Why Talent Analytics Matters

Albert Einstein once wrote, “Everyone is genius. But if you judge a fish by its ability to climb a tree, it will live its whole life believing that it it is stupid.” We need to ask the same question to our employees in our organization, “What is our employee’s genius?” Have we assigned the right people to the right job?


In our experience, we may have seen smart employees but somehow they are lacking from their performance. Can these employees considered as “bad performers”? Should we fire them? On the other hand, we may also have seen those with less competence people from different division/unit but they’ve always been able to meet their performance target, just because the target was too easy or too low. Can we label them as “star performers” and give them even higher reward?” As a company, would you rather pick the less competence employee to advance? We have seen this in many situations. How can we solve this problem?

This is where talent analytic matters. Using analytics, we can identify low performer with higher potential to become “high potential” to become star performer with condition if he/she is relocated to different division or if he/she is given a specific training. Talent analytics provides linkage from “candidate profile” to “business performance”, thus we can correctly assign the right people to the right job. Overall employee performance will increase, employee satisfaction will increase and ultimately the company’s overall performance and productivity will increase too.

On top of the sales performance review and talent re-allocation, there are many insights that we can leverage from talent analytics. According to unit4software.co.uk, a global cloud-focused business service provider, here are the predictive insights:

  • Employee retention – what drives high levels of engagement and retention?
  • Sales performance – what drive high achieving sales people?
  • Leadership pipeline – what makes successful leaders and can we developed the ones that we currently don’t?
  • customer retention and leadership gaps – where is our current talent gaps in the organization and what gaps can we predict in coming years
  • Candidate pipeline – what is the standard of candidates in our pipeline and can we make better decisions about who will succeed at our organization?

Are you ready to create your own talent analytics? Find some tips here.


Disclaimer: This is a personal blog. The opinions expressed here solely represent my own opinion and not those of my employer or any other institution.

Tagged with: , , , ,
Posted in Uncategorized

The Right To Be Forgotten: Erasing Bad Memories From The World


In the movie “Eternal Sunshine of the Spotless Mind” starring Jim Carey and Kate Winslet, the estranged couple decided to have each other erased from their memories; I wonder if we have this technology today, would people ant to wipe their bad or unwanted memories?

Even though the technology to partially erasing memory is not yet arrived, however, the ability to remove memory from the world does exist. About 2 weeks ago, the European court of justice said that Google  must comply with request from individuals to remove links on search result pages to newspaper articles and other web pages that might cause embarrassment. They call this “the right to be forgotten”.

Do you think this should be allowed? How about freedom of speech? Do we have the right to know if people have done bad things 10 years ago? Should this person have the right to be forgotten for what he/she did 10 years ago if the person has changed to be a better person?

When we talk about predictive analytics, it has an assumption that people habits are unlikely to change over time; this is the reason of why predictive model can be calculate potential of an event/outcome: who will buy, who will quit, who will get injured, who will go to jail, even to predict who will be dead in the next 1 year.

Using past data, a predictive modeler can identify someone’s probably future. The question is: how long back should we use the past data? Is 1 year data considered enough? What about 2 or 5 or 10 year? Would the past data from 10 years back still relevant to predict someone’s future?

For example, a person with criminal record 10 years ago may still have the same criminal habit in the past 1 year. In this case, we don’t need the past 10 years data, 1 year data is enough. On another example, a person with shop lifting record 10 years ago but she is now a successful lawyer. Would this still make her as a high probability to do criminal? In her case should she be forgiven that she has the right to be forgotten by newspapers and Google regarding her past criminal record?

What do you think?

Posted in Uncategorized

Talent Analytics: How to Find Top Talents and How Job Seekers Find the Right Job

How do you know that the candidate that you hire will become a high performer? Should you choose candidate with 3.5 GPA and reject candidate with 2.5 GPA? However, how do you know for sure that higher GPA will have better performance compared those with lower GPA? And why company like Google and AT&T start choose candidate with less academic records or from non prestigious schools?

