Create an AI assistant for your store today 

Getting your store ready for shoppers to discover via AI boils down to a few key areas: keep your product data clean and current, write content that answers the questions shoppers actually ask, and set up connections so agents can find and buy your products. 

AI can do so much more for merchants than just writing product descriptions. Point the same assistant your shoppers use at your real store data, and it can review last week’s sales, draft replies to customers, and plan your next promotion — all in your voice. Not only does this help customers discover your store, but it actually frees you up to run it.

Pick the AI tool you’d like to use: Claude, ChatGPT, or Gemini. You’re going to program it with your data so you can set it up to run tasks for you. 

We’ve created a prompt pack for you to use to set up 4 workflows. Each pack includes a README with instructions, as well as a ready-made brief and example prompts so you’re not starting from a blank page. They also have a click-by-click setup guide for Claude, ChatGPT, and Gemini, plus sample data so you can test it before using your real numbers.

For now, the connection is manual: you export the data, then you paste it into the AI chat. It’s a great starting point to see what the AI can do for you. But your developer can also set up a direct connection to your store via WooCommerce MCP if you decide to go that route.

Get the full set of workflow packs.

Start with context about your store

Open your AI of choice, create a project or start a chat, then personalize this template and paste it into the instructions field to set the context and tone for your assistant.

You are an assistant for [store name], a [vertical] store on WooCommerce. 

What we sell: [one or two sentences]

Target customer: [who buys, and why]

Brand voice: [e.g. "warm and direct, no jargon"]

When I give you data, work from it. If a number isn't in what I've given you, say so rather than estimating.

Upload 3 or 4 files alongside it that show your products, returns policy, and 2 things you’ve written that sound like you.

Then, create these workflows

Here’s the link to our ready-made packs workflow prompts. Pick one, feed it into your chosen AI, get it working, and add the next when you’re ready.

All four packs include a full brief for each workflow; setup guides for Claude, ChatGPT, and Gemini; and sample data to test that they’re working.

1. The weekly store review

Every week, feed your assistant last week’s store data. Then ask for a short, ranked list on a weekly basis that includes:

• What to restock
• What to discount
• Which reviews to answer
• What’s slipping

It can turn half an hour of scanning dashboards into a five-minute skim. This process can also surface trends or issues you might have missed.

What data to give it: The last 7 days of orders, traffic, reviews, and current stock levels.

Review my WooCommerce store's last seven days using the attached order export, traffic report, reviews, and current stock levels. Give me: (1) the 3 SKUs closest to selling out, (2) the 3 slowest movers worth discounting, (3) any unanswered 1- or 2-star review from this week, and (4) one trend I'd likely miss by skimming a dashboard. One page, ranked.

2. The content builder

While you can write your own AI-optimized product descriptions, you can also offload that task over to your new AI assistant. This workflow can draft FAQ blocks, buying guides, and category copy that makes it easy for AI to discover your store while also reading well for human shoppers. The same assistant can then check how your store shows up when it runs a shopper’s query.

What data to give it: A product export, or just paste in the details for one product.

Write an FAQ block for [product name] that answers the five questions shoppers actually ask before buying. Use plain language, and keep it under 40 words per answer, with no marketing filler. Then run the search query a shopper would type to find this product, and tell me whether my page would answer it.

3. Customer communications

With a few writing examples — like links to branded social posts, website copy, blogs, and other marketing materials — this workflow will draft replies to reviews, support questions, and post-purchase follow-ups. You approve the draft and send it. That way, you keep your brand’s unique voice while offloading the actual typing.

What data to give it: This week’s reviews and support messages, your shipping and returns policy, and 2 or 3 things you’ve written yourself so it can match your voice.

Draft a reply to each one. Match the tone to the situation, reference our actual shipping and returns policy, and keep each under 80 words. Flag any I should handle personally rather than send as drafted.

4. Campaign and promo planning

This workflow is based on your order history and product catalog, plus what you know about who’s buying your products. But that doesn’t necessarily mean you need to upload any formal personas or market research. Order history shows which products sell together, who comes back, and when your quiet months hit. Previous email campaigns reveal what people actually opened and clicked, and what they ignored. Reviews tell you, in customers’ own words, why they picked you over someone else. Hand over all that context and ask this workflow for a seasonal calendar, promo concepts, and email drafts. This is especially handy if your marketing team is stretched thin and you need a fast first draft.

What data to give it: Your last 6 to 12 months of orders and your product catalog, plus past email results if you have them.

Build a promo calendar for the next quarter. Give me: (1) three promo concepts built on products that already sell together, (2) the two quietest weeks worth filling, and (3) a subject line and opening paragraph for each campaign email.

Get the full set of workflow packs.

Try out these workflows and make it a habit to run them weekly. That way, you can get real ROI from AI right now, with the tools you have at your disposal. 

These 4 workflows are just the start of what AI can do for you. If you have Claude, you can take it further by connecting your assistant directly to your store via WooCommerce MCP. You can also try our pre-release plugin with 18 prebuilt WooCommerce workflows. Run these in Claude using simple slash commands, and they’ll use your store’s live data to check your store health, review top revenue channels, generate better product descriptions, and more.

Get started with our 4 workflow packs today.

About

Dave Lockie

Dave is a visionary at the forefront of the Web3 revolution and currently leading the charge at Automattic. A native of the UK, Dave now soaks up the sun and tranquillity in Portugal. Dave shares his Web3 expertise through his advisory role at Adnode and past collaborations with WordProof, Grant for the Web, and his founding venture, Pragmatic. He also co-chairs the BIMA Web3 council.