Level Insights 04.17.2017


Test Data

Related Industries
Related Capabilities

What the data tells us about the death of exact match and its impact

Google’s latest change to exact-match keywords toward a more “semantic” approach cedes more control to its algorithm, but we may be further from the Singularity than the update indicates. From Search Engine Land:

This represents a shift away from keywords and toward user-driven marketing, in line with what we’ve seen over the past year.

Google Exact-Match Change Summary:

  • Google has removed function words from consideration when serving an ad to a searcher. These words express grammatical relationships with other words within a sentence, adding no lexical value of their own. Examples are:
  • Prepositions — of, in, between
  • Pronouns — he, she, they
  • Determiners — the, either, more
  • Conjunctions — and, that, while
  • Auxiliary — is, have, do
  • Particles — no, not, as
    • While ignoring the determiners and auxiliaries doesn’t seem like a big deal (Pittsburghers have been omitting “to be” from their speech for generations), it’s easy to see how disregard of prepositions and particles could result in some ads showing as false positives. The study cited in the article found that including semantically related search terms yielded unqualified traffic; this is a cause for heightened vigilance in monitoring campaigns. Marketers also must consider the fact that Google entrusts their machines with responsibilities that they are not yet sophisticated enough to handle without error.
    • Homonyms — like the “trailer” example cited in the article — present an interesting challenge as always, which is now accentuated by the disregard for conjunctions and plurals. Like most other Google updates, this will require heightened vigilance on the part of Level and search marketers everywhere to ensure that we continue to only pay for clicks that have the highest change to resulting in sales qualified leads.
    • For more information on this change, see the Search Engine Land summary here: http://searchengineland.com/google-exact-match-close-variants-word-order-function-words-271395

    Or see Google’s official announcement here: https://adwords.googleblog.com/2017/03/close-variants-now-connects-more-people.html.

Level Insight:

  • Robust keyword builds for larger clients limit the negative effects of this change. For smaller accounts, this change will have the most dramatic impact, due to Google’s limitations on advertisers’ ability to expand keyword sets into rarely searched terms. These terms were previously covered in modified-broad match campaigns, but may now be covered in exact-match efforts (which are managed differently than those modified-broad efforts).
  • This change has the most dramatic (negative) impact on small- to mid-size, highly targeted and niche-product-focused accounts due to Google’s minimum impression requirements for keyword bids.
  • Close-variant search query performance for these accounts are performing worse than their actual exact-variant counterparts — to the tune of a 70% higher CPA, in this blogger’s data.
  • This change also forces advertisers to be more stringent in monitoring and optimizing negative keywords in exact-match efforts that previously did not necessitate such heightened consideration.
  • This change relinquishes control from the advertiser and relies more heavily on Google’s algorithmic system, in what was previously the most controllable bidding system offered by search engine platforms.
  • We at Level will remain vigilant as always to ensure that smaller accounts — in the range of $5–$20K/month in media spend — do not experience unnecessary rising costs due to assumptions made by machine learning that, no matter what Google claims, have not attained sentience.

Featured Perspectives

Our latest thoughts, ideas, and approaches.