Data-Driven Decision Making Training Course

Duration 5 days 10 modules · 33 contact hours
Level Intermediate Certificate on completion
Who attends Business Analysts, Product Managers and others
Fee from USD 1,100 KSh 95,000 for Kenyan organisations
Where 52 cities + online See dates by city
Next start 7 Nov 2026 Course code SD-DA-073

Before you join: Participants should be comfortable reading tables, charts and basic business metrics such as revenue, cost, volume, conversion or service levels.

Dates & prices

Two start dates a month in every city, and live online on weekdays or weekends. Fees are per participant; classroom fees include tuition, materials, lunch and refreshments, and the certificate.

Suggest a date or city

Showing the next six months. Later dates, other cities and private deliveries for your team are available on request.

Course overview

Many operational, policy and investment decisions are supported by reports but not by a clear chain from question to evidence, recommendation and follow-through. Teams may select convenient metrics, combine incompatible datasets, mistake correlation for cause, or present results without stating uncertainty. These weaknesses can lead to misallocated budgets, ineffective service changes, avoidable risk and low confidence in analytical recommendations.

This course develops the working methods needed to make defensible decisions from data. Participants frame decision questions, define hypotheses, assess data quality, establish meaningful KPIs and prepare analysis plans. They use SQL concepts to inspect and combine data, apply descriptive statistics and visual analysis, interpret sampling and statistical evidence, design A/B tests, assess forecasts and communicate findings to decision-makers. The course also addresses data governance, privacy, bias, model limitations and decision logging.

Instruction combines focused briefings with realistic cases drawn from banking, public services, customer operations, supply chains and programme delivery. Participants work through decision scenarios, analyse structured datasets, discuss assumptions with peers and practise presenting recommendations under scrutiny. They leave with a reusable decision-to-evidence framework, an analysis plan template, KPI tree, data quality checklist, experiment design canvas and 90-day application plan. The course is suited to professionals who use data to recommend actions, commission analysis or assess analytical work produced by others.

What you will be able to do

  • Frame business decisions using decision statements, stakeholder maps and testable hypotheses
  • Assess dataset fitness using provenance records, data dictionaries and quality dimensions
  • Construct KPI trees linking operational measures to financial, service and risk outcomes
  • Query and validate relational data using SQL filters, joins, aggregations and window functions
  • Interpret descriptive statistics, variation patterns and segment comparisons accurately
  • Evaluate statistical claims using confidence intervals, effect sizes and bias checks
  • Design A/B tests with primary metrics, guardrails, power calculations and decision rules
  • Present evidence-based recommendations using decision-focused charts, narratives and decision logs

Benefits of attending

For you

  • Improved ability to challenge unsupported conclusions in reports and dashboards
  • Practical templates for structuring data-backed recommendations
  • Greater confidence discussing statistical uncertainty with analysts and senior stakeholders
  • Stronger capability to define KPIs that reflect actual decision objectives
  • A 90-day plan for improving one live decision process

For your organisation

  • More consistent links between management decisions, measures and underlying evidence
  • Fewer reporting errors caused by unclear definitions, poor joins and weak quality checks
  • Better prioritisation of interventions through hypotheses, experiments and forecast comparisons
  • Clearer executive recommendations with stated assumptions, risks and decision thresholds
  • Stronger governance through documented data sources, approvals and decision outcomes

Course modules

10 modules over 5 days

Participants can convert an ambiguous business issue into a defined decision question and testable hypothesis.

  • Decision statements specify owner, choice, timing, constraints and success measure
  • SMART objectives convert strategic questions into measurable decision criteria
  • Problem trees distinguish symptoms, root causes and controllable drivers
  • Stakeholder maps expose competing objectives, incentives and decision rights
  • Testable hypotheses link proposed actions to measurable outcome changes
  • Value-of-information estimates determine whether further analysis justifies delay

Participants can assess whether available data is fit for a specified decision.