On the other hand, if you are a job seeker, with your current strength and weakness, education background and personal background, what kind a job that you should seek to enhance your career faster and live a happy life? Should you choose to become a management trainee or technical administrator?

recruiter2 jobseeker

Currently HR uses several methodologies to filter thousands of resumes to eventually identify top talents. Several criteria are set to separate “Potential” candidates vs. “Non-potential” candidates. For example: minimum education requirements e.g., bachelor degree or minimum GPA e.g., at least 2.7 or must reside within the city area. In addition, several test may be conducted to filter candidate. For example, numerical test, behavior test or aptitude test. All of these criteria and tests are utilized to identify “Good” candidate as “Top” talents. However, how do we know for sure tat these “Good” candidates will bring higher business performance?

Why Talent Analytics?

According to Deloitte study, based on a survey of 436 North American companies, the study reveals that advanced talent analytics is helping to achieve better talent outcomes in terms of leadership pipelines, talent cost reductions, efficiency gains and talent mobility – moving the right people into the right jobs

Talent analytics provides  linkage from “candidate profile” to “business performance”. Talent analytic can provide prediction of future performance for each candidate. This combination is only recently been introduced to correlate people with performance (source: http://www.talentanalytics.com).

Talent analytic will help us identify the best talent pool that will have the best potential for business performance (i.e., higher revenue, better precision, higher capability, etc). The best talent pool does not necessarily have the highest GPA from the best university. They can have medium GPA from mediocre university but perhaps have 2 year of experience. Talent analytic uses “candidate score” that match business performance criteria directly.

If you’re a job seeker, Talent Analytics can help you the best career track for you to advance faster. The analytic will compare your “career score” to several job descriptions such as (Safety staff, Technical staff or a Management Trainee). Based on this score, Talent Analytics can provide a list of available job offer with the highest “career score” for you. It will be a win-win solution for both the job seeker and recruiter.

How to build Talent Analytics?

First you need to ask, what is my main business challenges that need to be address. What kind of Talent that I need to focus to address this problem. This is a very crucial factor to start Talent Analytics. In addition, do you need to consider staff attrition into the model? What should be the key performance indicator?

Recent report by Bersin (2012) mentioned that to develop talent analytics, we need to have the data that generally falls into three categories:

  • People data: such as demographics, skills, reward, engagement
  • Program data: such as attendance, adoption, participation in programs ranging from training and development and leadership programs to talent management and key projects and assignments
  • Performance data: performance ratings and data captured from the use of instruments such as 360, goal attainment, talent, succession programs and talent assessment.

Let’s say your objective is to find talents for Petroleum Engineers for off-shore drilling company in Asia. You have a pool of talents. But how to pick the best candidate that will bring the best performance? First, you need to have these 3 types of data, People data, Program data and Performance data of your current staff. The people data should contain the data that you gathered before these staff join. Using modeling technique, we can compare these variables to the staff “future performance” that already existed today. You can do this modeling technique to each and every job description that you have. Each Talent Analytic model will provide you with a score that will help you identify you next best performer. Voila! Now you have your Talent Analytic Engine.

Disclaimer: This is a personal blog. The opinions expressed here solely represent my own opinion and not those of my employer or any other institution.

Further reading on Talent Analytic:

Deloitte: Oil and gas talent management powered by analytics

TalentAnalytics.com: The best of Talent Analytics yet!

Forbes: Big Data in Human Resources Talent Analytics Comes of Age

SHL Talent Analytics

Bersin HR and Talent Analytics

Tagged with: , , , , , , ,
Posted in Big data, Predictive Model

How To Become a Business Analyst (Data Scientist)

What is a business analyst? Business analyst is someone who can identify the needs for change and ensuring the message is delivered to business units, risk management, IT unit, operation, and eventually to support business to implement the changes. A good business analyst must have a good technical skills, business understanding, good communication and presentation skills.


Today, business analyst must also become a data scientist that possess a deep understanding of advanced statistics and data engineering. Using big data, a business analyst have the ability to analyze and provide solutions: from process improvement, understanding customer behavior (such as customer online shopping experience or customer’s attrition), to pricing optimization.

If you plan to be a business analyst and wondering how to become one, here are some thoughts to get you started:

1. Willingness to learn – Being a business analyst is like being a student of life. You learn new things every day. Even when you don’t know about one thing, people expect you to provide recommendation, a solution and guidance for business to excel. How do you provide recommendation when you don’t have any experience in that area? You learn. Before you do any analysis, walk around, ask people, visit your IT people, sales or operation. Ask them anything, learn from them. Once you have some idea, create your hypothesis and prove it using data. If your hypothesis is wrong start it all over again.

