Richard says: “I want to focus on content decay.

We spend significant amounts of time researching, writing (both humans and AI), building, and publishing content. Yet many treat content as a publish-and-forget mindset. We need to fundamentally change that in 2026.”

Okay, so how do you identify ineffective content?

“That's a very good question. It's kind of three stages: you have the disease, the diagnosis, and the cure.

We need to talk about the disease first of all, and try to understand why content decays in the first place.

We've worked with brands – both UK and internationally – over the last decade. They've invested significant amounts in content and their content strategies, and yet content can become old and can become out of date quite quickly, especially in sectors like travel or finance, where you're looking at changes in regulations, changes in what's happening in the global markets, changes in the economy, etc.

You need to bring your content up to date because you'll also see your competitors having more mature content strategies, and also, the web is changing significantly, especially in the last 12 months, with AI.

Search algorithms are more complex, and find it easier to understand content. So, if you don't have focus on your content, you're not going to be successful in this multi-platform search ecosystem world that we now live in.”

Are there any particular metrics that you look at in order to establish if a piece of content is worthwhile keeping or not?

“Oh, there are lots.

The first step is you need to identify all your content. Crawl your website, get GA data, get GSC data. At page level, build out a huge starting spreadsheet (depending on the size of your website) and then look at getting link data at a page level, and then identifying, for each page, the search interest and the topics that you want to target.

But then also add an additional set of data for the human aspect because, if you've got a huge website with thousands and thousands of pages, this can be quite a daunting task. So, what you also need to do is overlay that human layer because there may be content that's produced for brand or for partnerships or for PPC landing pages that, if you're just simply looking at the data, it may not have a great deal of sessions or be performing particularly well, but that might be for genuine reasons.”

I'm intrigued by this document that you produce that collates all the content on a website. Are you talking about every single page, no matter how big or small it is in terms of volume of content, and in terms of whether or not there's actually an intention to drive traffic to that page?

Is there a particular form that that document takes? Is there software that you use to do that, or is it a fairly manual process in terms of spreadsheets?

“Exactly. What you need to do is basically create a full map of your website – every single page, like you said.

You can use tools like Lumar or Screaming Frog to crawl the site initially. You can pull in your analytics data from Adobe or GA. You can use Search Console data as well to understand every single page and how it's performing at a session level.

What we're looking for in content decay is we're looking to see (ideally over 12-18 months, if not longer), are pages starting to decline in performance, before we can start to really get under the skin to understand why that page is underperforming? There could be genuine reasons why that page is underperforming.

There's a lack of search interest, for example. Topics in 2020-21 around COVID are no longer relevant anymore, for example. There could be other reasons where you need to refresh that page, bring it up to modern EEAT standards and principles. You need to adopt things like authorship onto the page to help that page perform.

First of all, you need to get all this data into position to allow you to do that quantitative analysis before you can then look at the qualitative analysis afterwards.”

I guess the big challenge is, as you've alluded to, the fact that search has changed massively over the last 24 months or so with the advent of AI search, and obviously, people finding more answers that they're looking for directly on the SERP, opposed to going to a website.

Those reasons that you highlighted there are all exceptionally relevant in terms of identifying which piece of content is likely to be decaying in terms of performance.

But how do you determine if it's decaying as a result of natural ageing of the piece of content and perceived reduced relevance, versus changes in user behaviour on the SERP because of new technology?

“I think you can address that in two parts. First of all, you're looking at changes in the SERP itself, such as AI Overviews or even Google's AI Mode, and how that's analysing and digesting the content in the first place.

If the content's been decaying, you can track that through the data you get within this keep, kill, merge, refresh-type audit, this huge content audit. But then, what you can also do is you understand the topic of each of those pages.

Then you can see if AI Overviews are starting to appear for each one of those topics that you're trying to focus on and target for those pages. So, you may start to see a decline in actual click-through and in visibility because AI Overviews are appearing. But on the flip side of that, large language models like ChatGPT, Perplexity, Gemini, are using the web and the web's content to learn from.

If we've got a web that has a large percentage of content that is out of date, these large language models are creating negative biases because they're learning from outdated content.”

Now, one phrase that you mentioned there that I'd love to go back to is a ‘keep, kill, merge, refresh audit’.

I've heard of that before, and maybe talked about it briefly before, but in terms of the context of content and whether or not a piece of content is likely to be worthwhile keeping or doing something with moving forward, how do you use that framework to make that decision?

“Again, the spreadsheet is the base of that, or the data we've got is the base of that. As I said, there are tools available, like Lumar and Screaming Frog. There are other tools that you can analyse, such as Keyword Insights and ClearScope, which will help you look at that content decay.

Ultimately, what you want to do is you want to create a content governance roadmap. You've got all this data, you've gone through, you've seen the content that's either performing really well, not performing, or there's content that, as I said, is brand or partnerships that you don't need to focus on as much.

What that gives you is a blueprint. That tells you what you're going to keep, because we don't need to do anything with it. There may be content that's out of date, there's no search interest, or it's just an old piece of content relative to a topic that we don't need anymore. That can be killed and redirected.

Then, you can merge. Sometimes we've seen examples (especially in the travel sector) where you've got two topics created by different stakeholders in the business, but they're very similar, so they're competing against each other.

Ultimately, what we're trying to do is reduce the site bloat, create a cleaner site, and create a simple list of gold, silver, and bronze tiers of content that you can build out into your content strategy. Then you can have content governance over that, so for any stakeholder who has anything to do with publishing content on your website, you know about it, and you know what its importance is.”

Would you ever advocate 404ing an old piece of content that hasn't had any traffic for a while and isn't getting any obvious rankings at all, rather than 301 redirecting it?

