Exact Match is widely accepted as the Cadillac of match types in paid search, and for good reason. Queries that exactly match the keyword triggering a paid search ad have the highest conversion rate of any query-to-keyword relationship. When used, a brand can rest assured the searcher is searching for the exact term targeted.
However, while some brands advertise offerings that can be adequately covered by exact match keywords alone, the vast majority of advertisers rely on broader match types to show ads for rare or new query variations.
In the case of brands with ever-changing, expansive product selections, broader match types are necessary to stay visible for queries which aren’t launched on exact match.
After exact match, which match type performs best?
My analysis of a number of high volume retailers that deploy multiple match types shows there might be little difference between phrase and broad match options for actively-managed accounts.
Little difference in conversion rate for broader match types
Below is a chart depicting Q4 2017 conversion rate relative to pure exact, based on Google’s definition of match type in search query reports. It is assigned based on the relationship of the query to the keyword rather than which match type the keyword triggered is assigned.
As with all match type analyses, these numbers are impacted by variables like keyword coverage and the degree to which paid search managers curate negative keyword lists to block out irrelevant traffic and shepherd queries to closer matches.
As you can see, pure exact matches had the highest conversion rate of any query-to-keyword match by far.
Exact (close variant) outpaced phrase and broad matches on desktop and tablet but was actually slightly lower on phones. The study below from Merkle (my employer) shows exact match close variant traffic share rose in the back half of 2017, this after Google’s change to the definition of what constitutes a close variant.
Comparing pure phrase match to pure broad match, the conversion rate is incredibly close, with no more than a one percentage point difference for any device type.
This is pretty surprising given the phrase match should result in tighter matches than full-on broad match and the general flack broad match has caught in the paid search industry over the years.
It also indicates that there might not be a clear difference in conversion rate between broad match modified (BMM) traffic (which is not reported on in the match type column) and pure broad, since BMM is essentially a looser version of phrase match in which the words must be present but can be rearranged.
These figures are impacted by how well brands deploy negatives to block out poorly-matched traffic, and also vary from keyword to keyword since different terms carry different possibilities for phrase and broad matches.
The answer for whether it’s better to use phrase, broad, or BMM for non-exact keywords might be different depending on your management and account specifics. It may also depend on which keyword you’re looking for.
Looking at the median advertiser, there’s not a clear answer that can be applied broadly to every type of brand. Even where differences in value exist, marketers should be able to account for them with proper management.
Bid right
There is an old expression in the paid search industry, “there are no bad keywords, only bad management”.
Even if the value a keyword is expected to drive for an advertiser is very low, the advertiser should set the appropriate bid based on that expectation. If a bid is too low for a keyword to show an ad in the search results, then the keyword won’t get any traffic.
If it’s high enough to get ad impressions, then the bid should ensure that the advertiser spends only what it can afford to when the ad does get clicks.
In the case of keywords with different match types, brands should try to get as much pure exact traffic to keywords set to exact match as possible and set the most precise bids possible for this traffic.
That doesn’t mean brands should forego broader match types. Broad and phrase match keywords can help pull in new or rare variations of a search query and can operate at the same return on ad spend as exact so long as bids are set appropriately.
Keep the negatives tight
Negative keywords are certainly an important part of this, as brands should be looking to block out as much irrelevant traffic as possible so that they may set appropriate bids for a relevant phrase and broad matches.
This is also true of keywords set to exact match, with the evolving definition of close variants steadily expanding the amount of traffic coming from queries that aren’t true exact matches.
Negatives are also important for funneling traffic to the best keyword matches possible to ensure the bids set for those keywords are based on the best possible collection of traffic.
For example, blocking broader match types from being triggered for keywords which exist on exact match within the account. This can be tricky, as Google sometimes clearly prefers odd keyword variations for some queries rather than exact match keywords that perfectly match the query.
Just because pure exact matches convert at a higher rate than other query-to-keyword relationships doesn’t mean brands should only ever rely on exact match.
Even if phrase matches did convert at a meaningfully higher rate than broad matches, it wouldn’t mean brands should only deploy phrase and exact match and ignore broad match. It just means that bids need to be set appropriately for each match type variation.
While my title did specifically frame the question of match types in terms of which was best, it’s really difficult to write a succinct title to describe a post that compares match type conversion rates and argues there is little difference in conversion rates.
It also doesn’t matter; we shouldn’t be pitting match types against each other.
Conclusion
A lot of paid search marketers have very strong opinions on the best match type strategy and many of those opinions have been formed by real-life examples of broad match pulling in terribly unqualified traffic.
This is something pretty much everyone involved in paid search for any length of time encounters to one degree or another. You have to wonder if broad match gives Google too much flexibility to determine relevant queries.
If we really look at how well Google matches the intent of the query when using broad match as judged by conversion rate, it’s not obvious that broad matches will perform worse than phrase matches.
Given that BMM is essentially a more flexible iteration of phrase match, it also follows there probably isn’t a huge distinction between broad and BMM.
As mentioned earlier, specific keywords might benefit more or less from a specific match type than others. However, the data presented here indicate overarching arguments lifting up one (non-exact) match type over others might not be accurate for all advertisers.
Outside of exact matches converting at a meaningfully higher rate than all other match types, the winner seems far murkier than many believe it to be.
The battle shouldn’t be used to eliminate specific match types from paid search campaigns. Any match type can “work,” so long as it’s properly managed.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.
About The Author
Andy Taylor is Associate Director of Research at Merkle, responsible for analyzing trends across the digital marketing spectrum for best practices and industry commentary. A primary contributor to the Merkle Blog, Dossier, and quarterly Digital Marketing Report, he speaks frequently at industry conferences and events.