STARTUPS 
DATA 
DECISIONS 
Sami Can Tandoğdu 
@sctandogdu 
www.sctandogdu.com 
November 20th 2014
Who am I? 
• TED – Bilkent 
• Corporate Life – PwC 
• Entrepreneurship – Fikrimuhal, AçıkDemokrası 
• Director of Finance – PUBLIK 
• Freelance Consultant 
• Currently working for Soostone, a NYC based artificial intelligence 
/ machine learning startup
What is data driven 
decision making?
Data driven decision 
means depending on 
data on every single 
decision you make for 
your startup
First of all lets call it data 
informed decision making 
Never underestimate the your 
intuition!
Cognitive bias is your 
worst enemy!
Anchoring 
Relying on the initial 
information way too much
Case of the Entreprenuer that did not 
calculate his target market size 
•An entrepreneur was talking about his partnership with 
a flagship SF startup 
•His business was built around this new startup, which 
was quite big and growing double digit each year 
•He was always talking about SF startups revenue and 
growth, neglecting his own metrics 
•We calculated his company’s target market 
•Global market size was below 100 mn $
Confirmation 
Point of view bias: thinking that 
the world around you is the real 
world
Bandwagon Effect 
Popularity bias: Follow the herd 
mentality
Further reading 
Check 
https://en.wikipedia.org/wiki/List_of_cognitive_biases
Be careful while building 
your Decision 
Mechanism
AND Gate decisions 
Director Other members Outcome 
YES NO NO 
NO YES NO 
YES YES YES 
NO NO NO
Perfect AND Gate: 
İş Bankası
OR Gate decisions 
Director Other members Outcome 
YES NO YES 
NO YES YES 
YES YES YES 
NO NO NO
ME Gate decisions 
Director Other members Outcome 
YES NO YES 
NO YES NO 
YES YES YES 
NO NO NO
NO Gate decisions
Autocracy vs Democracy
Come up with a theory! 
Act like a scientist, try to collect 
data around your theory
Case: 1-1 QnA Site 
•A market place like product, helping people with 
problems with meet problem solvers 
•They were worried about the quality of answers 
•They added a public comment section – BIG MISTAKE! 
•They allowed comments on Youtube and Fbook – NO! 
•They created a sentiment analysis system focusing on 
answers to spot out ‘negative vibe’ 
•Found a correlation between negative vibes and 
complaints 
•Build a strategy guiding problem solvers
You don't need big data, 
you need meaningful 
data
Latent metrics vs 
predictive metrics
Further reading 
Check 
http://a16z.com/2014/09/05/why-amazon-has-no-profits- 
and-why-it-works/
Vanity Metrics 
Metrics that don’t give you any 
kind of insight
Finding the truth in data 
140.0% 
120.0% 
100.0% 
80.0% 
60.0% 
40.0% 
20.0% 
0.0% 
Budget realization 
Jan Feb Mar Apr May Jun Jul Aug Sep Nov
Actionable Metrics 
Metrics that you can act on!
If there is a correlation use that to 
your advantage! 
50 
45 
40 
35 
30 
25 
20 
15 
10 
5 
0 
5.5 
5.0 
4.5 
4.0 
3.5 
3.0 
2.5 
2.0 
1.5 
1.0 
Number of meeting and sales to meetings ratio 
Jan Feb Mar Apr May Jun Jul Aug Sep Nov 
# of meetings Sales to meeting ratio
Building decision models 
It’s not that complex!
Your best friends
HR: Hiring with data 
• Try to define the position 
• Understand the requirement for this position 
• Hard skills (education, experience) 
• Soft skills (company culture) 
• Quantify - create a scoring card 
• Reduce cognitive bias 
• Conduct interviews with multiple candidates 
• Conduct multiple interviews with each candidate 
• Review your model each 6 month 
• Have you done the right hiring? 
• Does your model work?
Thanks! 
