Mission
To build applications that help you make better data-driven decisions in your day-to-day activities. So much data exists online today and our goal is to enable you to effectively use this data to make better personal decisions.
Background
We make a dizzying number of decisions - what to eat, where to live, what car to buy, where to go to school, what to study, what career to pursue, what to read etc. How do you make these decisions? Are you data-driven? Or do you mostly use intuition?
Evidence from the fields of social science, psychology and even economics indicates that we in fact are quite irrational in our decision-making. The popular bestseller by MIT Professor Dan Ariely, “Predictably Irrational” describes this irrationality quite well. Princeton Professor and Nobel Laureate, Daniel Kahnemann, in his bestseller, “Think, Fast and Slow”, describes an intuitive and automatic System 1 and an effortful and deliberate System 2 that we use in decision-making. System 1 plays a big role in decision making in our daily lives but is also quite prone to serious errors of judgment.
Not only are we irrational in many of the decisions we make but we invariably are influenced by hidden forces and cues beyond our control. In the sciences, knowledge comes from data, experimentation and observation. In many other realms of human endeavor such as business, we see a preponderance of data-driven decision-making. If the data-driven approach works quite well in many spheres of human activity, could it work as well in our personal activities?
Evidence from the fields of social science, psychology and even economics indicates that we in fact are quite irrational in our decision-making. The popular bestseller by MIT Professor Dan Ariely, “Predictably Irrational” describes this irrationality quite well. Princeton Professor and Nobel Laureate, Daniel Kahnemann, in his bestseller, “Think, Fast and Slow”, describes an intuitive and automatic System 1 and an effortful and deliberate System 2 that we use in decision-making. System 1 plays a big role in decision making in our daily lives but is also quite prone to serious errors of judgment.
Not only are we irrational in many of the decisions we make but we invariably are influenced by hidden forces and cues beyond our control. In the sciences, knowledge comes from data, experimentation and observation. In many other realms of human endeavor such as business, we see a preponderance of data-driven decision-making. If the data-driven approach works quite well in many spheres of human activity, could it work as well in our personal activities?
Benefits
Certainly, not all decisions require data; this is especially true when there is a moral or ethical dimension to the decision or in matters of religion, faith, love etc. In endeavors where we desire novelty and creativity, intuition, and leaps of imagination rather than data will likely yield a superior result. Data-driven decision-making also implicitly assumes that the future will be like the past - which may not always be true. Also, where fear, flight, or fight is the appropriate response, you probably are better off relying on intuition rather than data.
Despite these caveats, data can help us make better decisions and reduce unnecessary risks in many other aspects of our day-to-day lives. The massive amounts of data online, used in the right circumstances, will surely help us make better decisions where each of us can define what “better decision” means to them individually.
Despite these caveats, data can help us make better decisions and reduce unnecessary risks in many other aspects of our day-to-day lives. The massive amounts of data online, used in the right circumstances, will surely help us make better decisions where each of us can define what “better decision” means to them individually.
Options
Search engines like Google and Bing, that help you find information about almost anything, in essence help you make data-driven decisions in some respects. But is typing a few keywords in a text-box on Google the best way to get the data you need to make the important decisions in your personal day-to-day life? For example, if you are contemplating a career change, how efficient is the current model of keyword-based search engines at helping you quickly reach an informed decision? Is there a better way?
In consumer retail, a lot of effort and innovation has been expended to help you make data-driven decisions. Data comes in the form of easily available, detailed product information, expert product reviews, Top 10 lists, product recommendations based on buyers similar to you in some respect, consumer product reviews, QR Codes etc. In fact, in addition to established players like Amazon and eBay, a slew of start-ups have emerged to help you make product decisions using approaches such as:
1- Personalization – tell us your preferences and we’ll tell you what’s best for you
2- Expert – we are the experts and here is what we recommend in the form of a Top 10 list or an expert review
3- Recommendations – users similar to you in one or more aspects prefer these products so you’ll likely prefer these as well
4- Reviews/Ratings – here is what other users think about these products and how they have rated or ranked them
Other interesting areas are around health-related decisions; with companies emerging that help people make data-driven health-related decisions. Similarly, data in the form of consumer reviews plays a key role when we decide where to eat or which hotel to book. But beyond product and health-related decisions, there is a multitude of other decisions we make on a day-to-day basis in which data could also be brought to bear. Can the personalization, expert, recommendations, reviews or hybrid approaches be applied to these other realms in which we make decisions?
In consumer retail, a lot of effort and innovation has been expended to help you make data-driven decisions. Data comes in the form of easily available, detailed product information, expert product reviews, Top 10 lists, product recommendations based on buyers similar to you in some respect, consumer product reviews, QR Codes etc. In fact, in addition to established players like Amazon and eBay, a slew of start-ups have emerged to help you make product decisions using approaches such as:
1- Personalization – tell us your preferences and we’ll tell you what’s best for you
2- Expert – we are the experts and here is what we recommend in the form of a Top 10 list or an expert review
3- Recommendations – users similar to you in one or more aspects prefer these products so you’ll likely prefer these as well
4- Reviews/Ratings – here is what other users think about these products and how they have rated or ranked them
Other interesting areas are around health-related decisions; with companies emerging that help people make data-driven health-related decisions. Similarly, data in the form of consumer reviews plays a key role when we decide where to eat or which hotel to book. But beyond product and health-related decisions, there is a multitude of other decisions we make on a day-to-day basis in which data could also be brought to bear. Can the personalization, expert, recommendations, reviews or hybrid approaches be applied to these other realms in which we make decisions?
