Just the phrase “keyword match type” sounds incredibly dry, right? Even so, the incredible importance of a well thought-out (and executed!) strategy for dealing with match types can’t be understated.
Here is an example of two high-volume search queries for a client account. It shows their return on ad spend when the account was taken over, and then after the matching and negative keyword structure was refined.
Nothing about these terms was changed other than refining the match type and negative keyword structure they had:
No new ad copy. No different landing page. No different or better offer on the product. This was the result from a cleaned up matching structure, and nothing more.
So, want to know how to create a great matching structure? Let’s talk about that by starting with the foundational knowledge of how search queries are matched by the search engines.
What Are Keyword Match Types?
If you haven’t been a hard-core paid search manager, you may not be well-acquainted with keyword match types. When I look at the top three biggest issues when taking over a paid search account, the match type issues are usually at the top of the list.
A match type is the way that the search engine will match up the keywords you’re bidding on to the actual search queries a user types in. There are three main types, though one kind of has a sub-type as well:
1. Broad Match
1a. Modified Broad Match
2. Exact Match
3. Phrase Match
Here’s how these function in the wilds of paid search:
Broad Match
This is the default type when you set up keywords in a system like AdWords. Let’s say you sell hooded sweatshirts for this example.
If you bid on “red hooded sweatshirts” as a Broad Match term, you’re telling the search engine it can match you for any of the words in that search term. So, it could match you just to “sweatshirt” if it wanted to.
Pros:
- You get the most impression volume, which means you’ll get lots of juicy search term data to help you uncover a lot of keywords.
- Increase in potential sales due to the increased volume (in theory).
Cons:
- The Pro becomes a con quickly when you see how many irrelevant searches you get matched to.
- Irrelevant searches are a one-two punch to your account: if people click on your ad, you pay for it but won’t get a sale. If they DON’T click on it, eventually your quality score will start to tank.
Modified Broad
The idea of modified broad is that you get the reach that broad match provides, but with more control.
In our “red hooded sweatshirts” example, a modified broad match tells the search engine, “All three of these terms need to be in the user’s query, but it’s ok if they’re not in that order and some other terms are in there.” Modified broad is formatted with a “+” symbol in front of each keyword, so it looks like +red +hooded +sweatshirts.
Pros:
- You still get matched to a wide range of searches, some of which will provide some great keywords you hadn’t thought of.
- More targeted than straight broad match
Cons:
- It’s still a form of broad match, so you will get irrelevant searches.
- Since the search query can still include other words, you really have to stay on top of what you’re showing up for.
Exact Match
This is exactly what it sounds like, and for people not well-acquainted with match types, it seems new users assume this is how you’re matched with the keywords you bid on.
When you bid on “red hooded sweatshirts,” that’s the only query you’ll show up for. It has to be those three words, in that order, and nothing else in the query. It’s formatted like this: [red hooded sweatshirts].
Pros:
- You know the traffic is pinpoint-targeted.
- No irrelevant matches to worry about.
Cons:
- The person has to type in exactly that phrase. No permutations, no having words in a different order.
- This can cut down on a lot of traffic that is still very qualified, simply because you have to try and anticipate every single way a person might search for your products.
- The prices on Exact Match tend to be higher – this is ok if you find it converts well, but if it doesn’t, it might punch your pocketbook in the gut a bit more.
Phrase Match
With Phrase, the search query has to appear exactly as you’ve typed it, but it can be part of a larger search query. So, if you bid on “red hooded sweatshirts” as Phrase Match, it can appear in something like “Adam Sandler’s red hooded sweatshirts.” Phrase match is formatted like this: “red hooded sweatshirts”.
Pros:
- Like Broad Match, you’ll find some great new context for how people search for what you offer.
- There’s sometimes some great content gems in Phrase Match, because it can reveal some “how to” needs, or searches on specific things that might differentiate you from competitors. For example, a query such as “red hooded sweatshirts that don’t fade.”
Cons:
- Much the same as Broad Match, you’ll appear for a lot of stuff you probably wish you didn’t.
- That can lead to the cost and quality score problem also noted in Broad Match.
Ok, got that?
I’m going to address this in the context of existing account setups because it has some juicy data that will show you where the issues actually lie.
