We use cookies. You have options. Cookies help us keep the site running smoothly and inform some of our advertising, but if you’d like to make adjustments, you can visit our Cookie Notice page for more information.
We’d like to use cookies on your device. Cookies help us keep the site running smoothly and inform some of our advertising, but how we use them is entirely up to you. Accept our recommended settings or customise them to your wishes.

Bing Ads Match Types Drive Unexpected Traffic

We recently wrote about the definition of Bing’s Broad Match Modifier (BMM) and how it’s use of ‘equivalent expressions’ could cause query-to-keyword matches many advertisers wouldn’t expect with BMM.

Looking at how queries are matching to other, more traditional match types, however, it’s clear that Bing Ads is consistently serving keywords on queries which would be considered by most to be outside of the realm of possibility given the match type of the keywords.

While broad match keywords obviously give engines a wide net to cast in displaying ads for relevant queries, match types like exact and phrase match look by their definitions to be pretty cut and dried in terms of which queries can trigger ads. However, that’s not always the case.

When Phrase Match Isn’t Phrase Match

By Bing’s definition, ‘phrase match triggers your ad when all of the words in your keyword match the words in a user’s search query, in exactly the same order, even if other words are present in the search term.’

Pretty straightforward, right? All the terms in a keyword need to be present in the query in the same order as in the keyword for an ad to be triggered.

However, our analysts regularly find examples of queries which don’t fit this definition triggering phrase match keywords. For example, the query ‘silk thermals for women’ triggering the keyword ‘silk long underwear.’

Bing Ads considers the term ‘for’ to be a ‘stop word’, which is essentially a commonly used term that might not be necessary for understanding the intent of a query. While in the past Bing normalized queries and keywords with stop words to eliminate these terms from consideration in their matching algorithm, they announced in 2015 that this normalization process would no longer eliminate stop words.

Per Bing’s example below, the query ‘the office’ would have been normalized to just ‘office’ with Bing’s old normalization, but is viewed as ‘the office’ under the new system rolled out last May.

This doesn’t appear to be the case yet for all queries with stop words.

Aside from the issue of stop words, ‘women’ is hugely important to the meaning of the keyword, and it’s difficult to explain why this term (or a close variant, explained below) was not required.

Then there’s the issue of ‘silk thermals’ vs ‘long underwear.’

Close variants are at play with Bing phrase match, and keywords with this match type are supposed to show for queries that are slightly different than the keyword being bid on, such as in these examples provided by Bing:

However, the query-keyword example mentioned above does not seem to fall under any definition of close variants.

Instead, it seems like the ‘equivalent expressions’ possibility allowed in BMM might have been used in this case, with ‘thermals’ being equated to ‘long underwear.’ ‘Equivalent expressions’ are basically synonyms, with Bing’s literature including an example where the term ‘cabin’ would qualify as a similar expression for the term ‘lodge.’

Bing does not list ‘equivalent expressions’ as a possibility in its phrase match type definition, though, and this is not an expected match based on the definition provided.

Et tu, Bing Ads Exact Match?

We’ve also come across cases where exact match doesn’t seem to be serving as expected. For example, the exact match keyword for ‘[brand name redacted]’ showing ads for the query ‘what is a [brand name redacted].’

Bing’s definition of exact match reads, ‘exact match triggers your ad when the exact words in your keyword appear in a customer's search query, in exactly the same order.’

All three of 'what', 'is', and 'a' are listed as stop words by Bing, and were included in the list of words that would no longer be removed from a user’s raw query through Bing’s old normalization process in their announcement last spring.

So again, it seems Bing hasn’t yet fully eliminated the normalization of queries in this way.

Close variants are also allowed for exact match keywords, but this example doesn’t seem to qualify as a close variant either based on Bing’s definition and examples:

Bing’s Matching Has Become Much Less Restrictive over the Years

As recently as 2011 my colleague Mark Ballard was lamenting the fact that Bing Ad’s keyword-to-query mapping was much more restrictive than Google’s, at the time resulting in Bing/Yahoo generating non-brand clicks from broad-matched ads at just 70% the rate of Google.

However, Bing Ads expanded its use of broad match in 2012, driving up click and spend growth significantly throughout the year. In Q4 of 2012, clicks increased 39% and spend increased 54% year-over-year, according to our Digital Marketing Report at the time.

Bing Ads maintained this level of broad matching in following years, and when Google rolled out mandatory close variants for all phrase and exact match keywords in 2014 Bing moved soon after to include close variants in its match types.

Bing Ads has certainly come a long way in loosening the restrictions of their query-to-keyword matching, and advertisers must stay aware of what their keywords are serving ads for now more than ever before.

Stay on Top of Your Search Query Reports

While these wrinkles in which types of queries match to exact and phrase match keywords typically account for a very small share of traffic, it’s still surprising that they occur at all. This is something Bing is consistently working on, but for now the safe route is to make sure you’re checking your search query reports even for keywords with restrictive match types to search for potential negatives.

Those match types may not be as restrictive as you think.

Hat tip to Lauren Hale and Allen Wong.