Skip to main content
Data Innovation Accelerator

Using data science

The Data Innovation Accelerator supports companies to apply data science tools and techniques to their data, which can be a vital tool for business resilience and growth.

This allows companies to derive intelligent insights into emerging trends and behaviours, provide highly targeted sales and marketing strategies, streamline business processes and be highly responsive to customer sentiment and rapidly changing market conditions.

Through this process we have developed a selection of possible use cases.

Data science keywords

The topic of data science is teeming with terminology, a convergence of terms from computer science, statistics, mathematics, and software engineering. In addition, the language of data science evolves very quickly. This brief list includes some of the more common keywords.

  • Data engineering
  • Correlation
  • Time series prediction/forecasting
  • Social network analysis
  • Bayesian inference
  • Human behaviour analysis
  • Supervised and Unsupervised learning
  • Machine learning
  • Optimisation
  • Artificial intelligence
  • Modelling
  • Data mining/analysis
  • Data visualisation
  • Simulation

Use cases

These are potential data science use cases for business transformation and resilience:

Market analysis

  • Integration of market analysis with open public and third-party data sets (e.g., meteorological data, environment, public health, census and national statistics) in relation to company activities, to aid decision making, product development and optimisation of physical store locations
  • Social network data mining for real-time consumer behaviour analysis and product demand
  • Routing optimisation for on-demand services (public transport or positioning of hire bicycles)
  • Time-series predictive analytics for sales forecasting and lead generation (e.g., how store footfall is affected by weather)
  • Analysis of spatial and satellite image data to determine highly localised sales trends, sales opportunities, hot-spots and outliers

Sales optimisation

  • Leveraging of recommendation engines for product up-selling and cross-selling
  • Dynamic price optimisation (competitor comparison, spot pricing, seasonality adjustments)
  • "Deep" customer personalisation (e.g., context aware offers based previous search and retail history and group trends, loyalty rewards and retargeting opportunities)
  • Image recognition through deep learning to enable enhanced discoverability (e.g., visual and audio search, automatic image tagging)
  • A/B Testing frameworks and website micro-behaviour monitoring to improve customer experience and Conversion Rate Optimisation (CRO)
  • Analysis of product placement and prominence including product halo effects
  • Social graph analysis to determine like-groups, advocates and brand influencers
  • Augmented reality (AR) applications to overlay information within physical stores

Advertising and marketing

  • Developing highly personalised marketing communications
  • Use of Natural Language Processing (NLP) for context-aware advertising (e.g., optimal advert placements based on surrounding content)
  • Developing a coherent advertising strategy and performance evaluation across disparate platforms (omnichannel)
  • Bespoke dashboards and visualisation of real-time customer data
  • Image recognition for video commerce (e.g., product identification and tagging in customer/company generated content)
  • Automated marketing content generation
  • Social media analytics to identify target audience and to determine message and timing of social media communications

Services and support

  • Customer satisfaction analytics (e.g., conversation analysis, customer sentiment and emotion recognition through NLP)
  • Automating responses to customer feedback and social media trends
  • Development of AI "chatbots" for first line response and self-service support, and “listen-in” chatbots for assisted response and next action suggestions
  • Intelligent support call routing (auto-triaging, most capable agent, personality match)
  • Customer assistance with augmented reality in physical stores (e.g., location aware help, map/offer overlays)

Product improvement

  • Advanced modes of interaction and monitoring with digital products (gestures and anomalous motion behaviour)
  • Matching and recommendation engine optimisation (e.g., online dating)
  • Data mining to determine trends to feed into product development
  • Customer sentiment analysis following product modifications and updates

Business logistics and operations

  • Enabling data engineering techniques for inventory and supply chain optimisation (e.g., product information management, just-in-time pipelines)
  • Use of Internet of Things (IoT) technology for component failure prediction (predictive maintenance and automatic recovery)
  • Robotics and “cobots” for process automation
  • Effectiveness of cashier-less checkout systems (e.g., self-checkouts, RFID item tagging)

Business processes

  • Creating digital assistants for automatic parsing and understanding of large volumes of documents (e.g., ensuring regulatory compliance)
  • Hiring optimisation (e.g., determining the best candidate from text analysis of candidate information)
  • Employee retention and productivity monitoring including smart attribution and compensation based on employee performance
  • Assisting with internal communication and in-house processes (meeting scheduling, email prioritisation, expense and approval reporting, billing)
  • Building and estates management (e.g. smart energy monitoring of offices)
  • Marketing and PR analytics (collection of marketing data and KPIs)

Data management and warehousing

  • Efficient harmonisation and processing and visualisation of disparate large-scale data sources (pre-processing, cleaning, ETL, validation)
  • Creation of data processing pipelines to trigger actions on customer events
  • Leveraging cloud computing to scale rapidly with demand
  • Data labelling to enable supervised machine learning applications

Security and fraud

  • Motion sensors to detect anomalous activities (e.g., crowd monitoring)
  • Real-time fraud detection
  • Image and voice authentication
  • Malware information sharing across company infrastructure

Risk management

  • Predictive analytics for customer creditworthiness (e.g., calculating the probability of a customer not paying their credit card minimum payment to a business)
  • Streamlining debt collection processes and dispute handling