I don’t want AI to replace SEO strategists. I want it to remove the repetitive work.
That one idea shaped everything in this article. I run an online furniture store, Homely Furniture, and like every e-commerce owner I faced the same wall: hundreds of product pages, each one needing careful, manual on-page SEO. So I built an AI-powered SEO workflow that now optimizes our entire product catalog in a single day, without a single API key.
Here is how it works, what it does, what it deliberately does not do, and how you can build a similar workflow for your own store.
The problem: product-page SEO is a grind
Product pages are where e-commerce SEO is won or lost, and they are also the most repetitive pages to optimize. For every single product, someone has to:
- Rewrite the description so it is useful, original and matches search intent
- Pick a focus keyword and work it naturally into the title, headings and body
- Keep keyword density in a healthy range, not too thin and not stuffed
- Add internal links to related products and categories
- Add relevant external links where they help the reader
- Write an FAQ section that answers real buyer questions
- Add or improve product images, alt text and the featured image
None of this is hard. It is just slow. Multiply it by a full catalog of chairs, tables and storage pieces, and SEO turns into weeks of copy-paste work that leaves no time for actual strategy.
The tool: a Product SEO Bot that needs no API key
Most AI SEO tools ask you to paste in an OpenAI or Gemini API key and pay per request. My approach is different: the bot works through the ChatGPT and Gemini subscriptions I already pay for, directly inside Chrome. If you have an AI subscription, you already have the engine.
A small desktop control panel connects to three things in my Chrome browser: the WordPress product editor, Gemini and ChatGPT. Once all three show as connected, a run looks like this:
- Find product tabs. I open the product edit pages I want to optimize, and the bot detects them and lists each product by post ID and name.
- Start the run. The bot works through the list, several products at a time, in its own background window so I can keep using my browser.
- Write and optimize the content. Gemini and ChatGPT rewrite the product content following a prompt I control, built around a full on-page SEO checklist.
- Create images. ChatGPT generates new product images (I choose how many per product) and the bot can set one as the featured image.
- Update the product. The bot fills in the fields and clicks Update in WordPress by itself, then moves to the next product.
The prompt and settings are plain editable files, so the SEO rules stay in my hands, not hidden inside the tool. Every output is also saved to a folder, which makes it easy to review what changed.
What it optimizes on every product page
The bot follows the same checklist a human SEO specialist would work through by hand.
Area | What the workflow does |
|---|---|
Product content | Rewrites descriptions to be clear, original and aligned with buyer search intent |
Focus keyword | Sets the keyword and places it in the title, headings, opening and meta |
Keyword density | Keeps usage natural and within a healthy range, no stuffing |
Internal links | Links to related products and categories on the store |
External links | Adds relevant outside references where they help the reader |
FAQ section | Answers real buyer questions, which supports both SEO and AEO (answer engines) |
Images | Generates new product images, adds alt text and sets a featured image |
Meta data | Writes the SEO title and meta description |
The point is consistency. A tired human skips steps on product number 87. The workflow does not.
The results: 85 to 90 SEO scores, in a single day
Optimized product pages on Homely Furniture now consistently land at an 85 to 90 SEO score. More importantly, the whole catalog can be processed in one day instead of weeks of manual editing.
The best part is what that day looks like for me. I start the run, the bot works in its own window, and I am free. Some days that means planning our next category. Other days it means a few rounds of 8 Ball Pool while the bot handles the busywork.
That is the real return on automation: not just faster pages, but time given back.
What stays human
This workflow is built to support SEO thinking, not replace it. These decisions still belong to a person:
- Keyword strategy. Which terms matter for the business, and which products deserve the most attention.
- Customer understanding. What our buyers actually care about, and the language they use.
- Quality review. Checking AI output for accuracy, tone and correct product details before trusting it at scale.
- Growth direction. Which categories to expand, what content to build next, and how SEO fits the wider business.
AI is excellent at execution. Judgment is still the strategist’s job, and now the strategist has time for it.
How to build your own AI-powered SEO workflow
You do not need my exact tool to apply the same thinking. These are the steps that made it work for us.
- List the repetitive tasks. Write down every on-page step you repeat for each product. That list becomes your automation scope.
- Turn your SEO standards into a prompt. Spell out your rules for keywords, density, headings, links, FAQs and tone. A vague prompt gives vague pages.
- Use the AI tools you already have. A ChatGPT or Gemini subscription can do the writing and image work without extra API spend.
- Connect the AI to your CMS. Automate the hand-off into WordPress or WooCommerce so content lands in the right fields and gets saved.
- Test on a few products first. Run three to five pages, check the SEO score and read every word before scaling up.
- Keep settings editable. Store prompts and options in files you can change, so you can improve the rules without rebuilding the tool.
- Review, measure and refine. Track scores, rankings and traffic over time, and update the prompt as you learn what works.
A practical note: if your automation drives the ChatGPT or Gemini web apps, check their terms of use first, and keep a human review step in place.
Frequently Asked Questions (FAQs)
Is a repeatable process where AI handles routine on-page SEO tasks, such as content rewriting, keyword placement, linking and FAQs, while a person sets the strategy and reviews the results.
Not necessarily. My workflow uses existing ChatGPT and Gemini subscriptions through the browser, so there are no API keys or per-request API costs.
No. AI removes repetitive execution work. Keyword strategy, customer insight, quality review and business direction still need human judgment.
For our store, the full set of product pages can be optimized in a single day, compared with weeks of manual work.
Our optimized product pages consistently score between 85 and 90, because every page follows the same complete checklist.
No. Any WordPress or WooCommerce store with many product pages can use the same approach, from fashion to cosmetics to electronics.
Conclusion
The goal of AI in SEO is not to remove the expert. It is to remove the busywork that keeps the expert from doing expert work.
Building this workflow for Homely Furniture turned weeks of product-page editing into a one-day run, with consistent 85 to 90 scores and no API costs. The strategy is still mine. The repetition is not.
If you run an online store and want to build something similar, start small: pick five products, write down your SEO rules, and let AI handle the steps you are tired of repeating.
Have questions about building your own AI SEO workflow? Leave a comment or reach out. I am happy to share what I learned.