Escape the backlog

Michael says: “While AI is flooding the web with content, it’s also assisting teams to create huge strategies and big plans. However, we're still not seeing them getting executed.”

What does execution actually mean, and why is it not happening?

“Imagine that you’ve gone through and done an SEO audit; you've explored all the opportunities, you've gone through the competitor set, and you've gone through what you have for existing content and how you're performing.

Then, you get this beautiful list of what you should do to get things done – much like going to the physio, where you get a list of the exercises you should do when you're not at the physio. In the meantime, between visits, the work's not getting done.

There's definitely a place for using AI. It's great at analysing data. It's great for giving a first draft of a piece of content. We're leveraging it to get content briefs put together and get a first draft of content. However, a lot of the time, we're seeing things getting shelved.

Priorities start getting pulled. It's sometimes the fact that there is a whole lot that needs to get done. A lot of marketing teams are dealing with having fewer people to execute with, and they're having to rely on other sources to get things done.”

What can an agency do to get their project top-of-mind and increase the urgency on it?

“There are two approaches to this. One is getting the internal stakeholders for your client on board with it by explaining where the opportunities lie. When you come up with a list of opportunities, prioritise that list. What's going to give the biggest ROI, and what can you get implemented right now?

If they're having an issue with execution or they're being pulled in multiple directions, step up and offer to execute. Including execution in the strategy is something we've been rolling out quite frequently in the last 18 months.”

How do you determine the ROI of a content project?

“There are so many different levers for that. It's hard to say because there isn’t one solution. You can use the classic Impact-Effort matrix, with the four quadrants. If it’s low effort and high impact, then you can prioritise it right away.

Perhaps it's something that can be rolled out at scale and can be applied at a template level, like putting article schema on a handful of articles, or you can point out where they're going to get the most return.

Prioritise things based on where you can see a quick lift and implement very quickly – where it's really good practice, and you should start working on it now. The old joke is that the best time to start exercising was 10 years ago. The second best time is now.”

Why do you believe that most teams don't have a strategy problem; they have an implementation problem?

“This has been an issue that I’ve seen for the last eight years, since I've been out on my own – and particularly when I’m consulting.

When I was on the agency side, we were providing direction and then expecting teams to execute internally. We would even prioritise tickets in Asana or Jira, and you might get a first lift, where people start going through and implementing things, but then priorities shift. Their focus isn't 100% on SEO; their focus is marketing as a whole, and the queue keeps getting bigger.

If you keep providing more direction and discovering new opportunities, you're just adding to a queue that's not getting executed.”

How do you prioritise what to do first?

“Start with the things that require less effort but would provide a bigger lift. Build your foundation first.

What can you do at a template-level, to change the structure of article pages or product service pages, for example, in a way that gives you SERP features, helps search engines, and even helps LLMs digest that content in a better way?

Then, start moving on to the longer-term projects. For example, link building is something you want to start as soon as possible. You're not going to see the return for a while, but you want to get it implemented now so that you'll start to see a return before the project's out.”

What does good link building look like in 2026?

“Good link building is still using a community engagement-based approach. Is that link actually useful? Links for the sake of links are just noise. If you're starting with nothing, that can probably give an initial lift, but if you're dealing with established brands, getting a link for the sake of a link is not necessarily worth it.

When it's a link from somebody that's part of your extended community or somebody related to what you're offering, it can be a really strong play, and it can help build partnerships.

You want to use a more community-based approach, rather than just building links for the sake of links”

You say that the backlog is where good SEO goes to die, but can you automate some of those tasks with the help of AI?

“It does make sense, but you want to spot-check those automated tasks. It depends on the size of your site.

If you're dealing with an e-commerce site that has a whole pile of pages, you shouldn’t generate the titles with AI necessarily, but you can come up with a structure that's systematic and variable-based: using the product name and whatever's unique about that product, like colour, size, etc., to generate the title.

If you're starting to notice things falling off, then you can dig deeper into why that's falling off. If you're seeing something that's a pattern, you can roll it out broadly across the whole site.”

Are more clients nowadays getting audits and feeling good that they've got them, but not doing anything about it?

