Derek Rose vs Derrick Rose: Dunking search ambiguity

BACKGROUND

  • Derek Rose faced a unique and persistent challenge: name ambiguity. “Derek Rose” as a search term overlaps significantly with Derrick Rose, the NBA player. Whenever there was a spike in interest in the athlete, CPC’s for Derek Rose’s paid search campaigns rose sharply.
  • But we couldn’t simply eliminate all basketball-related terms, as some of those users were actually consumers misspelling the brand.
  • Trying to isolate these potential customers became the mission.

SOLUTION

  • Negative out all obvious Derrick Rose related keyword variants. Limitation - many people entering ‘Derek Rose’ were actually searching for the basketball player. This solution didn’t solve for that.
  • Separated campaigns into two structures, basketball audiences and non-basketball audiences, which gave us greater control over performance. Limitation - still experienced CPC volatility whenever interest in Derrick Rose spiked.
  • The final piece of the puzzle - shifted the bidding strategy to Target ROAS. This allowed Google’s algorithm to optimise not for clicks, but for value, prioritising users most likely to generate high-return conversions, even within ambiguous search terms.

RESULTS

  • Our hypothesis was that users searching for “Derrick Rose” queries fell into two distinct groups. 1) Users seeking information about the basketball player (low commercial intent). 2) Users intending to find the Derek Rose brand but misspelling the name (high commercial intent)
  • By adopting a tROAS strategy, we effectively instructed Google to prioritise the second group, those delivering higher revenue per click, while naturally de-prioritising low-value traffic.
  • Rather than fighting ambiguous intent through exclusions, we allowed value-based bidding to resolve it algorithmically.

IMPACT

CPC Reduction

54
- %

ROAS Improvement

x

Conversion Rate Uplift

+ %
Scroll to Top