peopleanalyst

library / lib817666d84297f06e

Analytics at Work: Smarter Decisions, Better Results

Thomas H. Davenport, Jeanne G. Harris, Robert Morison · 2010

In a sentence

A practical, implementation-focused guide showing how any organization can build the capabilities to put analytics to work in everyday decisions and processes to make smarter decisions and get better results.

Most companies are awash in data yet make 40 percent of major decisions on gut instinct, leaving money and competitive advantage on the table. In this sequel to Competing on Analytics, Davenport, Harris, and Morison move beyond the rare analytical competitors to the broad base of organizations that simply want to become more analytical—one decision at a time. The book organizes the prerequisites for analytical success under the memorable DELTA framework (accessible high-quality Data, an Enterprise orientation, analytical Leadership, strategic Targets, and Analysts), pairs it with a five-stage maturity model, and then explains how to sustain analytical capability by embedding analytics in business processes, building an analytical culture, and continually reviewing models and assumptions. Illustrated with examples from Best Buy, Progressive, Humana, 1-800-Flowers, Capital One, and many others, it offers frameworks, assessment tools, and pragmatic advice—a compass rather than a rigid map—for managers who want to unleash the value buried in their data.

The four lenses

  • Science
  • Statistics
  • Systems
  • Strategy

The model

A capability model in which five organizational design levers (Data, Enterprise orientation, Leadership, Targets, Analysts) plus sustaining practices (embedded analytics, analytical culture, continuous model review) drive fact-based decision making, which in turn improves business performance and results. Analytical maturity moderates the strength of these relationships.

Accessible High-Quality Datadesign lever

The degree to which an organization has clean, integrated, accessible, well-governed, and where possible unique/proprietary data structured and managed for analytical (not just transactional) use.

Enterprise Orientationdesign lever

The extent to which data, technology, analysts, and analytical decisions are coordinated holistically across organizational silos rather than fragmented into local, self-serving fiefdoms.

Analytical Leadershipdesign lever

The presence of leaders at any level who passionately manage by fact, push for data and analysis, set example and performance expectations, hire smart analysts, and build an analytical ecosystem.

Strategic Analytical Targetsdesign lever

The degree to which analytical efforts are focused on high-value, high-impact business processes and distinctive capabilities that drive performance and differentiation rather than scattered low-value projects.

Analyst Talent Capabilitydesign lever

The supply, mix, skills, engagement, organization, and deployment of analytical talent (champions, professionals, semiprofessionals, amateurs) building and applying models across the enterprise.

Embedded Analytics in Business Processesbehavioral pattern

The degree to which analytical models and decisions are integrated into core operational processes and workflow so that insights are routinely and automatically acted upon.

Analytical Culturepsychological state

A shared set of attitudes and behaviors—searching for truth, seeking data not just stories, valuing negative results, pushbacks for unsupported claims, and acting on analysis—that makes fact-based decisions the norm.

Continuous Review and Model Managementbehavioral pattern

The systematic practice of reviewing and renewing strategy, targets, competitors, customers, technology, and analytical models/assumptions to keep them valid as conditions change.

Fact-Based Decision Makingbehavioral pattern

The use of objective data and rigorous analysis as the primary guides to decisions across strategic, tactical, and operational levels, with intuition employed only where appropriate.

Analytical Maturity Stagecontextual condition

The organization's position on the five-stage continuum from Analytically Impaired to Analytical Competitor, reflecting how developed and balanced its DELTA capabilities are.

Business Performance and Resultsoutcome metric

The downstream outcomes—better/faster/more consistent decisions, improved efficiency, risk management, profitability, growth, and competitive differentiation—achieved by putting analytics to work.

How they connect

  • data quality and access predicts fact based decision making
  • enterprise orientation influences fact based decision making
  • analytical leadership predicts analytical culture
  • analytical leadership influences strategic targets
  • strategic targets predicts business performance results
  • analyst capability predicts fact based decision making
  • data quality and access predicts embedded analytics
  • embedded analytics mediates fact based decision making
  • analytical culture mediates fact based decision making
  • fact based decision making predicts business performance results
  • continuous model review moderates business performance results
  • analytical maturity stage moderates fact based decision making
  • continuous model review correlates continuous model review

A candidate measure

Analytics at Work: Smarter Decisions, Better Results — derived measurement candidates

Accessible High-Quality Data

number of conflicting/duplicate data sources; data integration coverage (% of transaction systems in warehouse); count of proprietary/unique data assets used; data quality error rates in key domains

self-report suitability: medium

Enterprise Orientation

number of redundant data marts/tools; percent of analytics projects that are cross-functional; existence of enterprise governance bodies

self-report suitability: medium

Analytical Leadership

frequency of data-based pushbacks; number of analytical hires sponsored; visible analytics communications by leaders

self-report suitability: medium

Strategic Analytical Targets

ratio of high-impact to low-value projects; number of cross-functional/strategic targets; ladder rung achieved per major process

self-report suitability: medium

Analyst Talent Capability

headcount by analyst type; skill proficiency ratings; analyst engagement/satisfaction scores; retention/turnover rates

self-report suitability: medium

Embedded Analytics in Business Processes

share of decisions automated vs human; number of embedded vs standalone applications; insight-to-action latency

self-report suitability: medium

Analytical Culture

frequency of 'use data' pushbacks; instances of decisions reversed by analysis; perceived norm of fact-based decisions

self-report suitability: high

Continuous Review and Model Management

existence of model-validation function; review cadence frequency; number of models retired/updated per period

self-report suitability: low

Fact-Based Decision Making

percent of major decisions based on facts vs gut; decision-process audit scores; decision quality tracked over time

self-report suitability: medium

Analytical Maturity Stage

composite DELTA-by-stage assessment score; number of elements at each stage

self-report suitability: medium

Business Performance and Results

profit margin change; cost savings from analytics; market share; forecast/decision accuracy; decision cycle time

self-report suitability: low

Run the assessment

The story

The reader A manager or executive whose organization holds massive data but underuses it, and who wants to make more fact-based decisions and get better business results.

External problem

The company collects and stores data but doesn't analyze it to inform decisions, leaving money and competitive advantage on the table.

Internal problem

The reader feels they are managing on autopilot or going with their gut, unsure whether their decisions are right and frustrated by missed opportunities.

Philosophical problem

It's just plain wrong to make important decisions based on unaided intuition, bias, or 'because that's how it's always been done' when facts and analysis are available.

The plan

  1. Get accessible, high-quality, and where possible unique Data in order.
  2. Adopt an Enterprise perspective rather than fractured local silos.
  3. Build analytical Leadership at every level, setting example and expectations.
  4. Pick strategic Targets where analytics will make the biggest difference.
  5. Acquire, organize, and develop Analysts as a scarce, valuable workforce.
  6. Embed analytics in business processes, build an analytical culture, and continually review models and assumptions.

Success

  • Smarter, more consistent, faster, fact-based decisions; better problem solving and business processes; the ability to anticipate market shifts; improved efficiency, risk management, and profits; and a sustainable enterprise-wide analytical capability.

At stake

  • Continuing to manage on autopilot and gut feel, leaving money on the table, falling behind analytical competitors, being unable to push back on poorly understood risks (as in the 2007–2009 financial crisis), and stagnating while rivals improve.

Related in the library