Is Voice Search Resurrecting Broad Match?
For Search Marketers, unmodified broad match has long been the lowest “don’t go there” option in targeted, profitable search campaigns. For those unfamiliar with keyword match types, broad match means that the designated keyword can appear anywhere in any user search string, or maybe not at all if Google or Bing’s algorithms deem that there is a “match”. Broad match can be “modified” by insisting that one or more words are included, but unmodified broad match is something of a dumping ground for random, long, or even untargeted queries.
We used to joke that Google threw everything but the kitchen sink into unmodified broad match traffic, so you can imagine our laughter (and tears) the day we found the query “kitchen sink” in a query report on an enterprise software keyword. As data-driven marketers it is this type of example that has trained us to hate, fear, and loathe unmodified broad match.
Before unmodified broad match became a dumping ground for unloved (or unsellable) traffic, it had a real role to play in search campaigns. It was a match type that could be used in “harvest mode” to understand query behavior and mine both new queries to move to phrase and exact match as well as negative keywords to block from campaigns. In the heyday of short searches and high volume keywords, broad match helped in sculpting keyword mixes for effective search campaigns. But when broad match modifiers were released, less risky options became available and unmodified broad match fell by the wayside. It made sense to abandon unmodified broad match since modified broad captured most of the relevant queries in most campaigns. Unmodified broad match wasn’t particularly useful any more.
But then voice search came along.
Broad match wasn’t as useful prior to voice search because typed search queries enforced a more limited range of query behaviors. Voice search changed that. Natural language is richer than typed search queries and therefore more nuanced and more varied. Add in Google’s developments in contextual understanding of threaded voice queries and suddenly the query range isn’t just long tail, it’s hyper-long tail. Tools are needed to capture this incredibly varied source of targeted searchers. And it turns out there is just such a tool: unmodified broad match.
Hopefully, the ad networks will carve voice search into a unique addressable audience where we can target and optimize separately, but for now old tricks might be the best for a new dog.