  • Data provenance records source systems, transformations, owners and refresh schedules
  • Grain definitions prevent accidental mixing of transaction, customer and monthly data
  • Data dictionaries document fields, units, allowed values and business meanings
  • Completeness, validity, consistency, timeliness and uniqueness assess data quality
  • Missing-data mechanisms distinguish MCAR, MAR and MNAR treatment choices
  • Reconciliation checks compare aggregates against controlled operational reports

Participants can define decision-relevant measures and document an analysis plan before examining results.

  • Analysis plans predefine population, measures, segments, comparisons and decision thresholds
  • KPI trees connect operational measures with financial, service and risk outcomes
  • Leading and lagging indicators clarify intervention timing and accountability
  • Baselines, targets and tolerances prevent ambiguous performance interpretations
  • Cohort definitions ensure retention, default and service analyses remain comparable
  • Metric governance assigns calculation ownership, approval and controlled change procedures

Participants can use core SQL concepts to retrieve, combine and validate decision data.

  • Relational tables, primary keys and foreign keys establish reliable joins
  • SQL SELECT, WHERE, GROUP BY and HAVING answer defined business questions
  • INNER, LEFT and anti-joins identify matched records and exceptions
  • Window functions calculate rankings, moving averages and period-over-period changes
  • Aggregation checks detect duplicated rows and unintended many-to-many joins
  • Reusable query logic documents filters, assumptions and refresh dependencies

Participants can identify meaningful patterns, variation and exceptions in operational data.

  • Distribution shape, centre and spread reveal typical performance and variability
  • Percentiles and box plots identify outliers without relying on averages
  • Cross-tabulations compare segment patterns while preserving denominator context
  • Time-series decomposition separates trend, seasonality and irregular variation
  • Control charts distinguish common-cause variation from special-cause signals
  • Funnel analysis locates conversion losses between defined process stages

Participants can interpret statistical evidence without overstating what the data proves.

  • Random sampling and sampling bias determine confidence in observed patterns
  • Confidence intervals express plausible ranges rather than false precision
  • Null hypotheses, p-values and effect sizes support proportionate claims
  • Correlation matrices flag associations without establishing causal direction
  • Confounding variables and selection bias threaten observational conclusions
  • Regression outputs interpret coefficients, uncertainty, fit and diagnostic warnings

Participants can design controlled tests and assess whether an intervention produced a meaningful effect.

  • Randomised controlled trials isolate treatment effects using assigned comparison groups
  • A/B test designs specify primary metric, guardrails, sample size and duration
  • Power calculations estimate detectable effects before allocating intervention traffic
  • Randomisation checks verify balanced groups before interpreting experiment results
  • Multiple-comparison controls reduce false discoveries across repeated metric tests
  • Difference-in-differences estimates policy effects using pre-post comparison groups

Participants can assess forecasts and use uncertainty ranges to support planning decisions.

  • Forecasting purpose determines horizon, granularity, update frequency and error cost
  • Naive and seasonal-naive benchmarks establish minimum acceptable forecast performance
  • Moving averages and exponential smoothing model stable demand patterns
  • Forecast accuracy uses MAE, RMSE, MAPE and bias measures
  • Prediction intervals communicate uncertainty for staffing, inventory and budget decisions
  • Scenario assumptions translate operational drivers into alternative planning estimates

Participants can create decision-focused visual narratives that state findings, implications and recommended actions.

  • Chart selection matches comparison, distribution, relationship, composition or time questions
  • Pre-attentive attributes direct attention to the decision-relevant pattern
  • Annotated charts state takeaway, comparator, period and material caveat
  • Dashboard wireframes separate executive decisions from operational monitoring needs
  • Data ink reduction removes decorative elements that obscure evidence
  • Executive narratives connect finding, implication, recommendation and requested decision

Participants can apply governance controls and plan the transfer of course methods to a live decision.

