Role of double AA- s in TCO Optimization - Framework for Value …documents.grenadine.co/PMI Bangalore India Chapter/PMPC... · 2017. 7. 18. · Role of double AA- s in TCO Optimization

Role of double AA- s in TCO Optimization - "Framework

for Value improvement"

Vision of future technologies - Optimizing Total Cost of Technology

Ownership

PMIBC-17-2-003

Project Management Practitioners’ Conference 2017

www.pmibanga lorechapter .o rg

Page 2

CONTENTS

ABSTRACT .................................................................................................................................................................. 3

AUTOMATION JOINS ANALYTICS AS A FORCE MULTIPLIER ............................................................................... 3

LIMITATIONS OF TRADITIONAL PM TECHNIQUES ................................................................................................ 4

CHANGING DYNAMICS - SHIFT FROM DISCREET TO CONTINUOUS MANAGEMENT IN PMO ........................ 6

THE 5C FRAMEWORK OF PROJECT MANAGEMENT – AA ENABLED ................................................................. 6

UNDERSTANDING THE FIVE PHASES OF “5C FRAMEWORK” ............................................................................. 8

CRITICAL SUCCESS FACTORS: ............................................................................................................................. 13

KEY BENEFITS ......................................................................................................................................................... 13

CONCLUSION ........................................................................................................................................................... 14

REFERENCES .......................................................................................................................................................... 14



Page 3

ABSTRACT

As organizations accelerate the pace of modernizing and transforming their IT footprint by adopting digital

technologies at a scorching pace, the IT functions are transforming into projectized organizations, running several

hundred projects at the same time.

Further, the rate of adopting new versions of currently deployed digital technologies is rapidly increasing as visible

cost of latching on to newer versions becomes near zero in the cloud era, with several projects being undertaken

just because they appear “free”.

More importantly, with IT getting fully integrated in business processes, traditional models for TCO computation for

IT are no longer valid, as several more business impacting variables need to be taken into account, making it almost

impossible to truly determine IT TCO in age old ways.

This paper recommends a new framework for using Automation and Analytics (AA) in such a “projectized

organization” covering three critical areas:

• How can AA help in computing TCO, by incorporating impact of parameters such as revenues and customer

satisfaction?

• How can AA help tame technology change management tiger by identifying TCO reduction projects in areas

like leakage due to cloud sprawls, optimization opportunities due to growing global public cloud?

• How can AA help Project life cycle management as IT departments are on an over drive to identify, define,

design, deploy and continuously modernize technologies – providing early warning signals on failure points

and increase success rates?

AUTOMATION JOINS ANALYTICS AS A FORCE MULTIPLIER

For more than a decade we have known about three types of analytics: descriptive, predictive, and prescriptive which

include machine learning, artificial intelligence (AI) etc. This trinity of analytics variations has worked well for most

organizations and today they have become integral to the way more mature organizations run their operations. A

key change today though is that

analytics is now increasingly

combined with automation. What

this means is instead of presenting

a recommendation to a human, as in

legacy analytics, automation directly

takes action on the results of the

analytics. For example, predict the

risk in the project proactively and

intelligently change the project

execution variable like change in



Page 4

critical path or addition or inclusion of new project activities. Also, the automation is not a new phenomenon, and

questions about its promise and effectiveness have long been accompanied by its advances. A survey from Narrative

Science in partnership with National Business Research Institute [1] reports the adaption rate of automation &

analytics (Fig. 1).

Leveraging Automation and analytics in today’s era is no more a technology implementation project, but is a business

imperative with the necessary support of right implementation. On one hand, automation is about providing high

quality and faster turnaround to clients while on the other end analytics is the systematic quantitative analysis of any

kind of data to derive meaningful conclusions and aid decision making. Hence, we can call analytics as method and

automation as an enabler to actually make it happen.

LIMITATIONS OF TRADITIONAL PM TECHNIQUES

Today's digital businesses demands IT agility, and agility demands a new breed of digitally minded, technology-

enabled enterprises with new breed of project management methodologies. Empowered customers, coupled with

increasing economic, product, and market change, force organizations to become customer-focused, outcome-

oriented business partners who must think about technology systems as cost / business drivers as well as revenue

catalysts.

