Struggling to keep up with what’s actually working in SEO right now? You’re not alone. The pace of change is relentless, and the tools that seemed optional two years ago are now table stakes. Recently I joined the SEO in 2026 podcast series hosted by David Bain, where we broke down the one SEO tip that matters most heading into 2026. The conversation focused on a challenge almost every SEO faces: having the knowledge to do great work, but lacking a repeatable system to do it at scale. Below, I’ve turned the key insights from that interview into a practical guide you can apply straight away.
Watch the video: SEO in 2026 — Victoria Olsina on Automating Your SEO Process
The Real SEO Problem in 2026: You Have the Knowledge, But Not the System
Most experienced SEOs know exactly what to do. They’ve run hundreds of audits, written dozens of briefs, and done keyword research so many times they could do it in their sleep. The problem is not the knowledge. The problem is the repetition.
Every time you start a new homepage optimisation or a new content brief, you’re rebuilding the process from scratch. That time adds up fast, and it’s time you’re not spending on strategy, clients, or growth.
The fix is not hiring more people. It’s building systems that handle the repetition for you.
Define the Process First, Then Automate It
This is the core idea from the SEO in 2026 interview, and it sounds simple because it is. But most people skip the first half and go straight to the automation, which is exactly why their automations break.
The sequence that works:
- Identify the SEO task you repeat most often
- Document every step of that task
- Build an AI assistant or automation around that documented process
Without step two, you have nothing to train on and nothing to automate reliably.
How to Document an SEO Process You’ve Never Written Down
Here’s the thing. Most experienced SEOs carry their process in their head. They know what to check, in what order, and why. But they’ve never written it down because they’ve never needed to.
Recording is the easiest way to surface what you already know.
Record Yourself Doing the Work
Open a screen recorder and run through your keyword research, audit, or brief process as you normally would. When you’re recording, you naturally do all the things you’d otherwise skip or forget to mention.
Modern tools like Loom, Fireflies.ai, and most video conferencing platforms now include automatic transcription. Once you have the transcript, take it into ChatGPT, Claude, or directly into Fireflies’ built-in AI and ask it to extract a step-by-step process from the recording.
You now have a documented process without writing a single line manually.
Prioritise by Repetition, Not Complexity
When deciding which SEO process to automate first, the instinct is to go for the most complex one. Resist that.
Start with what you do most often:
- On-page SEO optimisation
- SEO content briefs
- Keyword research
- Competitor analysis reports
These are high-frequency tasks where automation compounds quickly. Every time the system runs instead of you, you get that time back permanently.
Custom GPTs vs Make.com: Choosing the Right Tool for the Job
Once you have your process documented, the next question is which tool to build it in. This is where a lot of people get stuck comparing options rather than shipping something.
The answer depends on one thing: does your process require back-and-forth thinking, or is every step already decided?
When to Use a Custom GPT
Use a custom GPT when your process involves iteration, discussion, or open questions.
Example: you’re deciding whether a product should be positioned around “Bitcoin loans” or “Bitcoin-backed finance.” That requires a conversation. You need to discuss keyword volume, intent, competitor positioning, and brand fit. A custom GPT handles that kind of iterative, conversational work well.
Custom GPTs are also fast to build. Depending on complexity, you can have a trained assistant running in 15 to 60 minutes. And crucially, you can share them directly with clients. They receive a ready-to-use tool, not a set of instructions they have to configure themselves.
To train a custom GPT effectively, you need five things:
- Clear instructions — purpose, audience, tone, rules, what to avoid
- Your content — examples, sample outputs, existing templates
- Technical parameters — output format, response length, structure
- Test and refine — run realistic queries and adjust based on results
- Share — deploy directly to a client or team member
The CLEAR framework (Context, Length, Examples, Audience, Response format) is a useful structure for writing those initial instructions.
When to Use Make.com
Use Make.com when every step of your process is already defined and there are no decisions left to make mid-flow.
Example: a keyword is submitted, the system queries Surfer SEO or NeuralWriter, and a content brief is returned. No discussion. No branching. Just a sequence of defined steps.
Make.com automations take longer to build and test than custom GPTs. There are more variables, more modules, and more places for things to break. But once they’re running, they run without you entirely.
| Use Case | Custom GPT | Make.com |
|---|---|---|
| Process still being defined | Yes | No |
| Iterative, conversational workflow | Yes | No |
| End-to-end, no-decision workflow | No | Yes |
| Client-facing tool | Yes (shareable link) | Possible but complex |
| Build time | 15–60 mins | 1 hour minimum |
| Ongoing maintenance | Low | Medium |
What MCPs Add to This Stack
Model Context Protocol, or MCP, is worth understanding if you’re building custom GPTs connected to external data sources.