Click Bleed
It’s important to remember that a search query is matched to a keyword, which is matched to an ad group, which determines the ad copy that’s seen:
Search term –> Keyword you’re bidding on –> Ad Group location –> Ad populated
I started using the term “click bleed” to describe what happens when keyword match types aren’t structured succinctly with folks and it stuck.
Click Bleed: (adj.) The act of your broad match keywords being matched to multiple ad groups, likely not the sole one you intended it to be tied to. When click bleed occurs, you are matched to multiple ad groups, some of which aren’t relevant, which means less-than-optimal ad copy is shown.
Below is a visual of what click bleed looks like. This is the same client I showed at the beginning in regards to their ROAS, prior to when it was revamped and cleaned up.
This client offers a certain type of certification, along with training and courses related to it. Obviously, the ad copy messaging for someone seeking out training vs. certification isn’t going to be the same.
The green area shows the actual search term. Listed below it are the ad groups in which this search term was matched, along with the corresponding spend:
Notice how each term was matched to multiple ad groups? More concerning, notice how each term contained the specific reference to “online” and it was matched to some ad groups that don’t house/promote online certification ad copy?
This illustrates what happens when broad match keywords and Search Term Reports aren’t closely monitored.
Leveraging Broad Without The Bleed
1. House Match Types Separately
You’ll hear varying opinions on whether each keyword match type needs to be housed as its own Campaign or its own Ad Group – it really depends on the account. At some point, having each set of keyword themes as its own Campaign isn’t easily scalable when you’re dealing with enterprise-level spends and keyword volume. It can be helpful if you use it selectively on keyword themes that need their own spend levels because they’re budget-busters when shared.
That said, the key is this: house different match types on their own; at the very least, in their own Ad Groups.
This means you should have something like this:
Campaign A
[Keyword Theme Name Ad Group] – Broad
[Keyword Theme Name Ad Group] – Exact
[Keyword Theme Name Ad Group] – Phrase
Don’t ever put them all in one Ad Group. It gets messy, and when you look at Search Term reports to see the actual queries you’re being matched to, you’ll lose insight that’s instantly concise to you.
2. Limit Broad Match – Rely on Modified
I very rarely run straight broad match anymore, and rely much more heavily on modified broad to drive my keyword expansion/exploration and reach strategies. Let’s say you sell hats: surely you can make a judgment call on whether your terms should include “costume” or “party” or other identifiers. Bidding just on “hats” is a surefire way to get a boatload of junk traffic. (Note that the challenges you have with straight Broad Match will be similar in Phrase Match, so you need to be equally diligent there.)
Even with modified broad, you need to check your Search Term report constantly, and add negative keywords regularly. Which leads me to…
3. Don’t rely on AdWords for all of your keyword data.
Keywords are what you set up – the money is in the actual search queries. Remember that the keyword data you see in AdWords isn’t the actual story of the search queries you’re getting and what they’re doing.
Link AdWords to Analytics, and lean on latter. The AdWords interface works fine for cursory conversion details, but if you’re dealing with something that’s e-commerce driven and ROI-intensive, it won’t tell you everything. You will want to get very familiar with the Matched Search Query data in Analytics. This can be found in:
Acquisition –> AdWords –> Matched Search Queries
Export the revenue data and then back the cost data into it to figure out the true return on spend.
4. Negative keywords are a huge positive.
Be relentless in studying the Search Term report in AdWords for all of your Broad and Phrase match Ad Groups.
The Search Term report is under the Dimensions tab, in the “View” sub-menu:
Add negative keywords religiously. You can add them at the Campaign level or the Ad Group level. Think through this very carefully to maximize their use.
(Side note: Anecdotally I’ve heard that AdWords disregards keyword lists that exceed 50 negative keywords. If you need that many, you should figure out if you can break them into smaller ad groups anyway.)
Conclusion
Granular though it may be, the function of match types and negative keywords are one of the largest missed opportunities for Campaigns to run efficiently. Before revamping ad copy, bidding on more keywords, fussing with bids, or obsessing about what your competitors are doing, study the structure of your Campaign. Implement clean match type setups and smart negative keywords to laser-focus the users your ads are shown to so you’re reaching the most relevant ones.
Stop the click bleed, stabilize the paid search patient, and you’ll avoid major reconstructive surgery later.