Way Forward
Many of the decisions we make are about material things, what
we own, what we should own, but the really important decisions are about what
we should do. Should I quit my day job and start a business? Should I get a
graduate degree? Should I change careers? Decisions of this type are very different
from decisions about which shoes to buy – these are high impact decisions.
These types of decisions are personal and usually life-altering. So how should we go about making these high-impact
decisions?
As with product decisions we make, can we similarly bring data to bear on these high-impact decisions that greatly impact our personal day-to-day lives? So instead of using data to answer the question- which one should I buy, can we also use data to help to decide if we should buy it in the first place? At Manzia, we believe the answer is an emphatic YES.
As with product decisions we make, can we similarly bring data to bear on these high-impact decisions that greatly impact our personal day-to-day lives? So instead of using data to answer the question- which one should I buy, can we also use data to help to decide if we should buy it in the first place? At Manzia, we believe the answer is an emphatic YES.
Building Blocks
A new wave of technologies around Mobile, Big Data, Data Science and Cloud Computing are coming together to make very powerful applications possible. Google’s self-driving car is probably the most high profile example of this type of application but other applications like Google Now are pointers in this new direction where we begin to use the power of data-driven technology to help us make better decisions. Better and smarter predictions is the main promise of this new wave of computing that provides deeper and richer insights in the mountains of data online.
Unlike previous computing paradigms, where you had to know what you were looking for, in this paradigm, machines automatically glean intelligence from massive amounts of data and make predictions for you. The answer to the question, Should I quit my job and start a business? – would be a probability of success where I get to define what success means to me. These types of emerging applications are now being referred to as predictive applications.
So what are the building blocks of applications that would support data-driven high-impact decisions? They include:
1- Data – getting the right kinds of data to support high-impact decision-making is crucial. Data integration is also paramount as its likely to come from different sources and in different formats
2- User Experience – what is the best User Interface for this type of application? Is it extremely easy to use by the average user?
3- Personalized – does the application help a specific user make a better high-impact decision based on how they have individually defined “better decision”
4- Real-Time – does the application help the user make a data-driven decision in real-time, anywhere and anytime
5- Learning – does the application “learn” from the user in order to make better and better predictions?
6- Security & Privacy – Is the application secure? Is the user’s personal information protected?
7- Confidence Score – the application should provide the user with a confidence level of its predictions so the user knows when to ignore or not ignore the application’s predictions. Even better is not to provide a prediction unless the confidence level is above a pre-set threshold.
Unlike previous computing paradigms, where you had to know what you were looking for, in this paradigm, machines automatically glean intelligence from massive amounts of data and make predictions for you. The answer to the question, Should I quit my job and start a business? – would be a probability of success where I get to define what success means to me. These types of emerging applications are now being referred to as predictive applications.
So what are the building blocks of applications that would support data-driven high-impact decisions? They include:
1- Data – getting the right kinds of data to support high-impact decision-making is crucial. Data integration is also paramount as its likely to come from different sources and in different formats
2- User Experience – what is the best User Interface for this type of application? Is it extremely easy to use by the average user?
3- Personalized – does the application help a specific user make a better high-impact decision based on how they have individually defined “better decision”
4- Real-Time – does the application help the user make a data-driven decision in real-time, anywhere and anytime
5- Learning – does the application “learn” from the user in order to make better and better predictions?
6- Security & Privacy – Is the application secure? Is the user’s personal information protected?
7- Confidence Score – the application should provide the user with a confidence level of its predictions so the user knows when to ignore or not ignore the application’s predictions. Even better is not to provide a prediction unless the confidence level is above a pre-set threshold.
Reflections
From an economics perspective, these applications would help us compute the opportunity cost of taking one action over another action. Beyond the benefits, it’s also important to realize and appreciate the consequences of this type of application:
Will creating more options for people lead to “analysis paralysis”, where it becomes even harder to make decisions if even more options or more data is available? How does the application mitigate this?
How much and what kind of personal data does this kind of application require? Some predictive apps like Google Now that are attempting to predict user intent are on one end of the spectrum, is there a middle ground where the application is not “creepy”?
How does the application ensure the user always realizes that its answers are only probabilistic? The answers are not correct or wrong; they are in essence a data-driven “guess”, the user can and should ignore the suggested answer if its confidence level is not high enough.
In the final analysis, social scientists, neuro-scientists, psychologists, and now even economists tell us we are fundamentally irrational in the way we make decisions, this is a good thing is certain spheres of life, but in others where it makes sense we can and should use data to inform our personal day-to-day decisions. The pieces required to make this vision happen are in place, the technology, the data, and the algorithms, therefore why not try and see how far we go?
Will creating more options for people lead to “analysis paralysis”, where it becomes even harder to make decisions if even more options or more data is available? How does the application mitigate this?
How much and what kind of personal data does this kind of application require? Some predictive apps like Google Now that are attempting to predict user intent are on one end of the spectrum, is there a middle ground where the application is not “creepy”?
How does the application ensure the user always realizes that its answers are only probabilistic? The answers are not correct or wrong; they are in essence a data-driven “guess”, the user can and should ignore the suggested answer if its confidence level is not high enough.
In the final analysis, social scientists, neuro-scientists, psychologists, and now even economists tell us we are fundamentally irrational in the way we make decisions, this is a good thing is certain spheres of life, but in others where it makes sense we can and should use data to inform our personal day-to-day decisions. The pieces required to make this vision happen are in place, the technology, the data, and the algorithms, therefore why not try and see how far we go?
