When to Ignore Keyword Cannibalization

I’ve noticed a trend in my own thinking lately — some things aren’t worth fixing, even if they seem like they should be. I wrote about it with 404 Error reports in GSC, and it turns out keyword cannibalization is another example.

When progress has stalled, SEOs often start looking for “issues” and one of the issues that gets people riled up is keyword cannibalization. This is because it seems like low-hanging fruit — find pages competing for the same keywords and merge them so Google doesn’t have to split the searchers’ attention — but it almost never is that simple. 

Sometimes it’s a thorny mess that you can spend tons of time trying to fix and get absolutely nowhere.

Keyword Cannibalization Examples (of the Intractable Variety)

As is often the case, because I don’t need to ask permission, I’ll use examples from my own site, WineClubReviews.net.

I started this site in 2009. SEO barely existed back then, and it was mostly ideas about link sculpting, doorway pages, and keyword spam. Fast forward to today, the website’s main focus is a tiny niche within the already small niche of wine. Now SEO, as you know, is incredibly complex.

Here are two examples of keyword cannibalization issues from my website. After years of trying to fix these issues, I gave up and focused on other efforts. If you know how to fix ‘em, I’m all ears!

Issue #1: Best Wine Clubs

Many years ago I saw one of my competitors was doing pretty well with a series of subdomains. Each subdomain had one page on it. (This was MANY years ago.) So I built a subdomain (just one to start) to house my Best Wine Clubs lists (which were also a fairly new idea at the time). I knew one page per subdomain probably wasn’t a great strategy, so I created several lists (I think there have been 23 over time).

1. Best Overall Wine Clubs
2. Best Affordable Wine Clubs
3. Best Premium Wine Clubs
4. Best Red Wine Clubs
5. Best California Wine Clubs, etc.

Remember my comment earlier about how I’m in a tiny little niche? It turns out very few people search for “best red wine clubs.” VERY few. Same for the other specific types of wine clubs. Only those first three pages have ever mattered. Well, also the Best Wine Clubs for Mother’s Day page, that one ranks.

Then the pandemic happened and every publication in America wrote an article about the best wine subscriptions. And then Google started ranking those generic “overall” pages even for the specific queries like “best red wine clubs.” At which point I started to wonder… do I need all of these different lists of best wine clubs? Google doesn’t care about them… do my readers?

So it turns out my readers care about these pages. I won’t tell you how I know, because at least one of my competitors is probably reading this, but I know they care.

So now I’m back to wondering, is it better to fix this cannibalization issue or let my readers have access to the content they want? I voted for the readers and stopped worrying about how it’s playing out at Google (mostly). I still work on ranking these pages, but I don’t fixate on the cannibalization side.

The Segueway to Issue #2

The next issue I’ll discuss is directly related to Issue #1 and there’s a great SEO tip buried all the way down here.

If you’re looking for keyword cannibalization on your site, always use Google as your data source. First the SERP and then GSC to verify what you’re seeing in the SERP is a real pattern and not an experiment.

SEO GEM RIGHT HERE: Run a query for “Best Wine Clubs” and then add this parameter to the URL: filter=0. Now you’ll see all of the pages on my website that match “best wine clubs.”

If you’re in California, the first one you’ll find is not one of my best wine clubs lists, it’s a page titled “18 Wine California Wine Subscriptions” (i.e. all of the wine clubs with active reviews that feature California-made wine).

On the first few pages (with the filter parameter set) you’ll see my actual Best Wine Clubs page, my Best Premium Wine Clubs page, and then a whole mess of pages all in a row:

The first few links are evidence of a combination of keyword cannibalization and Google misunderstanding search intent (incorrectly applying my location to the query, because if that’s what they believe I really wanted, there’s a better page on my site for that: Wine Clubs that Ship to California.

The remaining links are indicative of a larger problem — I have a big site of pages about a teeny tiny, poorly understood niche. And I can’t fix that problem because the problem is at Google.

Here’s another example.

Issue #2: California Wine Clubs

In the wine world, California is one of the biggest and well-known areas of production. With so much wine produced here, it’s common for people in California to drink (or at least search for) California wine. Much more so than other states. Google Trends search: https://trends.google.com/trends/explore?date=all&geo=US&q=%2Fm%2F03gs5gm&hl=en

When it comes to “wine club” terms relating to California there is a lot of confusion. Here are some examples which Google regularly misunderstands the search intent for:

1. Wine clubs that ship wine to California
2. Wine clubs (retail) that feature California wines
3. The best wine clubs (retail) featuring California wines
4. The California Wine Club (a wine club company which features California wines)
5. California winery clubs (wineries in California which sell their own wine DTC)

I can sort of see why Google might be confused, but it’s not that hard for a human, so why is it so hard for them? 

It is situations like this one, which I have tried myriad solutions for resolving, which make me look forward to the implementation of LLMs (Large Language Models like Bard or GPT) into their search results pages. Of course, even if they correctly understand the search query and intent, they still have to know which pages actually match that.

The LLM doesn’t get it either.

Just to break that down for you:

• The California Wine Club is a good option
• Gold Medal Wine Club also focuses on California wine
• Diamond Wine Club is by Gold Medal
• Naked wines does not focus on California wine (but does ship there)
• The Wine of the Month Club is in California (doesn’t focus on California wine) but I’m guessing it’s here because it’s quoting an LA Times article
• Firstleaf does not focus on California wine (but does ship there)
• Bright Cellars does not focus on California wine (but does ship there)
• Winc is based in California and has a lot California wine

Under Ask a follow up, the other questions should be about California wine clubs, not wine clubs in general. 

My 18 California Wine Subscriptions page ranks 3rd in the 10 blue links section. It’s not included in the Generative AI section as a source.

What’s interesting is that The California Wine Club, arguably a great result for the #1 spot (and also a client), also has a boatload of other pages that could rank:

And then after all of those pages, the next one is my Best California Wine Clubs page. Hey, at least I outrank the LA Time article (today) 🙂

Some of what have I tried or not tried

• Disambiguation (tried). On each of the pages most likely to be confusing I have a series of links (right under the H1) which suggest other pages which might be more appropriate (this is for users, but I use the right anchor text to help the bots).
• A California Wine Club silo (not tried). These pages all belong to other silos, namely: Best Wine Clubs, Wine Club category pages, Wine Delivery, and Wine Club Reviews. Those silos are all well reinforced with navigation clues.
• The one thing I think may be contributing to the problem is a Best Wine Clubs shortlist I have at the top of the page for all of the wine club category pages. The phrase “Best Wine Clubs” used to be a heading tag, but I downgraded it to a p tag a while ago.
  
  Readers like these lists, so I’m willing to live with Google’s confusion. Interestingly, readers also preferred the generic Top 5 list rather than a context-aware Top 5 list (for example, I tested showing the Top 5 Red Wine Clubs on the All Red Wine Clubs page and that test wasn’t a winning version).
  
  Again, I vote for my readers and not for Google.

So should you fix keyword cannibalization?

I’m an SEO so, it depends. 

You can see from my examples there are times when resolving cannibalization will require a lot of work and get you absolutely nowhere. This is an opportunity cost. What else could you have done with all that time?

That said, if you have cannibalization, make some effort to resolve it but be comfortable hitting the pause button without seeing success.