Building a Smart Data Enterprise Architecture

CIO Academy Asia hosted a special leaders’ breakout session, as part of the Analytics insights Exchange organised by SAS.
10 JuLY 2018 | Singapore

Demonstrating business value and developing speed and flexibility in data analytics

Organisations today are convinced that data is critical to business survival, but further investments in data analytics must be proven by quick-wins that demonstrates a clear business upside to the board. For many corporations, board accountability on a quarterly basis imposes immense time pressures on digital leaders. Delivering business improvements on time and streamlining processes on existing infrastructure must be a priority. Underlying these quick-wins should be a longer-term objective to scale up and operationalise successful pilots.

Embedding analytics and leveraging the Internet-of-Things (IoT) in business operations

Existing infrastructure hold tremendous potential for analytics to be embedded. IoT has enabled physical assets such as machines and storage spaces to collect data and optimise resources on a real-time basis. With IoT, the number of sources, variety and volume of data multiplies, providing more opportunities to improve data insights.

Lending his expertise on the best practices to help organisations gain the most out of data analytics is Jason Mann, Vice President of IoT, SAS,

  • Managing the sprawl of data

Today’s operating environment has become more complex, where the sprawl of data workloads has increased exponentially over a hybrid infrastructure architecture – in the Data centres, the Cloud and at the edge. Compounded by rising expectations for real-time insights and automation, the traditional and expensive approach of data warehousing has to shift toward a decentralised and distributed approach, where data insights are generated at the point of collection. Although maintaining a high level of performance efficiency within such a distributed architecture at scale will require a re-allocation of compute and storage resources, such efficiencies have shown to generate significant cost savings as the volume of data grows exponentially.

  • Making use of existing infrastructure

The prevalence of connected edge devices such as mobile phones and laptops today is one of the most dynamic platforms to collect and process a variety of data points. One example cited by Jason Mann points to geolocation-based retail promotions. By analysing geo-location data, retail outlets are able to offer promotions to customers when they are within physical proximity of an outlet.

Intel has similarly developed easy-to-deploy compute capabilities to support analytics at the edge. To help organisations jumpstart their journey in analytics and IoT, Intel has developed Market Ready Solutions (MRS) for end-user organisations, a simple starter-kit solution that end-users can deploy on existing infrastructure. System Integrators are also empowered to bring an ecosystem of solutions together for their customers through the RFP ready kits (RRK) offered by Intel and its partners.

  • Leveraging flexible consumption models to jumpstart analytics projects

Everything-as-a-Service or what is now commonly known as “XaaS” is now a popular business model that has allowed for small-scale pilots requiring a suite of microservices to be operationalised on a flexible cost model. This approach has proven to be effective in getting organisations started on data analytics, where cost flexibility can be achieved across the critical components – from the expansion of network and storage capacity to the procurement of software licenses as technology teams grow.

SAS offers a flexible consumption-based pricing model that tracks and takes a reward from the business impact achieved, based on a set of metrics defined for each project. “Results-as-a service” as this model is called, has served as the business backbone for local energy efficiency company, Barghest Building Performance (BBP).

BBP utilises a mix of proprietary hardware and software technology combined with data analytics capabilities provided by SAS to reduce utility costs for industrial operations. The company aggregates sensor data and automates equipment energy usage optimisation at the edge. Besides static energy usage reports, automation algorithms are fine-tuned based on new data inputs.

BBP’s CEO, Poyan Rajamand, attributed BBP’s success to having access to highly accurate data collected on the company’s proprietary sensors. In addition, the deployment of analytics at the edge greatly enhances processing speed and saves significant costs in managing data on a network.

Konica Minolta has benefitted from an IoT upgrade on its production machinery. An established Japanese company known for its business and industrial imaging products, Konica Minolta has utilised SAS analytics to build capabilities in fault detection and predictive maintenance within approximately 80% of its existing business technologies. Since its implementation, the company has seen a significant increase in the lifespan of production parts as well as a notable fall in customer claims.

Meeting new expectations of the business – Not everything needs real-time attention!

 Business stakeholders are no longer satisfied with static reports generated by descriptive analytics. Rather, action-oriented analytics will be the new industry standard, where data models are operationalised at the edge to generate actionable data-insights.

These evolving demands are most apparent in the retail sector. Ascentis, a Singapore based company that designs rewards programmes and e-commerce platforms for retail organisations has realised increasing benefits from embedding analytics within its e-commerce services and loyalty programmes. The company has relied on automated decisioning processes based on data-insights about customer behaviour to boost retail sales for its customers. Bryan Ang, CEO of Ascentis, shared his recommendations in building a compelling case study for the board to strengthen their investments in data analytics. Although there is no silver bullet, the most critical success factors that technology leaders have to establish would be:

  1. A high value business objective and the role of data analytics to address the business gap
  2. The type of data for collection; the site of collection and the purpose for the data being collected
  3. A method of performance measurement to clearly track and demonstrate business impact.

Business stakeholders need work with their technology counterparts to prioritise real-time needs for data insights vis-à-vis insights that could be generated with cheaper data warehousing techniques that may require a lengthier throughput. Real-time capabilities take up significantly larger operational resources (such as compute and storage capacity) and should be deployed only when a clear business upside is expected to outweigh the associated costs.

Risk management is paramount in managing data

Another dimension of data analytics that is often the blind spot of organisations is in the area of data governance and cybersecurity. As organisations connect business critical systems and physical assets over the cloud, data connectivity is enhanced and managed services, such as those offered by BBP and Ascentis are better utilised. However, there exist a shared responsibility between service providers and end-user technology leaders to ensure that cyber-resilience is embedded into any new deployments, so that –

  1. Critical business data is protected and not vulnerable to loss or cyber-hacks
  2. Critical information infrastructure enabled by IoT capabilities are not vulnerable to service disruptions either by system failures or malicious cyber-attacks
  3. Clarity in data ownership is established especially with tightening regulations and heavy penalties for non-compliance
  4. Robust data back-up is in place to restore business operations when new deployments fail


The business value chain we know today is evolving. Unconventional partnerships and disruptive business models have brought together unrelated industries and upended well-established distribution channels; just like how banks are offering micro-loans services through e-commerce platforms, or how such platforms are disrupting brick-and-mortar pharmacies by working directly with drug companies to prescribe and deliver over the counter medicine. The rise of digital connectivity has been influential in creating efficient data sharing platforms for better business outcomes in data analytics.

While the possibilities are endless, even for the ecosystem of service providers that are constantly drawing new data insights from their growing portfolio of customers, digital leaders have to establish three key focus areas:

  1. Developing trust with stakeholders
  2. Narrowing down to high-impact business needs
  3. Achieving quick-wins on existing infrastructure

These focus areas will shape the D.N.A of a data-driven organisation!