MCPs were developed by Anthropic, the company behind Claude. They function like a simplified API connector. Where a traditional API connection requires you to write or generate a JSON configuration, an MCP typically requires just a URL or a short script. The AI then ingests and works with live data from any platform that supports it.
For SEO, this means connecting your AI directly to tools like Keywords Everywhere, Ahrefs, or Semrush without building a full integration from scratch.
MCPs are not revolutionary if you’ve already built API-connected custom GPTs. But they lower the barrier significantly for anyone who hasn’t. Think of them as a universal adapter for your AI stack.
How to Stay on Top of What’s Actually Working
There is no single newsletter or course that keeps you current in this space. What works is consistent, daily exposure across multiple channels.
LinkedIn, X (Twitter), and YouTube are the three platforms where new tools, model releases, and SEO shifts surface first. Being part of active communities in any of those, whether Slack groups, Discord servers, or X lists, means you pick things up by osmosis rather than by actively searching.
Three years of daily consumption compounds into a very strong signal-to-noise instinct. You stop chasing every new tool and start recognising which ones are genuinely worth integrating.
Putting It Together: The 2026 SEO Automation Stack
Here is a practical starting point for anyone building this system:
| Layer | Tool | Purpose |
|---|---|---|
| Transcript extraction | Fireflies.ai | Auto-transcribe process recordings |
| Process definition | Claude or ChatGPT | Extract step-by-step process from transcript |
| AI assistant | Custom GPT | Handle iterative, conversational SEO tasks |
| End-to-end automation | Make.com | Run fully defined, no-decision workflows |
| Data connection | MCP or API | Pull live keyword or SEO data into AI |
| Keyword data | Keywords Everywhere | Real-time keyword metrics |
| Content briefs | Surfer SEO | Brief generation within automation |
Conclusion
The SEO professionals who will do best in 2026 are not the ones with the most tools. They’re the ones with the clearest processes. Define what you already know how to do. Record it. Document it. Then let AI and automation handle the repetition while you focus on the work that actually requires your judgement.
Start with one process. Get it running. Then build from there.
Here’s every single interview of the Majestic podcast series:
- SEO in 2023: The importance of using hybrid models for AI
- SEO in 2024: Build your own prompt libraries for your clients
- SEO in 2025: Improve the output from AI by creating custom GPTs for your clients
- SEO in 2026: Customise AI to automate your SEO processes
Frequently Asked Questions
What is the best way to start automating SEO tasks with AI?
Start by identifying the SEO task you repeat most often, such as keyword research or content briefs, and document every step of that process. Once you have a clear process written down, you can train a custom GPT or build a Make.com workflow around it. Automation built on a vague or undocumented process will fail quickly. The process definition is the foundation everything else depends on.
When should I use a custom GPT instead of a Make.com automation for SEO?
Use a custom GPT when your workflow involves iteration, discussion, or open-ended decisions that require back-and-forth reasoning. Use Make.com when every step of your process is fully defined and no decisions need to be made mid-flow. Custom GPTs are better for newer or evolving processes; Make.com is better for mature, end-to-end workflows where the output is always the same type of deliverable.
How do I document an SEO process I’ve never written down before?
Record yourself completing the task as you normally would using a tool like Loom or Fireflies.ai. The recording captures the steps you take unconsciously, which you would likely omit if writing from memory. Once you have a transcript of the recording, take it into an AI tool like Claude or ChatGPT and ask it to extract a numbered step-by-step process. This produces a clean, trainable document without manual effort.
What is Model Context Protocol and how does it relate to SEO tools?
Model Context Protocol (MCP) is a connector standard developed by Anthropic that allows AI models to ingest and work with live data from external platforms that have an API. In practical SEO terms, it means you can connect an AI assistant directly to tools like Keywords Everywhere or Semrush without writing a full API integration. MCPs lower the technical barrier for building data-connected AI assistants significantly.
Is a ChatGPT Plus subscription worth it for SEO professionals?
Yes, for SEO professionals who want to build and share custom assistants with clients. The ChatGPT Plus subscription at $20 per month gives access to custom GPT creation, which allows you to train a bespoke assistant on your process and share it via a direct link. Clients receive a ready-to-use tool rather than a set of instructions. For agencies delivering repeatable SEO services, this capability alone justifies the cost.
How do I keep up with AI and SEO developments without being overwhelmed?
Focus on a small number of high-signal platforms rather than trying to monitor everything. LinkedIn, X (Twitter), and YouTube surface most significant AI and SEO developments quickly. Joining two or three active communities in your niche means you encounter important news through normal usage rather than active searching. Consistency over time, daily exposure across three years, builds a strong instinct for distinguishing signal from noise.
Need help turning ideas like this into working systems?
This is the kind of AI marketing work we do with teams that want repeatable workflows, not one-off experiments.
