As a person who has worked in marketing I have spent a great deal of time involved in an activity that I shall call “plunging the metaphorical analytics toilet.”

Allow me to expound.

In marketing, tracking metrics is hugely important. It’s the reason that advertising and marketing professionals love to quote “Half the money I spend on advertising is wasted; the trouble is, I don't know which half,” and attribute it to a famous retailing magnate (I couldn’t verify the authenticity of the attribution, so I won’t perpetuate it).

The principle is simple, but nuanced: even in the age of “drill-down” targeting for media like pay-per-click ads, only a portion of the audience is interested in what you’re advertising. Even a perfectly positioned offer isn’t going to convert 100% of viewers. Thus, any money spent showing ads to people who aren’t interested is, effectively, wasted. The amount of waste varies greatly, but 50% has a very authoritative ring to it.

Back to the analytics toilet.

There are tons and tons of metrics you can track. You set up Google Analytics, maybe a few custom spreadsheets and boom, the plumbing works and the water (data) is flowing. Until it doesn’t. At which point you grab a plunger and a plumber’s snake and attack the blockage. You look for unmonitored traffic sources, or robot impressions or poor email delivery – anything that might be clogging your report and making it stink.

Which brings me to another thoughtless truism that people regurgitate in business “Numbers don’t lie.”

You might as well say that a sledge hammer doesn’t lie. Strictly speaking, that’s true, but I can use that hammer to drive tent stakes or pulverize a human head – intent is everything.

I contest that numbers are, in fact, the finest instrument ever conceived to broadcast and harmonize deceit. Enron is a choice example.

I have mis-reported numbers more times than I care to admit, not because I was trying to cover a problem, simply because I didn’t know they were wrong in the first place. Only later did I discover my error.

Numbers as measurement (conceding that there are other applications) are approximate, a tool of separation and aggregation, as sharp and sticky as a scalpel and superglue, or as dull and weak as a rolling pin and chewed gum.