  • Data access classifications protect personal, confidential and commercially sensitive information
  • Purpose limitation and minimisation guide responsible collection and reuse
  • Model risk registers document assumptions, limitations, owners and review dates
  • Decision logs capture evidence, rationale, approvers and subsequent outcomes
  • Bias checks test whether decisions create unequal effects across groups
  • 90-day application plans prioritise decisions, datasets, stakeholders and review milestones

Exercises and examples are adapted to each cohort's sectors and roles.

What you take back to work

  1. Decision-to-evidence framework
  2. Analysis plan template
  3. KPI tree model
  4. Data quality assessment checklist
  5. A/B test design canvas
  6. 90-day decision improvement plan

Who should attend

  • Business Analysts — need defensible methods for turning operational data into recommendations
  • Product Managers — need evidence to prioritise features, journeys and service improvements
  • Operations Managers — need reliable performance measures and root-cause analysis
  • Policy and Programme Managers — need to assess outcomes and justify intervention choices
  • Risk and Compliance Professionals — need to challenge analytical claims, assumptions and controls
  • Data and Reporting Managers — need consistent decision methods across reporting teams

Skills and tools covered

Decision framingData quality assessmentKPI designStatistical interpretationExperiment designEvidence communication

Tools and standards

CRISP-DM, SMART criteria, A/B testing, Power BI

How the course runs

Training methodology

The course uses short instructor briefings to introduce methods, followed by case-based analysis, group discussion and structured problem-solving. Participants examine realistic decision scenarios, define metrics, test claims, interpret charts and compare alternative recommendations. Spreadsheet-based data inspection and SQL examples demonstrate how analytical choices affect conclusions. Simulated stakeholder discussions build skill in defending assumptions and communicating uncertainty. Each participant develops an application plan for a current decision, report or improvement initiative.

Requirements and prerequisites

Participants should be comfortable reading tables, charts and basic business metrics such as revenue, cost, volume, conversion or service levels. Some experience using spreadsheets, reports or dashboards is expected, along with the ability to discuss a current work decision. No prior statistics qualification, programming background or advanced SQL knowledge is required. The course introduces SQL concepts and statistical methods from a decision-maker's perspective, although participants will gain more from examples if they have worked with operational or customer data before.

A typical training day

08:30 – 10:30First session
10:30 – 10:45Refreshment break
10:45 – 12:30Second session
12:30 – 13:30Lunch and networking
13:30 – 15:00Third session
15:00 – 15:15Refreshment break
15:15 – 16:30Workshop and daily review

Live online deliveries follow the same structure in East Africa Time, with shorter screen blocks and longer breaks.

What the fee includes

  • Instruction by a practitioner facilitator
  • Full course workbook and materials
  • Exercise files, templates and case studies
  • Refreshments and lunch on classroom deliveries
  • Certificate of completion
  • Post-course application plan

Frequently asked questions

Participants should be able to read business reports, tables and charts, and should have basic spreadsheet familiarity. No statistics qualification, programming experience or advanced SQL knowledge is required.

Participants should bring a laptop suitable for spreadsheet and browser-based work, plus details of a current business decision or reporting challenge. Course materials and sample datasets are provided.

It is intended for professionals who make, influence, commission or review decisions supported by data. It is particularly relevant to analysts, managers, product teams, programme leads, risk teams and reporting functions.

This course concentrates on decision quality: framing questions, testing claims, interpreting evidence and making recommendations. It uses analytical tools and examples but does not focus on building reports or mastering one software product.

Participants can use the analysis plan, KPI tree, data quality checklist and decision log on an active initiative. The final module helps them identify a priority decision, stakeholders, data sources and review milestones.

Each training day runs from 08:30 to 16:30 and provides about 6.5 contact hours. Time is divided across instructor input, case analysis, group work, data interpretation, discussion and application planning.

Questions about this course?

Ask about dates, fees, group bookings or running it for your team. Our training team replies by email.

Ask on WhatsApp info@skillsetdevelopment.com

Related courses in Data Analytics