2. Data science for results.
Machine Intelligence partners with management teams and
investors to accelerate and scale data science innovation.
We guide product innovation, engineer machine learning
models and software, and build enduring capabilities.
We’re creative, rigorous, and efficient. We bring the
sophistication of a large strategy firm with the speed and
value of a focused boutique.
3. Managing Risk for Large IT Projects
Research and experience shows the biggest risk
indicators for large projects are:
• Length of time
• Business goal clarity
• Estimating time and budget
We can mitigate these risks through:
• Short cycle, iterative delivery
• Structured, inclusive business plan development
• Initial pilots and mockups to inform estimates
Sample Execution Risk Dashboard
4. Project Structure to Contain Risks
Business plan
Operate and improve
Deploy
Goals definition
Design
• Business process
• Product functionality
• Technical architecture
• Organization and capabilities
Business goals are achieved by
more than code. A project must
integrate change to processes,
organizations, and products.
This approach links business goals
to functionality, technical design,
coding and testing, and then IT
operations.
Confirm and stabilize goals
Articulate goals,
functionality, governance
Quantify economic value
Visual mockups to
accelerate design
Pilots inform time and
budget estimates
Align incentives
Evolve agile, QA, UX
capabilities
Automate, rehearse, and
run parallel
Track business plan metrics
Stability, scalability, cost
kaizen
2016 Budget and Delivery Plan
Business strategy
Build organization and
capabilities Build and
test
Mockups
and
pilots
Change
leadership
Governance
& engagement
5. Linking Strategy to Execution and Results
The methodology provides for
transparent and rigorous traceability.
Each business goals is linked to use
cases and product functionality.
Each business case has
corresponding metrics, baselines,
and targets.
Governance processes capture
production metrics and compare to
the business plan.
Example
• Strategy to respond to margin compression by reducing costs,
and grow by serving new customer segments.
• One business goal to reduce time and cost to configure core
application for new clients.
• Business plan sets a baseline starting points for metrics, and
defines a 50,000 hour / 25 FTE savings generating $5,000,000.
• Use case developed by interviewing and observing users.
• UX mockup developed collaboratively to streamline workflow.
The mockup is then wired to the flexible instrument model pilot.
• 2016 Budget based on mockup and pilot sets funding and
delivery date for October 2016.
• Business metrics capture savings, and compared to business plan
as part of governance.
• IT operations show latency is near instant with no unplanned
downtime.
Metrics
Test case
Technical design
Governance and
transparency
Application
functionality
Code component
Business plan
Business goals
Strategy
Process use case
6. Business Strategy
Business plan
Operate and improve
Deploy
Goals definition
Design
• Business process
• Product functionality
• Technical architecture
• Organization and capabilities
2016 Budget and Delivery Plan
Business strategy
Build organization and
capabilities
Business strategy is the starting point for most IT projects. This is a succinct statement of the
intended outcome for technology action. This should be generally understood prior to moving
to goals definition, but does not need to be precise.
The level of detail could be at:
• Respond to margin compression by reducing costs.
• Achieve growth by reaching new categories of customers.
• Expand into new geographic markets by localizing and extending applications.
7. Goals Definition
Business plan
Operate and improve
Deploy
Goals definition
Design
• Business process
• Product functionality
• Technical architecture
• Organization and capabilities
2016 Budget and Delivery Plan
Build organization and
capabilities
Goals translate business strategy into specific implementation. Goals include a measurable target
and date.
Example goals could include:
Costs
• Shift x% of headcount to low cost shared service centers by end of 2017.
• Reduce IT operations costs by 10% each year from 2017-2020.
• Sunset legacy system X by the end of 2019.
New Customer Segments
• Enable new customers to be configured on system X in a single day.
Business strategy
8. Business Plan
Business plan
Operate and improve
Deploy
Goals definition
Design
• Business process
• Product functionality
• Technical architecture
• Organization and capabilities
2016 Budget and Delivery Plan
Build organization and
capabilities
The business plan methodically integrates why, how, and when to achieve the strategy.
The business plan should be as limited as possible in scope and time. The business plan is a
confirmation to the organization of alignment across stakeholders.
Components include:
• Definitive and succinct statement of business strategy and goals.
• External analysis of competitive environment, client trends, and regulatory mandates.
• Internal environment analysis of system and alliances necessary to execute.
• Economic value generated by goals with clear and rigorous assumptions.
• Governance and engagement model.
Application screen mockups and technical pilots should proceed in parallel with business plan
development. Mockups and pilots produce learning that informs budget and time estimates.
Business strategy
9. Design
Business plan
Operate and improve
Deploy
Goals definition
Design
• Business process
• Product functionality
• Technical architecture
• Organization and capabilities
2016 Budget and Delivery Plan
Business strategy
Build organization and
capabilities
Design specifies how to implement the business plan.
Business process design drives use cases. We will iterate visual mockups, rather than static
documents, to confirm business design. Final mockups go to production: “code talks”.
Product functionality begins with analyzing incumbent competitors and emerging new
entrants. Functionality is designed to implement streamlines processes. Functionality includes
elegant user experience including a thoughtful and simply sequenced approach to change.
Technical architecture is a central decision. This stream can start earlier, at project inception.
