To speed up and democratise decision-making one needs an analytics capability that embraces actionable foresight.
Move business and market analytics eg. customer behaviour, from purely informational to actionable, by not just noticing what will happen, but also determining what needs to happen. Create an OKR/KPI analytics framework that helps you align business activities with your strategy and goals, but ensure that the data, the algorithms, and the outputs can all flow and support a mix of human judgement and machine-generated actions.
Let’s take a closer look at all the key components and how they build on each other to create that ultimate prescriptive model and unlock making taking action “automated” as much as possible. Accuracy and explainability become key attributes to keep in mind in this journey.
Information & Data Science
Powering your Business Intelligence, AI models, and prescriptive analytics will be your Data platform. Whether built in-house or in the Cloud, it’s the engine that will aggregate data from source systems, go through multiple stages of quality checks, and then be made available for exploration, reports, dashboards and automated actions.
Core Services include Data Aggregation (structures, unstructured, semistructured), Cleaning & Quality checks, Data enrichment (3rd parties), Data Models (consistent), LLM/RAG (domain enhancements), Data Science Modelling (Machine Learning), Generative AI (augment content), Agent Orchestration (automate).
We have partnered with multiple leading vendors to provide comprehensive solution capabilities, including Cloud technology stack, Data Engineering tech, Data Science tech, Devops/Dataops, Data Governance, and Analytics/Visualisation platforms.
Your data volumes are exploding, stored in multiple locations, with a very urgent need to extract meaningful value from it to stay competitive.
Descriptive and Diagnostic Analytics
Your business model, and strategic direction, including customer, product, operational and transformation plans, will need OKRs and KPIs that are most relevant and impactful. These metrics will be visible to all teams, whether C-level, Exec teams, front-line teams, or tech teams. Ensuring everyone stays aligned and focused on activities that make a difference to your business.
Marketing & Sales
Commercial insights and automation will unlock scaling outreach sales, lead generation campaigns, and a better understanding of customer segments, value and growth. Here are some use cases:
- Customer Segmentation and needs identification
- Growth modelling & Competitive intelligence
- Customer Journey and A/B testing (see below)
- Sales Performance, including impact attribution
- Demand forecasting and leading factors
- Churn prediction and preventative actions
- Prospecting, scoring and personalised engagement
- Automated recommended actions (e.g. support, cross-sell, up-sell)
Qualitative Insights
Leverage subjective judgment to analyze a company’s value or prospects based on non-quantifiable information. Whether from customers, employees, partners or execs, there are a wide variety of information-gathering methods, as well as decision-making analysis, often combined with quantitative insights.
- Customer Surveys e.g. Net promoter scores, Customer Satisfaction CSAT
- Colleague Surveys e.g. Workday Peakon employee voice
- Stakeholder Objectives, Pain Points & Goal Success
- Interviews and focus groups
- Phenomenological Method
- Ethnographic (Observational) Model
- Grounded Theory Method
- Case Study Model
Quantitative Insights
Derive more information and conclusions derived from data that can be measured and expressed numerically. A more objective and mathematical approach that balances accuracy versus explainability.
- Customer segmentation & value
- Customer engagement and churn
- Performance metrics Taxonomy
- Financial metrics Taxonomy
- Data modelling and calculations
- Visualisation tools (Google Analytics, Tableau, Power BI, Realtime SaaS dashboards)
- Optimisation or Automation using Machine Learning or AI methods
- Correlational & Experimental Research
Diagnostic Example: Customer Journeys
This example looks at all the questions one can ask to get a better understanding of the customer experience. Understanding the root cause of what’s driving the overall issue will lead to better actions.
What is the customer value chain and journey?
- Do the users achieve what we expect them to achieve?
- Are the main features of the product used?
- What are the product features they can’t do without?
What does the critical funnel look like?
- At which step do users drop off?
- What do they try to do instead?
- What content is most effective?
What does our onboarding conversion look like?
- How many people make it through the onboarding?
- How many people reach the “aha” moment?
- How Can A/B testing be used to optimise approach
What reputation and word-of-mouth referrals are you tracking?
- Understand customer sentiment and buzz in all channels
- What and where are the customer service pain points?
Predictive Analytics
Predictive analytics goes beyond the what and why to the what’s next. Using predictive models, one can forecast what is likely to happen in the future and assess what may result if certain actions are taken. Forecasting itself has evolved, and new models and the amount of data and machine processing power available allow machine learning techniques to be applied 24/7, improving the balance between bias and accuracy.
There are many types of predictive analytics models, which we can help you create using a range of tools and software algorithms. Simulators are often used to mimic the business or operational model.
- Classification Model e.g. Predict binary outcomes such as will the customer churn
- Clustering Model e.g. Customer segments for personalised targeting
- Forecast models e.g. Determine the best price for a product to sell
- Anomaly/Outliers Model e.g. Is a particular product return fraudulent
- Time Series model e.g. What are expected sales for the next 3 months?
Prescriptive Analytics
This stage of actionable analytics is the culmination of the other forms of analytics. By ensuring good quality data to document what happened, understand why it happened, and predict what may happen; then prescriptive analytics is about ensuring your business can have a clear set of actions to take to achieve desired results.
Example use cases include making capital investment decisions; deciding where your product stock should be located and moved; optimising patient healthcare plans; deciding which new features to build; adjusting marketing campaigns based on context; and setting product prices based on customer demographics and sensitivity.
Understanding the optimal balance of human judgment with pure data-driven analytics methods remains challenging, but the value being demonstrated is enormous.
