Presentation by piyush jain

pg. 1
Mantra for Innovative Project Management
Piyush Jain
Senior Delivery Manager
Infosys Limited

pg. 2
Effective Talent Management
Predictive Model for Skill Based Forecasting
By
Piyush Jain, Senior Delivery Manager, Infosys Limited
Vinay Prabhu, Delivery Manager, Infosys Limited

Contents
Abstract .................................................................................................................................................................4
Introduction ............................................................................................................................................................4
Context to the paper ...............................................................................................................................................4
Typical Talent Forecasting Model ............................................................................................................................5
Shortcomings of the Typical Talent Forecasting Model ..............................................................................................5
Our approach to forecasting talent needs .................................................................................................................6
Talent skill repository ...........................................................................................................................................6
Skill based Talent Forecasting Model .......................................................................................................................6
Input Parameters (for each skill) ...........................................................................................................................7
Derived Parameters (for each skill) .......................................................................................................................8
Working of the Model ..............................................................................................................................................8
Observations ..........................................................................................................................................................9
Assumptions & Scope for further development ....................................................................................................... 10
Conclusion ........................................................................................................................................................... 10
References .......................................................................................................................................................... 10
Acknowledgements .............................................................................................................................................. 10
About the Authors ................................................................................................................................................. 11
pg. 3
Note: All data shown in this paper is simulated. Actual data has not been used due to confidentiality reasons.

pg. 4
Abstract
The world of today is a fast place. Patience is no more a virtue but a bane. Longevity is now measured in quarters and
not years. Clients are scrambling to appease customers by trying to release new products, new software versions,
more features, software upgrades, release patches and so on almost on a quarterly basis. The pace is relentless and
its consequences are being felt across project management functions.
One function that organizations have to rapidly focus is the talent management function. As project timelines get
crunched and clients demand higher productivity with fast ramp ups, there is constant flux in terms of people demands
for staffing project engagements. Availability of right people with the right skills at the right time is often the di fference
between project success and failure. In light of this it is extremely important for IT services companies to ensure they
have a good model for talent forecasting leading to right sourcing and optimal utilization.
Almost all companies have some model or the other that is used for forecasting talent demand leading to talent
acquisition. However, most models tend to focus only on future demands to arrive at absolute acquisition numbers. In
our opinion this is a sub optimal model. In this paper we pres ent a skill based talent forecasting model that would help
predict the skill utilizations more accurately. We believe this model would assist talent managers in managing the talent
pools more efficiently, thus optimizing their talent acquisition costs and ensuring optimum utilization levels.
Introduction
Talent Management has always been a key function for any enterprise. For a human resource intensive industry like IT
services, its importance is magni fied many times more. Talent Management refers to the anticipation of required
human capital for an organization and the planning to meet those needs. It is the science of using strategic HR to
improve business value and to make it possible for companies to reach thei r goals. Everything done to recruit, retain,
develop, reward and make people perform forms a part of talent management as well as strategic workforce
planning[1].
The challenges in the current business scenario have precipitated the need for strategic and innovative approaches in
talent management for the purpose of achieving business objectives and gaining competitive advantage.The cycle of
workforce planning includes filling resource requests, analysing resource utilization, forecasting capacity, managing
and identifying the human resources to fill that capacity, and then restarting the cycle[2].
The scope of this paper is limited to the forecasting aspect of workforce planning. Through this paper we will explore
how skill plays a crucial role in forecasting of talent requi rement. We take the opportunity to present a predictive model
that considers skill attributes for talent forecasting and how that would help in ensuring the right focus on optimal talent
utilization and better talent sourcing strategies.
Context to the paper
The paper focuses on our experience of deploying skill based forecasting method for workforce planning. We take this
opportunity to share how bringing in detailed skill view in talent forecasting led to better clarity and purpose in
workforce planning leading to higher efficiencies in talent development and deployment.