This leads organizations today to revamp the technologies and in next few years there will be an entirely different IT

landscape compared to today. For example, most of the infra or apps will be on cloud, world will move towards being

100% on mobility etc. To achieve the same, organizations are running hundreds of technology projects in rapidly

evolving technology areas and this requires a different set of project management methodologies altogether.



Page 5

This is well collaborated by a recent survey by a well-known consulting firm [2] indicates that (Fig. 2) the traditional

methodology of project management has started to hit road blocks in this world of fast paced technology

transformation. So, what is causing this high failure rates of technology transformation projects? Here are some

pointers:

As organizations move to leverage cloud based technology solutions aggressively, new version or new capability is

available on the tap and appears that it can be is deployed for free, and hence the users tend to adopt them even if

they are not necessarily

required and have an

established business

benefit. This leads to

some “junk projects”

being taken up and a

strong prioritization is

required. Similarly, as

organizations deploy

consumer facing

applications, revenue

and user experience

impact become critical

parameters in evaluating

technology TCO in

addition the existing

known parameters. As a

result, performance,

availability and risk of

technology is equal to risk of business thereby necessitating a different approach to computing project ROI.

Lastly, with number of projects in technology transformation increasing many fold, average project duration shrinking

and technology budgets remaining broadly unchanged, the pressure to deliver lot more at same cost has gone up

manifold. All this demands a new mind set and new set of tools and technologies to be deployed in PMO. And this

is where AA step in. (Fig. 3).



Page 6

CHANGING DYNAMICS - SHIFT FROM DISCREET TO CONTINUOUS

MANAGEMENT IN PMO

At the heart of this struggle of Project Managers is the explosion in data points that need to be correlated, reduction

in reaction times available to

take corrective actions and

inability to estimate impact of

project delays on future

revenues and user

experience (Fig. 4).

Today organizations run more

than 1000+ project running in

parallel, with each project

consisting of 100+ tasks and

more than 10 types of

resources – and status of

each of them changing every

week or even faster with new technology project durations shrinking to 6-8 weeks. As a result, Project Management

needs to move from being discrete to Continuous – almost real time.

THE 5C FRAMEWORK OF PROJECT MANAGEMENT – AA ENABLED

The paper proposes a

“5 C Framework “to

address this – where

Continuous

management is the core

in all the stages of

Project Management –

from Planning to

Evaluation. Automation

and analytics then get

embedded in each

stage of the framework



Page 7

– as it is not possible to understand, prioritize and balance costs/value/risks/schedule of all the projects in any other

approach. The typical phases of “5 C Framework “is shown in Fig. 5.

This Framework proposes a 5 phase approach which comes with the ability to continually identify, prioritize and

execute projects from a holistic vantage point. The framework also moves organizations from merely project status

reporting on siloed data to true project analytics, by consolidating and organizing millions of data points compared

to few thousand data points in traditional methodology across all projects

The data generated - both ground-level to the top level data such as time sheets, budget records, plans and

schedules – becomes an input to the Automation and Analytics engine to answer questions such as

• What are the projects that need to be deprioritized considering new sets of resource constraints that have

come up?

• How quickly do we need to onboard new vendors and with what niche skills considering 7-10 new projects

that CTO has launched and wants to put on fast track, knowing fully well that these may be more expensive

than existing vendors?

• Is it best to abort some projects as the total ask from business is much higher than what PMO can handle

or can we scale rapidly by leveraging AA?

This models helps organization to transform their Project Management Lifecycle to break down organizational silos,

automate most of the PM tasks, streamline project deliveries and drive agility into the overall lifecycle by leveraging

AA.

The 5C model

is directed to

focus on

delivering the

best possible

outcome by

leveraging AA

in all major

project phases

as shown in

Fig. 6. This is

just not

possible by

managing

individual projects in traditional methodology which tends to remain more tactical.



