Solved: Keyword Cannibalisation with Exact & Phrase Match?

Hi there,

Happy to give you some of my initial thoughts and guidance on your question about keyword cannibalisation. It’s a really common issue and something that trips a lot of people up, so it's good you're thinking about it. You asked if running exact and phrase match keywords in the same ad group would cause problems, and the short answer is, yeah, it probably will. It gets a bit messy.

Let me walk you through my thinking on this, how I'd approach fixing it, and how it fits into the bigger picture of running a successful Google Ads account. This is the sort of stuff we go over all the time, so I've seen what works and what doesnt.

I'd say you should focus on account structure first...

Right, so the core of the problem is control. When you chuck both phrase and exact match keywords into the same ad group, you’re basically telling Google, "Here, you figure it out". And while Google's smart bidding, like your Max Conversions strategy, is pretty powerful, it's not magic. It works best when you give it a clear, logical structure to work with.

In theory, Google is supposed to prioritise the most restrictive keyword that matches the search query. So if someone searches for exactly your keyword, the exact match version [keyword] should be the one that triggers the ad. But in practise, it’s not always that clean. The algorithm might see that your phrase match keyword has a better ad rank in that specific auction, or a longer performance history, and decide to show that one instead. I've seen it get muddled up countless times. This leads to what you're calling keyword cannabalisation, where your own keywords are essentially competing against each other.

The bigger issue is that you can’t properly analyse what’s working. You can't see which match type is actually driving the best results because all the data is lumped together in one ad group. You also can't write super-specific ad copy. An ad that works perfectly for an exact search for "emergency electrician london" is different from an ad that needs to cover a phrase match search like "find an electrician in london". The first is urgent and specific, the second is more general. Lumping them together means your ad copy will always be a bit of a compromise, which hurts your click-through rate (CTR) and Quality Score.

So, the first thing I would do is seperate them. I'd create a much more granular structure. It takes a bit more effort to set up, but it pays off massively in the long run. You get more control, better data, and ultimately, better results. Think of it like this:

With this structure, you can write ads in the "Exact" ad group that are laser-focused on that one specific search. In the "Phrase" ad group, your ads can be a bit broader to match the wider range of searches it will trigger. This improved ad relevance almost always leads to a higher Quality Score, which means Google rewards you with lower costs per click (CPC). It’s a fundamental step in optimising any search campaign.

We'll need to look at negative keywords...

Now, just separating the ad groups isn't enough. If you do only that, you'll still have the same overlap problem. The key to making this structure work is using negative keywords strategically. It's how you build walls between your ad groups and force Google to send the right traffic to the right place.

It's quite simple, really. In your Phrase Match ad group, you need to add the exact match version of your keyword as a negative keyword. So, in the "Service - Phrase" ad group from the example above, you would add [emergency electrician london] as a negative keyword.

What does this do? It tells Google: "For this ad group, show my ads for any search that matches the phrase 'emergency electrician london', unless the search is exactly 'emergency electrician london'". This clever little trick ensures that the exact search query can only ever trigger your ad from the Exact Match ad group. This gives you back complete control and stops the cannibalisation dead in its tracks. All the data for that precise, high-intent search term will now collect neatly in one place, and you'll know for sure which match type is perfoming best. It cleans everything up and makes your data much more reliable.

You'll need to rethink your bidding strategy a little...

You mentioned you're using Max Conversions, which is a good strategy for getting as many leads or sales as possible within your budget. It's often the right choice. However, automated bidding works on data. The better and cleaner the data you feed it, the better the results it'll give you. When you feed it a messy ad group with competing match types, its trying to optimise based on noisy, confusing signals. It might not know whether a conversion came from a brilliant exact match or a lucky phrase match, so it can't learn effectively.

By implementing the clean structure with seperate ad groups and negative keywords, you’re giving the Max Conversions algorithm much cleaner data to work with. It will now be able to see clearly, "Aha, the Exact Match ad group is converting at 20%, while the Phrase Match ad group is converting at 10%". It can then adjust its bids far more intelligently to prioritise the ad groups that are actually working. Your overall campaign efficiency should improve as a result.

Once you've run this for a while and have a good amount of conversion data (say, 50+ conversions a month), you could even think about switching to a Target CPA (tCPA) or Target ROAS (tROAS) strategy. This gives you even more control over your profitability, telling Google exactly how much you're willing to pay for a conversion or what return you expect on your ad spend. I remember one case where we helped a software client reduce their cost per user acquisition from £100 down to just £7 by refining their account structure and then layering on a tCPA strategy once we had reliable data. We achieved this by using both Meta Ads and Google Ads.

You probably should focus on the bigger picture...

Fixing your keyword structure is a great first step, but it's only one piece of the puzzle. A perfectly structured campaign won't save you if the user has a bad experience after they click your ad. I see this all the time. People spend ages obsessing over keywords and bids but forget about the actual user journey.

Think about your entire funnel. Once someone clicks your ad, where do they land? Is the landing page a perfect match for the ad they just clicked? Is the headline on the page reflecting the promise of the ad? Is there a clear, compelling call to action that tells the user exactly what to do next?

I remember working with a B2B client who was selling an accounting system. Their ad campaigns were struggling. They had their keywords set up reasonably well, but their ads led to a landing page that didn't offer a free trial, which is standard in their industry. No one was going to switch their entire accounting system without trying it first. We helped them fix the offer on the landing page—introducing a proper free trial—and wrote some much more persuasive copy. The campaign performance transformed almost overnight, not because we changed the keywords, but because we fixed the broken step in the funnel after the click.

So, as you’re tidying up your ad groups, take a hard look at your landing pages too. This kind of holistic optimisation is what seperates campaigns that just tick along from those that generate serious results.

I've detailed my main recommendations for you below:

As you can see, it's not just about one quick fix. Getting paid advertising right involves getting all these different elements working together. While the steps I've outlined are things you can certainly implement yourself, it can be tricky to get right and takes constant monitoring and adjustment to get the best performance.

This is precisely the kind of strategic work we do for our clients. We handle the structuring, optimising, and scaling so they can focus on their business. If you'd like an expert eye on your account, we offer a free, no-obligation initial consultation where we can go through your specific setup in more detail and identify the biggest opportunities for improvement. Just let me know if that's something you'd be interested in.

Hope this detailed breakdown helps you out!

Regards,
Team @ Lukas Holschuh