“I've seen it my whole career. People get that audit, that beautiful report, and it tends to sit on the shelf – or they action one or two things, and then it just gets deprioritised.

Keeping an SEO on your team on a long-term basis (either as a contractor or in-house), who's actually executing on that, can really keep pushing things forward and moving in the right direction.”

Why do you think that SEOs need to work with real-world CMS and dev constraints, not against them?

“WordPress, Webflow, Squarespace, Wix – they all have their own restrictions, and their own things that don't quite work.

I've got a smaller service-based client right now who is using Squarespace, and there are things you simply can't roll out in Squarespace. I can't copy over a CTA to multiple pages, for example, I have to do it manually on each individual page, which takes a lot of work.

Essentially, you need to accept the things that can't be done. Don’t insist that you have to change the CMS. If the dev team and the marketing team want to keep the existing CMS, you're not going to be able to do certain things, so focus on the things that you can do. Focus on what you can change.

Don't die on that hill. Switching CMSs may give a little bit of a lift, but is it worth pushing everything, especially if it's well out of the budget?

It's still worth asking why you are with that CMS. Why were these decisions made, and is it possible that they might be willing to move in the future? You can create a prioritised list, with the things that must happen, the things that would be nice to have, and the things you need to consider if there's ever a site redevelopment.

One of those things could be the restrictions of the current CMS you're working with, which other CMSs don't have those restrictions, and what needs to be part of the discussion if you decide to switch.”

What new opportunities can you reveal through execution that you won't find in a spreadsheet?

“These opportunities come up all the time, especially when we're implementing content. One thing we'll do right now with clients is we'll work with a content writer. We'll use AI for a first draft and get a content brief. Then, we'll get it in front of a content writer who specialises in that space, and they'll act more editorially. Then, we'll take over the execution and publishing of that piece.

Some of the things that won't come across when we're pulling that out are things like getting an unordered list in the table of contents at the beginning of the piece. As you’re reading through, should you just blurt out that this is a table of contents? How can you make that come across more naturally?

One thing that comes up a lot is internal linking opportunities, and variants on the keyword that you didn’t consider. You did a Screaming Frog crawl for all of those variants, but as you’re reading through, you notice other topics that you are talking about that don’t use the exact phrase that you had flagged already.

As you’re reading through that content, new topics and new ideas come up, and new ways to expand on that piece.”

Are specific keyword phrases for internal linking still important, or should you be optimizing for long tail questions with less search volume?

“I think we're definitely going that way, especially with LLMs. No one queries ChatGPT with the term ‘HR software’. They ask very specific questions about that HR software, and there will be a lot of follow-up prompts to that.

We see that type of thing happen with AI overviews as well, directly in the SERP. That's definitely the way things are going.”

What elements within content can make it more appealing to AI search engines?

“Proper heading structure and classic SEO best practices. A table of contents is helpful.

Segment the content out into a proper heading structure that cascades down, starting with an H1, down to an H2. Make sure that you don’t jump to an H3 or have random headings that don't make sense and don't cascade the way you'd expect them to.

Another thing that helps is chunking your content up and having logical pieces that fit under those headings.”

Are there other teams that SEOs need to be working with more in order to get more done?

“I'd love to see teams working with CROs. If somebody's not engaging with that piece of content, then what's the point?

Talk to your social team and your community management people. Whatever content the social team is putting out on different social channels, what kind of questions are coming back? What kind of comments are coming back? How can you refine your content to answer those types of questions and comments?

With the community engagement team, what kind of things are happening? What are they hearing on Reddit? Reddit's a great example of an unbiased source where people are asking questions about your product. If somebody's asking those questions on Reddit, there's a good chance they've asked it on a Google search as well – especially if you see them more than a couple of times.”

Michael, what's the key takeaway from the tips you shared today?

“Get into your backlog and figure out what you can get done.

It doesn’t matter how good the strategy is if it just sits there. Execution will push your SEO forward in 2026.”

Michael MacMillan is Founder and Principal at MacMillan Search. You can out more over at MacMillanSearch.com.