I was recently on Seahawk Media’s WP Legends podcast talking about GravityKit, MCPs, and how AI is changing WordPress. One thread that kept coming up was the business model question: if AI can handle support tickets and even build custom features, what are plugin customers actually paying for when they renew a license?
It’s a question I’ve been thinking about a lot, and I wanted to unpack it properly here.
For as long as WordPress plugins have been sold commercially, the deal has looked roughly the same: you buy a license, and in return you get ongoing updates and expert support. Development teams ship new features, fix bugs, and patch security holes. Support engineers answer your tickets. That’s the value exchange.
It’s a model that has worked for over a decade. But both of those pillars are now under serious pressure, and plugin companies that don’t adapt risk finding themselves selling something customers can increasingly get for free.
Support is being commoditised
This one is already happening. If you run a WordPress product company, you’ve probably noticed support volume trending down. AI is the reason.
When a customer hits an issue with a plugin, their first instinct isn’t to file a ticket anymore. It’s to paste the error into ChatGPT or Claude, along with the plugin’s documentation URL, and ask for help. And the AI is usually right. It can parse docs, cross-reference forum threads, and walk someone through a fix faster than most support queues can even assign the ticket.
At GravityKit, we have over 700 documentation articles. That’s a huge asset, but it also means an AI model can absorb that knowledge base and handle many of the straightforward, docs-based questions that used to require a ticket. As Gautam from WP Legends put it during our conversation: “Just go to ChatGPT and say, read this documentation and help me fix this.” He’s right. That’s what people are doing.
This doesn’t mean support is worthless. Complex, edge-case troubleshooting still requires human expertise. But the volume of tickets that fall into the “read the docs and follow the steps” category is shrinking fast, and that was a big chunk of the perceived value of a paid license.
Updates and custom development are next
The second pillar, ongoing updates and feature development, is harder to replace. But the gap is closing.
AI coding tools like Claude Code and Cursor are already capable of extending plugins, writing custom hooks, and integrating tools that don’t have native compatibility. An agency that previously needed a developer to write a custom integration between Gravity Forms and some niche CRM can now get Claude to do it in minutes.
And it goes further than custom work. If an AI agent can read your plugin’s codebase, understand its architecture, and ship a feature that your dev team was planning for Q3, what exactly is the annual license renewal paying for?
I’m not saying plugin development is about to be fully automated. Security patches, performance optimization, and architectural decisions still need experienced humans. But the “new features” line item in your renewal pitch is going to carry less weight when customers can prompt their way to similar outcomes.
To be clear, having experienced developers maintaining a plugin you rely on is genuinely valuable. You’re paying for security audits, compatibility testing across WordPress versions, performance work that happens behind the scenes, and the peace of mind that someone is accountable for the code running on your site. AI can extend a plugin, but it can’t take responsibility for it. The point isn’t that this work has lost its value. It’s that releasing an update every few months and calling it a day is no longer enough to justify a renewal. Customers expect more, and the companies that deliver more will win.
MCPs are not optional anymore
Here’s where things get interesting, and where I think the real opportunity lies for WordPress plugin companies.
At GravityKit, we built what I believe was the first MCP (Model Context Protocol) server for Gravity Forms: the GravityKit MCP. If you’re not familiar with MCP, think of it as a standardised way for AI agents to interact with your software. Instead of the AI guessing how your plugin works based on documentation, an MCP gives it direct, structured access to create forms, submit entries, configure feeds, and manage data programmatically.
The difference is massive. Without an MCP, an AI agent working with your WordPress site has to navigate the REST API, parse HTML, and hope the block editor content doesn’t break when it makes changes. With an MCP, it just calls the right tool with the right parameters. Clean, reliable, fast.
We also built a Block MCP that lets AI make targeted edits to specific Gutenberg blocks on a page, rather than having to grab the entire page content, make one small change, and republish everything (which almost always breaks something).
My workflow now looks like this: I go into Claude Code, have it spin up a local development site, install all the plugins I need, then use the GravityKit MCP to set up forms, entries, and feeds. My job is basically to take screenshots and drop them into the guide. The setup that used to take an hour takes minutes.
I think within 12 months, every serious WordPress plugin will have an MCP. And plugins that don’t will be at a real disadvantage, because AI agents will simply work better with the ones that do.
Where the value is shifting
So if support is commoditised and feature development is accelerating beyond what any single team can monopolise, where does the value go?
Towards SaaS hybrid models. I think we’ll see more WordPress companies keeping their plugin as the on-site component but moving core logic, analytics, or premium features into a cloud backend. The plugin becomes the connector, not the product. With open-source plugins, an AI agent has full access to the codebase. It can read, modify, and extend the code freely. With a SaaS backend, the core logic lives behind an API that the AI can’t inspect or replicate. That makes the business model more robust and creates stronger lock-in.
There’s a pricing angle here too. A SaaS layer gives you more flexibility: you can bundle AI-powered features into a single monthly or annual cost rather than asking customers to manage their own token usage, which is what happens when AI features live entirely inside a WordPress plugin. I’ve been thinking about this a lot, and I suspect the companies that figure out this hybrid model early will have a significant advantage.
Towards being the product that AI recommends. This is the GEO (Generative Engine Optimization) angle. When someone asks Claude or ChatGPT “what’s the best way to build a member directory on WordPress?”, you want your product in the answer. During the WP Legends episode, Gautam mentioned that his own team had AI recommend GravityKit for a client project. That kind of organic AI recommendation is going to be increasingly important, maybe more important than traditional SEO rankings for discovery. I’ve written about how to track this traffic and it’s already a measurable channel.
Towards expert services. At GravityKit, we have Gravity Assist, a service where customers can submit custom development requests directly to our team. As the “can AI do this?” bar keeps rising, the work that actually requires human experts becomes more specialised and more valuable. Plugin companies are uniquely positioned to offer this, because nobody knows the software better than the team that built it.
What this means for WordPress
WordPress still powers roughly 33% of the web. It’s a hugely important platform, and the innovations happening around it (WordPress 7.0’s AI connectors, MCPs, structured documentation) are genuinely exciting. But its open-source nature does present challenges in an AI-first world, where any model can read, learn from, and replicate code that’s freely available.
It’s going to be up to WordPress businesses and plugin companies to navigate that reality. “Pay us annually for updates and support” worked when those things were hard to get elsewhere. In 2026, they’re becoming table stakes. The companies that thrive will be the ones that shift their value proposition toward things AI can’t easily replicate: trust, deep integration, managed infrastructure, expert services, and being the product that AI agents know how to use.
What does your plugin’s license actually pay for? If the honest answer is “updates and support,” it might be time to add a third pillar.
I discussed these ideas recently on Seahawk Media’s WP Legends podcast (Episode 154). If you’re interested in the full conversation about MCPs, AI workflows, and where WordPress is heading, give it a listen.