Each organization has their own method/approach to talent forecasting and it is dependent on the context surrounding
their business. Our attempt here is not to dictate a particular model or methodology. The purpose of this paper is solely
to bring the focus on why skill view is important in talent requi rement forecasting and how by doing that y ou can
achieve more desirable results both in sourcing and in utilization.
pg. 5
Typical Talent Forecasting Model
At a broad level traditional forecasting model for talent requi rements focuses on future demands and current attri tion
levels to determine the shortfall in talent needs. The demands do encapsulate the skill requirements under them, but
the focus is more on the overall number of talents requi red to bridge the attrition short fall and also address the growth
needs projected as demands.
The table below provides a view of one such typical model.
Legend Parameter Formula
Current
Q
A Current Total Talent Strength 10000
B Current Utilization (%)** 79
B1 Implies no. of people on production work B1 = A*B/100 7900
C QoQ expected growth (%)*** 2
C1
Implies expected people in production in next
quarter C1=B1 + (B1*C/100) 8058
D Target utilization % for next quarter 80
E Projected Total Talent Strength by next quarter E=C1/D * 100 10073
F Current Attrition % 3.5
F1 Implies talent shortfall due to attrition F1=A*F/100 350
G Gross Talent Shortfall for the quarter G=F1+(E-A) 423
H Projected Trainees to join in the quarter 125
I
Net Talent Shortfall for the quarter that needs to be
sourced I=G-H 298
Green Cells indicate input parameters to the model. Amber Cells indicate derived values based on input parameters
** Utilization is def ined as people w ho are on production projects being billed for their services.
*** grow th is assumed to be linear in terms of number of people billed in production
Data shown in above model is simulated
Table 1 – Traditional Talent Forecasting Model
The said model relies on inputs like current and expected utilization level, current attrition level and projected l iner
growth in terms of manpower growth to arrive at the overall talent requirements
Shortcomings of the Typical Talent Forecasting Model
The typical model for forecasting as shown above is good for overall projections. It relies on the macro level inputs
around utilization, attrition and growth projections to forecast the net talent requirements. However, it is laced with
shortcomings that can lead to sub-optimal results in getting the right talent.
The foremost shortcoming of the typical model is that it subsumes the skill view under the macro level growth
projections. This skill view is mostly based on skill requirements captured in the form of talent demands. Demands can

be for incremental growth in existing programs, attrition replacement or for completely new pro grams. However, it goes
without saying that talent demands tend to be rather liberal in terms of requirements. What is not explicitly captured
and known in this model is the current utilization of these skills, the impact of ramp downs if any on the utilization and
how that impacts the availability of the skill pool to meet the projected demands . Going only by the demand view and
ignoring the utilization view can lead to a skewed view of skill requirements, which can lead to mismatch between what
was required and what got sourced and thus affecting the effectiveness of the overall talent utilization.
Hence, while we totally agree that a typical forecasting model gives a good view on the overall projected talent
requirements, it needs to be supplemented by a model that explicitly captures the skill based utilization view, leading to
better workforce planning and more accurate skill based talent projections.
Our approach to forecasting talent needs
We have considered a unit / division of the organization as the base for explaining the skill based forecasting model.
This is not to say that it is restricted only in its application for a unit / division. The application of the model whether for a
unit / division or the entire organization is entirely a prerogative of how the organization approaches workforce planning
and management. Organizations that choose to de-centralize workforce planning and management can have this
model adopted and adapted at each unit / division level.
We would also like to state that the frequency of forecast we have assumed here to be on a quarterly basis, where in
the skill requirements of a said quarter are done one quarter in advance. This approach is mainly adopted keeping the
sourcing and recruitment timelines in mind. If an organization follows hal f yearly or annual forecasting cycle, then the
model can be suitably adapted for the same.
Before we get into the details of the model, the functioning of this model is dependent on having a robust system that
captures the skill details for each employee in the organization and follows the discipline of keeping it upto date at all
times.
Talent skill repository
A central repository that captures the skill details for every employee in the organization.
As any organization that deals with a wide variety of skills, the system designed needs to provide flexibility in
categorizing the skills according to domains, technology, lifecycle stages and management levels. This needs to be
overlaid with experience levels categorized in terms of proficiency metrics.
The accuracy of the forecast for a skill is completely dependent on integrity and accuracy of data in the system at all
times.
Skill based Talent Forecasting Model
The skill based talent forecasting model is a predictive modelbuilt around utilization views of each skill set at the start of
current quarter and its translation to projection of skill requi rements for next quarter. The utilization view of skill set
taken at the start of the current quarter is considered the most reliable view considering it represents the actual posi tion
of utilization at the end of previous quarter which has not been affected by production / non production movements in
the current quarter. Attempting to use the model with skill utilization data captured any other time in the current quarter
pg. 6