We often start by examining leading approaches to similar needs in other industries.
Architecture will be confirmed iteratively through pilots.
Organization design includes deciding sourcing, balancing variable costs for surge capacity
with fixed costs, team structure, talent development, and aligning incentives.
Capabilities begins with a frank assessment of collective team abilities and talent. Large scale
projects impose new demands on organizations. This phase focuses on proven key needs for
large projects: project leadership, agile development, QA and test case management, and UX
design. As needed, we may assess data quality, controls, compliance, or security.
Capabilities extend beyond people to include technical environments. Specifically, we will
ensure that test environments match production and include appropriate data. We will target a
production environment allowing low-risk parallel deployments. The team will work with the
finance function to plan and control investments.
10. Budget and Delivery Plan
Business plan
Operate and improve
Deploy
Goals definition
Design
• Business process
• Product functionality
• Technical architecture
• Organization and capabilities
2016 Budget and Delivery Plan
Business strategy
Build organization and
capabilities
The budget builds on design and-importantly-learning from mockups and pilots to commit to
funding, business outcomes, and delivery dates.
The budget and delivery plan includes:
• Funding requirements developed by tracing business goals to process to functionality, and
to develop capabilities
• Date milestones within a consistent tempo of iterations, e.g. functionality each month
• Finance decisions allocating capex and opex, and capitalization periods
Large projects often start with top down budgets, and evolve to estimation. Intuitive estimates
of complex, multi-year projects are not effective. Companies can instead execute pilots. Based
on this learning, capital commitments can evolve from top down to bottom up.
11. Connecting Plans to Agile
• Agile practices can fulfill hard date and budget commitments.
• Change will happen, e.g. customer RFPs, regulatory events, or M&A. Development processes need to accommodate change.
• Last minute changes to releases affect quality and stability, and demotivate teams.
• Disciplined, consistent iterations allow the business to confidently change which features go in which release.
Business planGoals definition
Design
Business process design
determines use cases, which are
implemented within the technical
architecture.
2016 Budget and Delivery PlanBusiness strategy
Design phase pilots provide baseline references for
budget and delivery commitments.
Unsuccessful
pilots can be
stopped.
Begin to stabilize spring
timing, e.g. monthly.
Design phase pilots can vary in time and
scope. They should be built with intent to
bring features to production.
12. Agile Development in Global Organizations
Design
Process design guides use cases.
2016 Budget and Delivery Plan
• Agile streams work well within a site, while less well across sites.
• Agile can take many forms such as Kanban or scrum. Sites will adopt their own practices.
• Within a site, there is a limit to how many agile teams are effective concurrently. Smaller numbers of effective people produce more than big teams.
Prague
New York
Tampa
Assign sites well-bounded, discrete processes.
Ideally assign complete responsibility, with local managers accountable.
Plan for reasonable T&E, and send people between locations. Ideally
move a small number of experienced people to new locations.
13. Leading Change
• Ineffective change leadership is a top cause of “black
swan” projects with very large cost and time overruns.
• Large projects produce uncertainty, affecting productivity
and quality.
• IT managers should explicitly discuss, learn, and plan
change leadership. Creating written plans works well.
• Communication is the central need. Show how change
produces new career opportunities, enables learning
interesting technologies, and is positive for people.
The Kotter change model fits well with the methodology.
A bottom up, Socratic approach is particularly effective: form
teams of junior people to recommend how to change.
14. Governance and Engagement
Product management
• All large IT projects are business projects.
• Start by involving business leaders in
change leadership communication, e.g.
video interviews.
• Always hold forums and send reports,
even at project inception.
• Track and report how strategy connects to
business goals, and then to process
design and use cases.
• In operation, show realization of business
plan targets.
Customer service
Audit
Finance
HR
Sales
IT
Engagement opportunities:
• Within communication campaign
• Business sponsors of sprints
• At the completion of each sprint
• Defining business processes and use cases
• Optimizing visual mockups
• Defining test cases
• Performing QA checks
• Governance forums and reports
15. Summary: Traditional vs. New Approach
Create enduring new capabilities for IT to operate with more efficiency and agility.
Long waterfalls
Design expressed in long documents
Intuitive estimates
Manage only to time and budget
Test after coding.
Address non-functional needs like
security and scalability after launch.
Big bang deployment
Tight, rapid iterations. Each sprint engages the business and is
nearly ready for production.
Visual mockups, and technical pilots. Provides a running start.
Facts from mockups and pilots inform estimates
Achieve business outcomes and motivate people
Test continually. Potentially code to tests.
Quality, security, and scalability starting with planning and coding.
Automated, rehearsed, parallel deployment
Traditional Target
16. Contact
Machine Intelligence Partners LLC serves clients
globally. Our people are centered in Boston,
Bozeman, Grand Rapids, London, New York, San
Francisco, and Washington, D.C.
Client relationship leaders:
New York
Jeremy Lehman
917.225.2011
jeremy.lehman@machineintel.com
Washington, D.C.
Philippe Berckmans
804.405.6009
philippe.berckmans@machineintel.com
Machine Intelligence is an Amazon Technology Partner
and member of the Microsoft Partner Network.
We are a veteran-owned small business.