2. Earn business experience – It would be an advantage if you start your career in the business area. Understand what are the main challenges when you run a business: creating a profitable product, obtaining necessary marketing budget, obtaining approval from legal and compliance, making sure that the operation support ready with the sales script and necessary equipment, and everything should done be within limited time period. Plus, you have to meet your ROE. Its a big responsibility. If you can understand the challenges in running in business, you’d become a great business analyst. What if you don’t have business experience? Volunteer. Yes, you can voluntarily ask business to run (or co-run) a project.

3. Ask the right questions – Most of the time, in business, people actually already know what the problems are and they might also already know how to solve it. But most of the time, no one dare to take ask or take action to solve it. A good business analyst need to have the attitude of asking the right questions. The question may be silly at first. For example, you may ask: “Why do we have to go to process A then process B? Why can’t we go to have 1 process called process C?” People may laugh at your question because it sounded silly for them. To them the process is working properly and it has been there for many years. Why change it? Since it is the analyst job to be the agent of change, using big data, you can propose a new process to go straight to process C.

4. Communication skill – Even after you have found the best analysis and on paper it was proven to increase the company’s revenue by 100%, its not always easy to sell your idea to the business. Business people understand the language of business. Thus, if you want to sell your idea, you need to use their language. Scrap too much details in statistics, put them back in your appendix. If you can get their attention in the first 2 minutes, you might be able to sell your idea. Yes, a good business analyst should be a good marketer too.

5. The skill and background – A business analyst can come from any background. You could have engineering degree or history or economics degree. What you need is a strong interest and understanding mathematics and statistics. When I mean strong interest,  you don’t have to have PhD in math or statistics, you should love them. You also need to be able to use statistical tools, from excel spreadsheet, SAS or Hadoop.

Do you think you can become a business analyst? Drop a line below should you have any question.

Tagged with: , , , , , , , , , , ,
Posted in Uncategorized

What I learn from my morning run about achieving a business target

I just started running again after pausing for 1 month. It’s quite tough to get 5km on my first day. I just realized after my 3rd day run of why I couldn’t finish my 5km target on the first day.

I realized that I didn’t start with the end in mind. When I started running, I didn’t really focus to finish the 5K, thus my breathing to the second kilometer is not regulated, I didn’t anticipate of what will happen after the 2nd kilometer; things get tougher, my breathing became heavier. I should’ve started the running with regulated 2 by 2 breathing (2 inhales and 2 exhales) since beginning, even though my body said I didn’t have to. My body said I only need a short 1 by 1 breathing, as if I’m only walking. That was the problem.

In business, we usually set a yearly target between end of Q3 or Q4. Say, we’re aiming to double our revenue from $1million this year to $2million in 2014. This is a huge target. Thus, starting January, we should have set the pace of reaching this 100% increase in revenue by end of year. Each month, we will need to have a new regulated monthly target, weekly target and even daily target. We set the additional marketing budget, new additional staff, etc to meet this goal. This way, we are ‘breathing‘ this target in our daily business activities. What will happen when we just do ‘business as usual’ in January? By end of Q1, we will realized we’ve lost too many opportunities and it will be ready too late to achieve the original goal. We need to start with the end in mind even the journey is still 1000 miles away.

Have you start planning the 2014 business target?

Tagged with: , , ,
Posted in Uncategorized

Who Has The Largest Predictive Data Analytics?

With vast amount of data that currently available, who should be the best one to predict the future? Should it be Google? Facebook? Twitter? Google claim to be able to predict the flu epidemic in the future, and other things. Is it because Google has the biggest data in this world? Or it is not?


If big data can predict the future, then those who have the biggest data are the ones who can better predict what will happened next. If we could only know what are people are doing, thinking, eating, drinking, buying, spending, talking, listening at all times, we should be able to understand the patterns, the repeated behavior across time, thus, predicting what will happen again in similar situation in the future.

These things should be very familiar isn’t it? People really do share what they are doing, thinking, etc at all times. Should it be Facebook to have the best predictive ability? Or twitter? Or is it a more traditional channel like credit card? or perhaps VISA? They should know each and every little thing people buy, spend, where and when.