can lead to erroneous results due to daily variations in production / non production status of the skill sets. Hence, we
recommend freezing the snapshot of utilization data taken at the end of previous quarter and use that as a baseline for
forecasting skill requirements for the next quarter.
[Illustration – If current quarter is Q2, then consider the skill utilization snapshot view taken at the end of Q1 for forecasting skill
requirements for Q3]
Following is a template of skill based forecasting model.
pg. 7
Table 2 – Skill Based Talent Forecasting Template
The template is built around a view of skills that you want to forecast for talent requirements. Using the skill
classification defined in the skill repository for employees, you create each row for a particular skill set. The model can
have as many skill rows as you wish depending on the focus you want to give to skillsets that have high utilization, you
see high demand and therefore you need to forecast against them. As a suggestion, we advise having the model focus
on top 10-15 skill sets basis of thei r demand and utilization. The model requi res certain parameters to be input for it to
derive the forecast numbers for each skill set.
In the model template above, cells marked in Green are for the input parameters and cells marked in Amber are
derived from the input parameters.
Input Parameters (for each skill)
 Current Actual Production Head Count
o Implies total current people who are doing production work and are being billed for their services
 Current Actual Total Head Count
o Implies total number of people having this particular skill set.
 Current allotted trainees
o Implies the number of trainees allotted for the unit and expected to join in current quarter
 Expected lateral joins
o Implies the laterals who have accepted offers and likely to join in the current quarter
 Projected Attrition
o Implies number of people on notice period

pg. 8
 Overall Growth % for next quarter
o Implies the expected growth through linear increase in manpower in the next quarter
 Desired Skill Utilization % for the quarter
o Implies the target skill utilization you want to maintain for the particular skill set for the current quarter.
This is dependent on the demand that you are seeing for the said skill set to get deployed in the
current quarter.
Derived Parameters (for each skill)
 Utilization %
o Implies utilization snapshot of the skill as on end of previous quarter.
 Derived as C = A/B * 100
 Expected Total Head Count for the current quarter
o Implies total HC derived for the particular skill set after adding the talent additions (trainees + laterals)
and minus the projected attrition numbers
 Derived as G= B + D + E - F
 Projected HC x% growth
o Implies the anticipated head count in production for the projected x% growth in manpower terms
 Derived as H = A+(A*N/100)
 Skill Util % at x% growth
o Implies the projected utilization of the particular skill set based on the projected increase in production
head count
 Derived as I = H / G * 100
 Projected HC required for next quarter
o Implies the projected head count for the particular skill set . This is based on two factors. First is the
increase in production head count that you anticipate for the skill set based on the growth % projection
provided and secondly the desired utilization you want to maintain for this skill set so as to not over
leverage the talent pool for this skill set.
 Derived as K = H / J * 100
 Net Shortfall / Excess
o Implies the net forecast in either excess or shortfall for the particular skill set
 Derived as L = K - G
Working of the Model
Now let us look at the working of the model using simulated data (actual data has not been used due to confidentiality reasons)
Shown below is the model template populated with simulated data. We have kept the data same as that used for
typical model (shown in Table 1)to explain how this model provides a more realistic view on talent requirements that the
unit needs to forecast and plan for the next quarter.

pg. 9
Table 3 – Skill Based Talent Forecast Model with simulated data
As indicated above, once the input data is provided for each skill in the amber cells, the model goes through various
intermediate calculations to finally arrive at the forecasted numbers for the said skills in Col L.
Some key aspects about the working of the model are as follows:-
1. Utilization level for each skill is captured separately and it provides a view on which skills are being heavily
utilized implying higher demand and which skills are exhibiting lower demand and thus, lower utilization
2. Projected joiners + attrition for each skill provides a clearer view on the available current talent pool for the
said skill
3. Projected skill utilizationconsidering growth numbers (as given in col I)provides a view on how the skill
utilization pattern would look like i f continue at the same rate of deployment for the said skill. This is an
important decision making point fordeterminingthe desired utilization level for the said skill, which would
help to determine the precise numbers needed to maintain an adequate supply for the said skill
4. The net forecast requirements for the skill (col L)can either be a short fall in number or excess depending
on their current and future utilization levels. What this implies is that the skills that have a shortfall need to
be sourced to improve their supply, whereas skills that have excess need to be focused on generating
demand for improving their utilization.
5. By ignoring the skills that have an excess number, one can tweak the numbers in col M (with the projected
shortfall / excess requirements (in col N) as the basis) to decide on the number of new joiners for each of
the skills.
Observations
This model has been effectively used over multiple quarters and fine-tuned during the course of its usage. Some of the
benefits and observations are:
 By tracking the utilization at a skill-level, the average and maximum utilization levels for a particular skill
(for the last 6 months) gave an indication of the utilization levels that the Unit can possibly operate on.

 People with skills that were not in demand were cross-trained to meet the demand of over -utilized skills.
pg. 10
This helped train the people on time and achieve the business demands of the Unit.
 Certain niche skill sets that were found to be on the excess side, were brought to the focus of the sales
team to help them work on generating business demands suiting the said skill sets, thus, improving their
net utilization.
 With focused hiring, the time taken to put a new joinee into Production reduced by a substantial margin.
Assumptions & Scope for further development
The skill based talent forecasting model presented in this paper provides a better and more accurate view on workforce
planning and talent requirements. Albeit, the model presented assumes that the lineargrowth % in terms of manpower
requirements applies uniformly across all skill levels, we plan to address this in the next version of the model .The
model going forward should also consider the experience levels of the talent and the location aspects of the
availability/demand.
We strongly believe that through this extension the model can really provide a near accurate view on skill forecasting
that can go a long way in the workforce planning for the future.
Conclusion
The Skill-based Forecasting Model aligns the Talent Management strategies with the business goals of the Unit. As
companies increase the focus on improving productivity and efficiency, it will be of paramount importance to provide
the people with the right skills at the right time and the right place. To ensure that this is possible Workforce Planning
through the Skill-based Forecasting Model will play an important role.
References
[1] Carpenter, Mason, Talya Bauer, and Berrin Erdogan. Management and Organizational Behaviour. 1. 1.
FlatworldKnowledge, 409. Print.
[2] Rudolf Melik. "Rise of the Project Workforce, Chapter 9: Workforce Planning". PM Hut. Retrieved July 9, 2010.
Acknowledgements
The authors would like to sincerely thank Mr. Nagabhushan a Samaga for his invaluable inputs, Ms. Anju Chawla
Takkar for doing a thorough proof reading and review of the paper and Mr. Manohar Atreya for his guidance and
encouragement.

pg. 11
About the Authors
Piyush Jain is a Senior Delivery Manager with Engineering Services at Infosys Limited, Bangalore. As part of his
current role, he heads the engineering business for a large telecom client and additionally holds the Unit PMO
responsibilities. Prior to this, he was the head of Talent management function at engineering services.
He has 20+ years of industry experience most of which is in the software engineering and Telecommunication space.
He has been instrumental in incubating and establishing large offshore development centers for engineering clients
across multiple geographic locations.
He is a certified PMP and has published and presented papers in the prestigious PMI conferences, forums and other
technology journals.
He is a graduate in computer engineering from S.V. NIT, Surat (formerly R.E.C Surat ), Gujarat. He is a sports
enthusiast and an avid reader with specific interest in current affairs and how it impacts business paradigms.
Vinay Prabhu works as a Delivery Manager in Engineering Services group of Infosys Limited, Hyderabad. He manages
the delivery of Engineering R&D programs for global clients in the Healthcare and Li fe Sciences segment. He has
managed complex Product Engineering programs for clients across geographies and industrial segments.
He is a graduate in Computer Engineering from M.S.R.I.T, Bangalore and has about 18 years of experience in
delivering projects in the Product Engineering space.
He is passionate about technology and people engagement related activities. He is an avid reader, an F1 enthusiast,
loves quizzing and spending his free time with family and friends.

Presentation by piyush jain

More Related Content

What's hot (20)

Similar to Presentation by piyush jain (20)

More from PMI_IREP_TP (10)

Presentation by piyush jain