Page 8

UNDERSTANDING THE FIVE PHASES OF “5C FRAMEWORK”

Continuous Planning: The traditional project management methodologies are not architected to support the agility

required to meet dynamic project demands. Hence, it is required to eliminate the bottleneck of sequential project

management phases. This is where the “Continuous Planning” phase of 5C model plays a vital role. In this phase,

data of all the past projects irrespective of domain or department is fed in to the analytics model. This analytics model

based on the past data is converted into a simulation model for a particular organization. Data of a new project is

passed through this simulation

model to understand the

accuracy of planning and

identify bottlenecks in planning

areas of budget consumption,

resource optimization etc. The

conceptual model is shown in

Fig. 7

Key building blocks

• Collection of all the past data of projects across the organization

• Inclusion of both structured as well as unstructured data

• Continuous inputs from the current projects where project managers are updating the data (PM tools,

collaboration tools, excels etc.)

Role of Automation and Analytics

• Applying analytics techniques – Pattern Recognition, Gap Analysis and Historical Trending (Time Series)

• Model standard algorithms based on the used techniques

• Build the Simulation Model to accept the data

Sample variables/data inputs to be considered.

• Number of projects

• Start and End date of projects

• Status reports with highlighted misses and on-time delivery

• Number of resources involved and their types – technical, functional or any other

Note: These are some key inputs however; a lot more can considered to build the accurate simulation model.



Page 9

Continuous Prioritization: Once the execution plan for the multiple projects is in place it is now important to classify

the projects into multiple

categories and choose

the ones that really

matter. Further, since

new projects keep

getting added every

week and some get over,

there is a need to

reallocate resources in a

real time. The conceptual

model is shown in Fig. 8.

Key building blocks

• Collection of all the master project data from the “Continuous Planning” phase

• Define the categories for the bucketing or classification of projects


• Applying analytics techniques – Data Clustering and Classification and Historical Trending (Time Series)


• Automatically define the project classification or bucketing based on the data

• Map the existing standard bucketing defined in this phase with the automatically defined bucketing


• All the master data from “Continuous Planning” phase


• Project categories based on different types – technology, strategic etc.

• Organization goals and priorities as per their business requirements, including new projects that get added

Note: These are some key inputs however; a lot more can considered to build the accurate analytical model.

Continuous Optimization: Once the right set of projects with accurate planning is identified the next step is to see

what kind of resources are required to run the projects and how these resources can be optimized without affecting

the delivery quality, budgets and project success factor. This phase identifies the resources used across the projects

and can be optimally shared across various projects. This not only helps keeping the budget under control but also

helps in bringing the right expertise so that ultimate quality of delivery is improved.



Page 10

In this phase the master data and project classifications are most important inputs from the earlier. With the help of

these inputs the common set of resources, their performance trends and utilization factor are identified based on

their involvement in the past

projects. The resources are

not only the human resources

but also infrastructure

resources, financial

resources or any other

resource required to run the

projects. The conceptual

model is show in Fig. 9. The

sample outcome shown in Table 1 is considering the resource utilization parameters however, unit of utilization and

related parameters can be changed as per the PMO standards of organization. This outcome normally does not vary

from one organization to other.

Key building blocks

• Collection of all the master project data from the “Continuous Planning” and “Continuous Prioritization” phase

• Define the resource utilization metrics that needs to be measured

• Define the types of resources – Financial, Human, Infrastructure, Technology etc.


• Applying analytics techniques – Event Correlation, Supervised Learning and Trends (Time Series)

• Model standard

algorithms based on the used

techniques

• Automatically define the

areas of resource

optimization by giving the

exact insights

• Project outcome

dashboard for resource

utilization

Sample variables/data inputs

to be considered.



Page 11

• All the master data from “Continuous Planning” and “Continuous Prioritization” phase


• Status reports for all the project resource utilization across all types of projects


Continuous Intelligence: The way the intelligence for executing the project is built in traditional project management

is more reactive however, to get today’s complex projects running successfully, a proactive model needs to be put

in place. What it means is that the key stakeholders should be informed proactively on the project risks in the areas

such as budget overruns, under/over utilization of resources, external factor impacts etc. This is where the phase of

“Continuous

Intelligence” plays a key

role. The conceptual

model is shown in Fig.