I think, those who have the biggest data are the ones who make people carry or wear them all the times: smartphones, credit cards, wearable devices (such as google glass, or iwatch). Even right now, the new Google’s Moto X can actively ‘listening’ to whatever the owner says all the time. If only Google can records all the words of what people say, their emotion, what people hear or listen at all times, then the future is on their hands. I’m not saying that Google would do that, but they can if they want to. These things are the first main sources of where big data come from.

Who else can have these predictive ability? I think the telecommunication company should be the one who should have this ability the most. They are the ones who bridge between people and these smart phones or internet companies.

Do telecommunication companies have the largest predictive data analytics?

I googled the information if telco companies have big data and predictive analytics ability. They do. However, most of telco companies only use predictive analytics to improve their traditional key drivers, such as Average Revenue Per User (ARPU), Minutes of Use (MOU) and Churn. I would think that the reason they only use predictive analytics up to this extent is because they care about their (current) customers (needs). Their main income is mostly coming from the customer’s bandwidth usage. Their main investment is building the biggest and fastest communication bandwidth across the world and (hoping) customers are flocking to use these bandwidth at premium price, thus they can get a good return on investment in the long run.

This business model has been proven to provide a great result for telecommunication industries in the past 50 years. However, I think this things will change very soon. Just like a traditional newspaper companies is losing to a more agile and faster internet news company or community news sharing. The telecommunication industry might as well suffer a defeat from community based wifi sharing. They don’t need use the voice line anymore. Even they don’t need to subscribe a data plan anymore. So where should the telco companies earn their income?

In order the adapt to the future, telco industry should adapt their business model. Some of them already did. For example, the biggest Indonesian telecom company PT Telkom Indonesia, since early of year 2000 they already broaden their services to include TV cable, internet portal, multimedia, e-payment, system integration, e-commerce and other content business. They built a new subsidiary company called PT Multimedia Nusantara (Metra) to focus on these telco adjacent industries. This way they could potentially get the biggest consumer data available in Indonesia, the 4th largest population in the world.

Using predictive analytics telecommunication companies should be able to move “intelligently” toward the next revenue generation services by listening to dynamic market requirement to adapt their business model constantly. However, the question is, are they doing it? Are the telecommunication companies are currently able to move “intelligently” toward their next revenue generation services? Are they doing it across their business or still divided in a silo?

If the telecommunication companies utilize the power of predictive analytics to the extreme, then the information of what people are doing, thinking, talking, watching, reading, spending, buying, having breakfast, having holiday, having new born babies, getting married, etc, you name it; they should be able to predict it. Or should they?

What do you think?

Tagged with: , , ,
Posted in Big data, Predictive Model

What I Learn From My Morning Run

If you’re a runner, you’d probably know what I’m going to talk about. There is always this feeling of steady state when I hit the first 3km during my morning runs: your breathing feels like you’re relaxing on a sofa; your vision becomes a crisp clear; your hearing become so acute; and time seems to stop. Usually during this time, my mind goes blank, just focusing on the next stride my legs to move forward. The feeling is so beautiful and makes me feel so alive.

When I looked at wikipedia, a system in a steady state has numerous properties that are unchanging in time. Steady state would be a situation in thermodynamic called equilibrium. It’s in a balance state where the behavior of the system will continue into the future.

This feeling however, couldn’t be achieved if I tried to rush it by running at a faster pace hoping to get into the steady state sooner. If rushed, at kilometer 3, the heartbeat would be too fast, breathing would be heavy and the next 3km would not be so enjoyable.

I think in life, this rule applies. If we think that we can rush-in and push-away to go to the next level at a quicker pace, we might get there faster than anyone else, however, the next level after that would not be so enjoyable. We’d probably get there faster but we might lose what’s the most important in us: family, friendship and self-fulfillment.

I think we should aim to have a balanced-equilibrium life and have a steady state from the inside. No matter what life brings you, your behavior as a person will always be the same and continue unwavering into the future.

When I was a kid, I thought this was given for adult: wise, balance and mature. It turns out that we really need to put the effort, continuously, and cannot be rushed.

Tagged with: , , , , ,
Posted in Uncategorized
%d bloggers like this: