dms: infersystems pitch-a-kucha: data analysis: math or religion?
DESCRIPTION
Sunil Sharma, CEO of InferSystems, will use a light-hearted history of mathematics to illustrate how objective thinking facilitates early innovation and why innovators quickly abandon it. The near-religious fervor that follows sparks ego battles and turf wars that hinder progress and hurt the ad tech industry. Whether you believe that ad tech has arrived or is still on its way, this critical examination of the market will both entertain you and make you think.TRANSCRIPT
A Brief History of Math
Perspectives for the ad tech market
June 26, 2012By: Sunil Sharma
Most advanced math, including the “Pythagorean” Theorem, seems to have originated in India at least 6,000 years
ago
By 500 B.C., Pythagoras and his friends regularly cruised the Mediterranean on ancient party boats
Seriously…but they were much more than that…
They believed two central things about God
1. That God exists, and:
2. That God would never create “irrational” numbers (fractions)
The Pythagoreans, led by Pythagoras, were a fanatical religious cult that practiced mathematics and took lots of cruises
…. And what was his reward?
Until one day, when a Pythagorean named Hippasus proved that irrational numbers exist (while aboard a party boat)
They murdered him by throwing him overboard
*This photo is a reenactment
This incident teaches us two key lessons about human nature
1. Nobody likes a “know-it-all”
2. Ego; the defense of our viewpoints can close our minds to the truth, to innovation
Similarly, Vilfredo Parto fell victim to a classic cognitive fallacy when he proposed the “80/20 rule” to describe wealth distribution
1906: Pareto Principle shows 20% of people have 80% of wealth in all nations; accepted as a rule
1
Faulty reasoning: Only European nations studied; all had similar conditions2
As conditions changed globally, the 80/20 “rule” did not hold up and was debunked3
The 80/20 “rule” is a classic case of going by what the data “tells” us
Data tells us nothing… We infer from it… sometimes correctly
….In ad tech, 20% of companies are innovating, and the remaining 80% want to throw them overboard
“I have the best algorithm, and the best managed service, and the best
technology, and the shiniest objects, and the biggest ego…
For example, only a very tiny fraction of data is actually being leveraged to provide scalable performance to advertisers
Often less than 1% actually used; creates risks for market participants
Company type Risks Mitigation tactics
Buyers and sellers:•Advertisers •Publishers
• Wasted proprietary data• Commoditized optimization• Lack of scale
•Active engagement, not “set it and forget it”• Compel innovation into the supply chain
Intermediaries: •Agencies•DSPs•Networks•Optimization companies
• Commoditization• “Leapfrogging” of market leaders
• Avoid the “we can build everything” mentality• Management can make decisions too
No one company can do it all “the best,” and collaboration is necessary to provide maximum value to advertisers
Don’t be this guy
Be a lion… attack the innovation challenge