10

Most the technology

projects are dynamic

wherein change in one

variable can impact

more than 100 variables, thereby impacting overall project timeline and budget. Hence, it is important to have

automatic alerts sent via different mediums like email, SMS or mobile push notifications to alert project sponsors and

managers at the right point of time to keep the show running. This kind of intelligent alerting is called as “Industrial

Intelligence”. This “Industrial Intelligence” uses the analytics domains like Artificial Intelligence(AI) & Machine

Learning (ML) to accurately predict the good or bad state of the project proactively. This automated alerts signal any

anomalies and with the help of corrective actions defined in the automation workflow, automatically solves the issues

Key building blocks

• Collection of all the master project data from all the subsequent phases

• Define the thresholds on specific data points of projects


• Applying analytics techniques – Artificial Intelligence, Machine Learning, Supervised Learning, Regression

and Historical Trends


• Automatically define the dynamic thresholds and improve the alerting

• Predictive Alerting with automated actions wherever applicable



Page 12


• All the master data from all the subsequent phases


• Key details of the stakeholder names and other demographic details


Continuous Evaluation: This phase is the most important phase since it is spread across all other phases

mentioned above. This phase primarily deals with the continuous evaluation and feedback to each of the 5C model

phases, by identifying opportunities for improvement and by measuring impact of improvement efforts at each of the

project phases. This phase ensures that all the project stakeholders understand the opportunities for improvement

and their responsibilities by giving the individual feedback at the resource level or project phase level. These key

insights are again provided by analytics model uses in the subsequent phases and then automate the actions

wherever applicable. This

phase makes both analytics and

automation to the core by

triggering various dashboards,

that bring in data from millions of

project data points to showcase

the required key indicators.

These dashboards are

actionable dashboards and will

alert in case of any major issues or thresholds are breached. The conceptual model is show in Fig. 11

Key building blocks

• Collection of all the master project data from all the subsequent phases

• Define the right point of feedbacks on specific data points of projects


• Applying analytics techniques – AI & Machine Learning, Pattern Recognition and Historical Trends



• All the master data from all the subsequent phases

• Status reports from each of the phases with highlighted misses and on-time delivery




Page 13

CRITICAL SUCCESS FACTORS:

The key areas of focus (Strategic Success Factors) are defined to identify areas where organization must be ahead

with its project management methodologies to achieve higher success rate

KEY BENEFITS

Analysis of millions of data points

provides various benefits by

providing real-time KPIs to

measure achievement of project

analytics implementation. The

model should be cascaded down

to an organization to measure its

achievement at both enterprise-

wide and business unit levels. Fig.

13 explains the benefits of the

framework.



Page 14

CONCLUSION

As evident from the above, analytics and automation can help organizations take more informed decisions thereby

keeping the projects on-schedule and on-budget. In today's complex project management environment with large

number of data points and need for real-time analysis, it is essential to make use of the analytics and automation

tools to manage project data overload, assess feasibility, analyze project portfolio for project selection, classification

and prioritization, initiate automated alerting & actions, improve project stakeholder management, enhance data

visibility and control, project focused dashboards, manage projects risks by avoiding project schedule delays and

cost overruns and finally to improve the overall management of project lifecycle.

It has also been seen in the critical success factors that to implement such a framework there is major need of skill

re-alignment and cultural change. What it also necessitates is a major change in existing project management

standards & methodologies and demands a change in the overall structure of the PMO, making it more horizontal

compared to traditional vertical PMO structures.

REFERENCES

[1] NarrativeScience, “Outlook on Artificial Intelligence in the Enterprise 2016”

[2] Deloitte, “Predictive Project Analytics”, https://www2.deloitte.com/content/dam/Deloitte/ca/Documents/risk/ca-

en-ers-predictive-project-analytics.pdf

[3] Harjit Singh, “Project Management Analytics – A Data Driven Approach to Making Rational and Effective

Project Decisions”

[4] Vishal Sanghi, “PM 3.0, Lets welcome the Outsourced PM!”, PMPC 2014 Bangalore, 20-22 November 2014

[5] Neil Chandler, Bill Hostmann, Nigel Rayner, Gareth Hersche, “Gartner's Business Analytics Framework”, 20

September 2011

From a real experience it has been established that project success rate improves significantly – failure rates

dropping by almost 50% compared to the data presented in Fig 2 above.

https://www2.deloitte.com/content/dam/Deloitte/ca/Documents/risk/ca-en-ers-predictive-project-analytics.pdf

https://www2.deloitte.com/content/dam/Deloitte/ca/Documents/risk/ca-en-ers-predictive-project-analytics.pdf