“It depends on the page metrics, because if, let's say, that page has got no search traffic anymore – there's no search interest – but, for example, it has attracted some links in the past, then it'd probably be worth redirecting it to a similar or relevant page that still exists on the site.

Ultimately, what we're trying to do is clean up that site, so we'd probably look more on the redirect rather than just simply 404ing that page.”

Okay, and in terms of merging pages, how will you make the decision to ensure that the pieces are quite relevant in terms of likely user preference or user desires?

“Again, you'd take a look at those page elements and page data.

Are the topics very similar? Maybe even where they are in the navigation or hierarchy on the side, looking at the performance of the page, like engagement and rhetoric, not just the sessions.

Then, using tools to bring them together; those tools like ClearScope and Keyword Insights. Some of the tools use Google's natural language processing to analyse the relevance and the semantic relevance of the content.

You're bringing them together to create one really strong page on that topic, rather than two that are competing against each other.”

What about refreshing? If a page is dropping, what elements will you try to freshen up to try and reverse that trend?

“Again, you're looking at some of the tools I just mentioned there, but also looking at the adoption of other techniques. What you've hopefully got from this huge audit process is that you've got your content on your site categorised by gold, silver, and bronze.

Let's take your top product pages. These, you may need to look at refreshing or just analysing every 2 or 3 months, but other content may be every 12 months. What you can do, looking at those pages, is look at things like, could you bring authorship onto the page? If you are bringing authorship, who are the authors, and what are their bios and expertise?

Analysing that page against Google's EEAT principles and, again, using tools and techniques like Keyword Insights and ClearScope to look at how relevant your content is for that topic against your competitors in the market.”

That's interesting that you mentioned refreshing a piece of content maybe every 12 months, if it's not that important, and perhaps more often if it is.

I was thinking that you might just focus on refreshing a piece of content based upon performance on the SERP, and if there was an obvious drop, then just dive in and do something then.

Do you think, in general, that you would refresh a piece of content if there weren't any significant changes in its performance in the SERP?

“I think it is getting yourself into that habit, and getting your business to understand the value and importance of content.

We know this from working in the financial service sector, where a number of the price comparison services we worked with would schedule content analysis on a probably six or eight-week basis on their hero pages/top category pages.

If you're looking at things like car insurance, life insurance, home insurance, etc., they would look at analysing on that regular basis and even do things like, if you've made a change, you update that page, and change the actual refresh date on the page. They would also use authorship on the page and authorship for whoever has looked at, reviewed, and refreshed the page.

Then, in these super competitive sectors where these ‘marginal gains’ approaches can make a difference, setting that routine and getting into that routine on that process is key to driving success.

But then it could be a blog page, in the travel sector. That may not need to be looked at until every 12 months, because it could be ‘Travel trends for 2023, 2024, or for 2025,' and ‘these are our top travel trends’. You just need to tweak it and update it, potentially, if you're keeping that piece of content as an evergreen piece of content.”

That's a great point, and it also leads into content strategy as well, doesn't it, because a relatively novice content producer would write a great blog post on travel trends for 2026 without actually viewing future years as part of the strategy and having an overarching vision there.

If they select a URL that doesn't incorporate the year, it makes it a little bit easier to update that particular URL and just use it as you mentioned for future years – 2027 to 2028 as well, moving forward – if there's a lot of authority that eventually passes to that link.

“A better example could possibly be, if you’re looking at the travel sector, around visas or travel restrictions – where you create a single page, but that changes on a regular basis or an annual basis, depending on what's happening in the world economy.

What we're trying to do, again, is trying to create a simple, clean, strong site that is authoritative on a topic. In those examples, that's more around the evergreen, but going back to the financial services, those are pages that you need to look at on a regular basis.

Another example we had is with the changes in algorithms and the changes in the way Google values content, actually, removing a percentage of the content from the page, and reducing that down, saw significant impacts in performance. I think that was in the life insurance sector.

More content isn't always the best. You've got to use the tools, you've got to use the analysis, you've got to look at the SERP results, you've got to look at the competitors, and take in all these different data points to come up with that real insight, because ultimately this is what's going to drive the performance.”

Have you changed the way that you suggest content be produced or tweaked a little bit over the last year or so?

For instance, I'm thinking that, with more AI search engines, perhaps the incorporation of a summary of the piece of content towards the beginning may assist those types of search engines.

Is there anything that you've done like that, that has tweaked the structure of the content to be more relevant nowadays?

“Yes, we have. Again, let me use the example of the travel sector, because this is one that's top-of-mind.

We took a lot of analysis to see whether the brand we were working with was within AI Overviews or not. A great example was around a travel guide to Costa Rica. If you were looking at the AI Overviews, the brand wasn't visible in the AI Overviews. So, what we did was we looked at the competitors, and we looked at the way that the questions were being answered in the SERPs.

What happened was, instead of having a long article that just talked about the best times to go, the SERP was answering it by saying ‘January to March’, ‘April to June’, ‘July to September’, and so on. So, we restructured the content on the page into these clear monthly segments, with what to do, and then brought it into things like simple tabulated or bullet point-type answers of what to do at the best times to visit.

That impacted the performance of the content in a positive way and started to get us into the AI Overviews. So, using that data to analyse not just your content, but how the SERP is responding, and then editing your content to try to match those different techniques came up with a positive response.”

Richard, what's the key takeaway from the tip you shared today?

“The key takeaway for me is to make sure that this content process isn't a one-off.

It isn't a publish and forget; it is a publish and manage, constantly looking at your content as an ongoing process.”

Richard George is Head of Organic Performance and Managing Partner at Wavemaker. Find out more over at WavemakerGlobal.com.