Sami Can Tandoğdu 
@sctandogdu 
www.sctandogdu.com 
Find the presentation @ 
http://www.slideshare.net/SamiCanTandogdu

Startups Data and Decisions

  • 1.
    STARTUPS DATA DECISIONS Sami Can Tandoğdu @sctandogdu www.sctandogdu.com November 20th 2014
  • 2.
    Who am I? • TED – Bilkent • Corporate Life – PwC • Entrepreneurship – Fikrimuhal, AçıkDemokrası • Director of Finance – PUBLIK • Freelance Consultant • Currently working for Soostone, a NYC based artificial intelligence / machine learning startup
  • 3.
    What is datadriven decision making?
  • 4.
    Data driven decision means depending on data on every single decision you make for your startup
  • 6.
    First of alllets call it data informed decision making Never underestimate the your intuition!
  • 7.
    Cognitive bias isyour worst enemy!
  • 8.
    Anchoring Relying onthe initial information way too much
  • 9.
    Case of theEntreprenuer that did not calculate his target market size •An entrepreneur was talking about his partnership with a flagship SF startup •His business was built around this new startup, which was quite big and growing double digit each year •He was always talking about SF startups revenue and growth, neglecting his own metrics •We calculated his company’s target market •Global market size was below 100 mn $
  • 10.
    Confirmation Point ofview bias: thinking that the world around you is the real world
  • 11.
    Bandwagon Effect Popularitybias: Follow the herd mentality
  • 12.
    Further reading Check https://en.wikipedia.org/wiki/List_of_cognitive_biases
  • 13.
    Be careful whilebuilding your Decision Mechanism
  • 14.
    AND Gate decisions Director Other members Outcome YES NO NO NO YES NO YES YES YES NO NO NO
  • 15.
    Perfect AND Gate: İş Bankası
  • 16.
    OR Gate decisions Director Other members Outcome YES NO YES NO YES YES YES YES YES NO NO NO
  • 17.
    ME Gate decisions Director Other members Outcome YES NO YES NO YES NO YES YES YES NO NO NO
  • 18.
  • 19.
  • 20.
    Come up witha theory! Act like a scientist, try to collect data around your theory
  • 21.
    Case: 1-1 QnASite •A market place like product, helping people with problems with meet problem solvers •They were worried about the quality of answers •They added a public comment section – BIG MISTAKE! •They allowed comments on Youtube and Fbook – NO! •They created a sentiment analysis system focusing on answers to spot out ‘negative vibe’ •Found a correlation between negative vibes and complaints •Build a strategy guiding problem solvers
  • 22.
    You don't needbig data, you need meaningful data
  • 26.
    Latent metrics vs predictive metrics
  • 27.
    Further reading Check http://a16z.com/2014/09/05/why-amazon-has-no-profits- and-why-it-works/
  • 28.
    Vanity Metrics Metricsthat don’t give you any kind of insight
  • 29.
    Finding the truthin data 140.0% 120.0% 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% Budget realization Jan Feb Mar Apr May Jun Jul Aug Sep Nov
  • 30.
    Actionable Metrics Metricsthat you can act on!
  • 31.
    If there isa correlation use that to your advantage! 50 45 40 35 30 25 20 15 10 5 0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 Number of meeting and sales to meetings ratio Jan Feb Mar Apr May Jun Jul Aug Sep Nov # of meetings Sales to meeting ratio
  • 32.
    Building decision models It’s not that complex!
  • 33.
  • 34.
    HR: Hiring withdata • Try to define the position • Understand the requirement for this position • Hard skills (education, experience) • Soft skills (company culture) • Quantify - create a scoring card • Reduce cognitive bias • Conduct interviews with multiple candidates • Conduct multiple interviews with each candidate • Review your model each 6 month • Have you done the right hiring? • Does your model work?
  • 35.
    Thanks! Sami CanTandoğdu @sctandogdu www.sctandogdu.com Find the presentation @ http://www.slideshare.net/SamiCanTandogdu