Data is funny.

Let's say I am an amateur bird watcher.

If I tell you that after watching blue jays every morning I generally see more
blue jays when there is more seed on my birdfeeder, you might start putting more
birdseed on your feeder to attract more blue jays.

Then, my pro bird loving friends inform me red-headed woodpeckers and blue jays
don't get along.

Despite carefully measuring the seed, meticulously counting the unique blue
jays, and logging my data, as I begin to measure a new data point (presence of
woodpeckers), I find that my seed volume data has actually misled us into
wasting tons of bird seed when we should have invested in woodpecker
countermeasures to increase the amount of blue jays at our feeders.

This is why data is funny. You can talk about specific data points, have
published sample sizes, be transparent with your results, and have no bias on
how things turn out, but if you inaccurately attribute causation to the wrong
data point, data will turn itself against you. This stuff takes real
professionals to understand and make useful.

Outreach has strived for 3 years to find a better signal rather than continuing
to rely on signals, tactics, and silver bullets that sound good but don't create
outcomes.

Check out this better signal in the comments.



Posted by Mark Kosoglow on LinkedIn
link: linkedin.com/in/mkosoglow