Newsletter: Here’s an effective AI workflow to make sure your content is never out of date

Hey everyone, Araminta here.

A lot of talk out there on AI and content is about scaling content. Specifically, about creating more of it so your brand ranks for more keywords, appears for more prompts and the surface area of your brand is as large as possible.

Although that is a good use case for AI, most people implement it badly which is why “Mount AI” keeps happening: companies scale and create a lot of content, only to get penalized by Google and lose all rankings (and therefore lose all surface area).

​Mount AI keeps happening to companies that use AI to scale way too much content, crashing their traffic and rankings

There are ways to scale AI, but it has to be done carefully and with the right tools. We’ve developed our own writing tool – Mint Writer – but we only use it for certain clients, there's always a human in the loop and we've trained it extensively.

But in this newsletter, I don't want to talk about this use case. Instead, I want to talk about a way to use AI which is less complex, a lot easier to implement, less risky and possibly, just as effective as creating new content: automatic content updates.

The problem this AI content workflow solves: stale content that loses rankings, citations and is outdated

If you publish a lot of content, it’s only normal that it gets out of date after months or years. It could be that it references something from a specific year, has old statistics or just a product feature that has been rebranded. Or it just doesn’t include that new feature because you launched it recently.

What also happens with Google and LLMs is that they like fresher, newer content. So it’s totally normal for older articles and GPT articles (what we use to improve LLM visibility) to slowly lose their rankings and citations, because it’s just out of date.

For example, if you’re looking for a list of PSPs specifically for the field services sector and it’s an article from 2023, will you really read it? Some industries are moving so quickly that just being up to date is a big trust factor.

Typically, to solve this, you’d need a person to manually check which articles have lost rankings or citations, analyze why this happened, put a plan together to fix this, and then implement the changes.

It’s very time consuming, and in some ways, more energy consuming than creating a new article because you have to think deeply about how to improve an existing article.

Turns out, this is an excellent use case for AI, and here’s how...

How the AI content workflow works

This workflow can be done with Claude Skills and n8n, but for those who want to have a workflow that is easy to share across teams and is for the non-technical, we use a tool called AirOps. It’s a content workflow tool that connects to your CMS and Claude, and allows you to deploy agents and set up workflows specifically for content.

We now have 2 clients where we’ve set this up with AirOps to do automatic updates. Here’s how it works:

1. The workflow is connected to your Search Console and tracks which articles have lost rankings. You can also add LLM prompts you want to track to see if any of your pages have lost citations
2. It identifies 3 articles that have lost rankings, diagnoses why this happened and proposes a fix
3. A second workflow then takes that fix and the article, rewrites the article and then sends a Slack notification to our workspace for me to review
4. I can review directly in the tool, chat to the AI if there are bigger things to change, and then at a click of a button, export to Webflow
5. I have this running every 2 weeks, and our rankings are now stronger than ever because the content is constantly refreshed

You can see how it works here:

What’s exciting about using AI for this?

• The tool automatically checks all the links to ensure they are not broken and out of date. For example, in my “Top Fintech Podcasts” article, there were a bunch of podcasts that are no longer running. The tool was able to find those, delete them and add new ones. It actually did research on Google to add new podcasts to the lists.
• It detects broken links, replaces them and proposes new ones
• It has a "Brand Kit" trained on your existing content which you can also edit, so that anything it rewrites maintains the tone of voice of your other articles. If you talk about certain product features, it will make sure to include that to
• You can add a lot of variations to this flow: for example, send an email instead of a Slack message, connect to Asana to send to compliance, or allow you to manually select which articles you want to optimize.
• The formatting remains. Fellow content people here will care about this. It actually keeps all the images and will re-add them when exported. No need to do thousands of clicks just so the article looks good, which means this frees up your content team to do more of the exciting stuff: sales enablement content, brand work, a report.

What’s the outcome of this workflow? Stronger rankings, stronger visibility and therefore more customers

As I explain in the video, the main outcome is that because your content is constantly updated, fresh and relevant, your rankings on Google and your visibility on LLMs are higher.

We only started doing this a few weeks ago and already the rankings for our own website are improving:

There’s nothing magical about AI here: if a person was going in and updating these articles manually, the rankings would also improve. But this is so time consuming, few content teams will actually do it.

Whereas before we would optimize maybe 1 article per month, now we can do 6 per month. Companies that have hundreds of articles can also ensure all their content is now regularly updated.

And also from a brand perspective: no longer will your content be full of product features that you no longer offer, or messaging that's out of date. You can ensure consistency across your content.

How to get started? A few ways

1. Implement this yourself. If you’re not technical, I recommend trying with AirOps. Having said that, it takes quite a lot of time to set up all the workflows yourself. If you’re technical, Claude Skills and n8n do this as well, but the challenge is it’s harder to coordinate the workflows as a team.
2. Work with a partner like Mint Studios. Specifically, for us this could be:
   1. A workshop where we show you how to do it and share the template
   2. We set it up for you, train the team and then you do it
   3. We work together on acquiring customers with content and this is part of our work together

We go into more detail about how we work with companies here: Put AI to work in your content engine​

This is just the beginning, here are some other use cases for AI and content

This is one of the first use cases that we’re experimenting with, but we’re excited to put in place many others. Some examples include:

Ensure product messaging and positioning is up to date

Whenever you launch a new feature, product or just change your messaging, dozens of pages are suddenly out of date.
​
What if with just a click of a button, all those pieces of content were consistent with your latest positioning and messaging?
​
With the right AI workflow set up, this is now possible. For example, we can detect when the phrase “Product X” is mentioned in multiple articles, and ask AI to update it to “Product Y”.

Whereas the workflow we discussed above is ongoing, this workflow would allow you to update all your content in one go when you've released a new feature, for example.

Find new high-intent keyphrases and prompts to rank for

You're already targeting certain keywords, phrases and prompts. But there are probably more you could be going after.
​
With AI, you can now scan what your content is currently ranking and appearing high on LLMs for, find gaps and automatically generate ideas with an outline of new content to create.
​
With the right AI workflow set up, you can ensure you’re always appearing for the most relevant keywords and prompts.

Repurposing content

This has always been a high need in content teams: how can you take an existing article or video, and turn it into sales enablement content, LinkedIn posts and newsletters?

With an AI workflow, you can set guardrails to ensure you’re constantly repurposing and making the most of what you already have.

Scale content carefully while maintaining quality

Scaling is of course a key use case. Everyone is trying it, but few people are doing it well. The big risks with AI slop are:
‍
- Low quality content which loses trust with readers
- Low quality content with gets you penalized by Google
​
With the right writing tool (like Mint Writer) and some solid workflows, you can ensure you're scaling content carefully while ensuring quality.

​

It’s an exciting time to be in content and it also shows there are many ways to use AI without it producing slop.

We hope this inspires you to try one of those workflows, and if you’re interested in learning more about how AirOps works and whether these workflows make sense for you, just respond to this email and we’ll get back to you!

Mint Studios Recommended Reads

• Interesting newsletter on the fact that content marketing is getting more and more technical, which is what we’re finding as well: Content Marketing Is Getting Technical. That's Not the Problem.​
• You may have seen a big increase in traffic from LLMs in the past couple of months. Profound explains why: Is Zero Click marketing dead? The branded link update​​
• If you want to learn more about AirOps you can check out their university: AirOps University​

Thanks for reading,

Araminta & the Mint team 🎉

Mint Studios is an agency that turns content into measurable pipeline for fintech and financial services. Learn more about what we do.