At DAAG 2014, in the Ubiquitous DA: Embedded DA session on Mar 27, 2014 at Boston, David Matheson facilitated an audience workshop to reflect on the differences between Embedded DA and Consultative DA. The first four slides are a summary of speaker talks in that session, while the rest are the result of audience contributions. The original flipchart photos are also included.
2. Focus of Software
Consultative Embedded
“Knobs for nerds”
Powerful, flexible tools
for analysts, address any
situation
“Get my work done”
Support multiple roles
with simple interfaces
that focus on what I
care about
Examples
Dow Agro, Eli Lilly, Chevron
3. How DA is Represented
Consultative Embedded
Tell them what you
are going to do. Train
before start.
Focus on problem and
deliver results, creating
self-motivated interest
Examples
Eli Lilly
• Knowledge —> broader vs specific training
• Culture Change
• Example - Embed DA Expert at Siemans H.S.
• Consultative Model is used in certain circumstances —>
Embedded DA at other times
4. Transparency of Information
Consultative Embedded
Analyst passes relevant
(possibly secret) information
to decision makers
Create standard work
processes that drive
transparency and deal
with the conflict
Examples
Eli Lilly
Issue: Information gets prettied up
5. Mid-Level Decisions
Consultative Embedded
Never see decisions or try to
train everyone. Opportunity lost.
Provide fit for purpose tools
that make relevant
information available to
people as they make routine
decisions
Examples
Chevron
Short Fuse Decisions
Asset Integrity
6. Six Elements of Decision
Quality: What do we want?
Consultative Embedded
Discovery interviews with
decision-makers
Policy statements,
mission statements,
clarity of objectives
7. Six Elements of Decision Quality:
What are our alternatives?
Consultative Embedded
Framing =>Strategy Table Teach/Train teams =>
Standard Frame Templates
8. Six Elements of Decision
Quality: WHAT DO WE KNOW?
Consultative Embedded
Framing & SMEs Databases, policy
capturing (debiasing)
9. Six Elements of Decision
Quality: How do we decide?
Consultative Embedded
Model alternatives (supplement
DM judgment)
Achieve clarity quickly with
simple standard models
10. Business Process
Consultative Embedded
Issue: Do we embed tools or people people into processes?
Continuum
• Skills
• People
• Processes
• Tools
• Domain Knowledge
• Problem Complexity
• Problem Frequency
Example
Empowered reservoir engineer —> improved normative
decision-making
11. Decision Process
Consultative Embedded
Required for projects requiring
a discovery process
Needs a frame upfront
Iterative process over a
short time frame
Not necessarily reusable frame
More time-intensive
Suitable for Repeatable
Projects
Lends itself to standard
templates
Process is well-established
Decision criteria clear/
well-established
12. Transparency of Information
Consultative Embedded
Dedicated team of analysts
Work on the projects one at a
time
Individual projects are time
bound
Gives an ‘answer’ at the end
of each project (yes or no)
Training
Simple, robust templates &
tools
Accept continuous
improvement
even when ‘answer’ is not
obtained
Issue:
1. Non-routine maintenance in O&G production facilities
2. R&D of Technology Portfolio
13. Broader use of Decision
Framework
Consultative Embedded
Bring in a consultant,
professional facilitator
knowledgable in techniques
Framing template - easy to
follow for asset teams
- generic between projects
Make it easier to push framing into the organization
Example
Shell Person! Framing Templates, VoI Templates, OF’s OTrio
Framing
14. Issue: Critical, Stuck, Not a typical
problem, Mid-sized decisions
Consultative Embedded
Outside firm, one-off focus on
tools
Have a system that shows:
• Assemble these kinds of
people
• Ask these types of
questions
• Test these items
• Community of Interest
• “Sophisticate” LUI of DA
ability
15. Issue: Decision for
production planning
Consultative Embedded
Very centrally owned and not
well understood
• Commonality of term
definitions (i.e. high vs low)
• Transparency to understand
key drivers
• Less about rolling
optimization
• Automating forecasting
process
• Lookbacks and learning
loops for future decision
16. Issue: Protect DA from
Change Management
Consultative Embedded
• Create/coin names for tools &
processes
• Our org performs due diligence on
every decision because we won’t let
ourselves use a simple bang for the
buck chart
• How to protect DA processes from
change in exec champions? To
embed org, needs to make it their
own - decision fitness. Consultative
has their own names for everything.
• We don’t make decisions as a portfolio
stream. We make many little decisions
separate from each other. Could we
reference an ongoing bang for the
buck chart?
• Templates useful when orgs are initially
learning new processes, but risky if
using unclear terminology (excellent vs
very good not defined) and not
ensuring alignment