jacqui — love seeing the mix of answers here. 
A few simple ways I’ve seen AI help marketing without losing your brand voice:
-
Refresh product pages: write the first draft yourself, then use AI to tighten clarity + structure (great for outdated listings).
-
Creative variations: generate a few image/background options and short video scripts, then pick the best and polish.
-
Ads + email testing: ask for 5 angles + 10 headlines/subject lines, then you choose and test (AI is great for volume).
-
Customer insights: summarize reviews/support questions to spot the top objections and turn them into FAQ blocks + ad hooks.
Curious — are you focusing more on organic, ads, or email right now?
we are using https://asyntai.com/ai-chatbot-for-shopify/ - can recommend
@Jacqui Most replies here are “I use [tool] for [thing]” which is where I started too. But the gap between using AI and getting good output from AI is massive. Here’s the specific workflow that made the difference for me.
The one change that 10x’d my AI output quality:
I stopped prompting from scratch every time and started giving AI my brand context upfront.
Here’s what I mean. Most of us do this:
“Write a product description for a magnesium supplement”
And get this:
“Unlock your wellness potential with our premium magnesium supplement. Crafted with the highest quality ingredients, this powerful formula supports muscle recovery, sleep quality, and overall well-being. Take control of your health journey today.”
That could be any supplement brand on earth. It’s filler.
Now here’s the same request, but I paste a short brand context doc first - who my customer is, my brand voice, my positioning, my specific product claims:
“You already know you need magnesium — 75% of adults don’t get enough. The difference is the form. We use magnesium bisglycinate, not oxide (which your body barely absorbs). 300mg per capsule. No fillers, no ‘proprietary blend’ mystery. Take two before bed. Most customers notice the sleep difference within a week.”
Same AI. Same model. Completely different output. The second version sounds like a brand that knows its product and its customer. The first sounds like a robot filling space.
My actual workflow (for product pages, email, ads):
-
I built a simple brand doc - about 2 pages. Covers: voice (how we sound), customer (who they are, what they care about), positioning (how we’re different), and product claims (the specific stuff we can back up).
-
Every AI prompt starts with: “Here’s the brand context: [paste doc]. Now write [the thing].”
-
First draft comes back 80% right instead of 20% right. I spend 5 minutes polishing instead of 30 minutes rewriting.
-
For email specifically, I add a line about the goal of the email and where the customer is in the journey (just browsed, abandoned cart, bought once, etc). That context alone changes the output from generic blast to targeted message.
The result:
Product descriptions that actually sound on brand - not like every other store that typed “write me a description for x.” Email open rates went up because the subject lines stopped sounding like templates. Ad copy that mentions specific product details instead of vague benefit claims.
The tool doesn’t matter as much as what you feed it. I’ve gotten great output from ChatGPT, Claude, and even Shopify Magic once the brand context is dialed in.
If anyone wants to go deeper on building the brand context doc, I wrote up the full framework on my site DTCskills. But honestly, even a half-page of “here’s who we are, here’s who we sell to, here’s how we sound” will dramatically change your output quality.
I use sceneo.io for product images and videos, and gemini for description or ad script etc.
Great tips, JBallard! I’ve found that adding brand context up front really helps AI produce more on-brand content. It’s a game-changer for consistency and saves a ton of time in revisions!
I’m learning OpenClaw to automate my workflow and things like that. For example, it’s really great that it can help me get keyword trends in the competitor industry, and it can be automated. Although there are many keyword research tools on the market, I think none of them are as good as being able to customize my own. Or there are some ready-made software that you can find in the Shopify app store, such as 4Seller, which are more specific tools for automating order and inventory workflows.
This is a really interesting thread! I’ve been following the discussion around AI for content, visuals, and voice support, but I didn’t anyone mention AI for sales assistance. I think having an in-store like shopping assistant who could sound board with you would be such a value add. Has anyone here explored anything like that before?
I use it for summarizing long threads or articles. It helps me stay updated without spending hours reading every single detail.
Gemini and ChatGPT. Some other tools that claim to optimize listing performance, such as Pipe17 and 4Seller, actually connect to their large models. It is better to write prompts by yourself, as you can also DIY some settings.
We built https://ask-ai-data-connector.co.uk to improve AI understand the business as a whole. It connects your Shopify store (plus Klaviyo, GA4, Google Ads, Meta Ads, Gorgias, Triple Whale, Xero, and 16+ other tools) directly to Claude, ChatGPT, Gemini, or Perplexity.
Once connected, you can ask things like:
-
“What were my top 10 products by revenue this month?”
-
“Which Klaviyo campaigns drove the most repeat purchases?”
-
“Compare my Meta Ads ROAS to Google Ads — include refunds”
-
“Which products generate the most support tickets relative to sales?”
The key difference is it uses your real data not generic AI advice. It cross-references across sources too, so you can ask questions that span
Shopify + Klaviyo + GA4 in a single query. Read-only access, everything encrypted, works with all the major AI tools. Free 3-day trial if anyone wants to try it.
Product photos used to eat up so much of my time — setting up flat lays, getting the lighting right, editing backgrounds. I started experimenting with AI tools a few months ago and it’s been a game changer honestly. Right now I’m using Snappyit that generates flat lays and ghost mannequin shots from a basic product photo. Attached an example — that flat lay was generated from a pretty rough original. Saves me hours every week and the results convert better than my old DIY shots. Would love to hear what others are using too.
i’ve shifted from using ai for just “writing” to using it for the heavy lifting, like building saas betas in a weekend. it’s less about the bot doing the work and more about it giving me the freedom to actually focus on the big-picture growth.