1. Rand Fishkin, Wizard of Moz | @randfish | rand@moz.com
Great Failures in Competitive Intel
Why so many analyses of the competition fail, and
how web marketers can improve the process
8. Recommended Action:
Test removal of our paid search
campaigns for a subset of keywords.
This isn’t necessarily a crazy, terrible idea.
So why do most of us know it won’t work?
9. Observation:
We saw X.
AFlawed Competitive Intel Process
Analysis:
We collected data.
Correlation:
We validated X’s
frequency.
Recommendation:
We proposed an experiment.
11. Observation:
We saw X.
Analysis:
We collected data.
Correlation:
We validated frequency.
Recommendations:
We proposed validation methods for each potential cause.
Hypotheses:
We brainstormed potential
root causes.
20. Cast a Wide Net
Analyze:
Our four biggest
competitors in the
search rankings.
21. Cast a Wide Net
Analyze:
Our four biggest
competitors in the
search rankings.
Analyze:
A broad range of
competitors, noting
correlations that predict
both failure and success
28. Via Visual Website Optimizer
A/B Test Results
Page without the shield icon had 400%+ CR
improvement vs page with the icon.
[Note: results ARE statistically significant]
29. We need to remove the
security messages on our
site ASAP!
33. Via Kayak’s Most Interesting A/B Test
A/B Test Results
“So we decided to do our own experiment about this and
we actually found the opposite that when we removed the
messaging, people tended to book less.”
- Vinayak Ranade, Director of Engineering for Mobile, KAYAK
36. #1: We don’t know for certain why or whether a
tactic worked (or if it will work for us).
#2: Copying a competitor’s strategy ensures
that we lag rather than leap them.
46. Now the 2 Questions Become…
TwitterPR
Social Ads Video
Instagram
Pinterest
Image Search
Facebook
Which of these channels/tactics can
successfully reach our audience?
47. Now the 2 Questions Become…
TwitterPR
Social Ads Video
Instagram
Pinterest
Image Search
Facebook
Who’s great at this stuff, and what
are they doing that works?