Posted: 17 Dec. 2020 05 min. read

6 easy steps to transforming your retail and consumer business on cloud

How do I leverage technology to drive innovation in retail?

Step 1: Connect the dots to innovate

Q - How can businesses best optimise data, to proactively engage with customers and drive innovation?

Instead of trying to industrialise the data foundations from day one, organisations can implement a Proof of Value approach. This is where Cloud plays a big role as it enables them to quickly produce an iterative low-fidelity prototype and test it with customers for immediate feedback. Then they can iterate with the business over a couple of sprints and translate the Proof of Value back to the ‘industrialised data pipeline’.

Initially, this process begins with identifying one or two high-value use cases. These need to have a ‘crunchy question’ behind it, which is a question about a metric or a tangible, measurable element that will move a business forward. For example, it might be improving an engagement score, moving a product from the shelves faster, replenishing the supply chain in line with demand - specific and measurable elements like that. 

From there you can distil the problem into potential analytical use cases — that can be either descriptive or prescriptive — and you pinpoint how it can be made into a reality. 

Next, you interrogate the use case with business stakeholders. It’s important to have a detailed conversation with them to properly understand the business logic that drives that particular metric. Through that, you will understand that there is always a data element involved. People are not always explicitly aware of this, but once they start engaging with the problem it becomes clear that there’s always a data component behind decision-making. 

When organisations have an operating model that lets you distil these problems into use cases, very soon there is a backlog — that’s why it’s an innovation pipeline. It’s full of use cases that you can quickly prototype and test with customers. For example, you can pinpoint a specific problem within your sector such as increasing customer propensity to buy a product by a certain per cent. 

Now you can leverage the Cloud to rapidly build prototypes and collect feedback and insights from users. 
 

Step Two: Laying your data foundation

Q - How does one build a data foundation?

Building the data foundations is akin to building a scalable ‘architectural runway’. Imagine you are building a runway horizontally (the platform) and then you create vertical elements on top of the horizontal. These are your use cases. 

Once you’ve distilled the problem and understand the insight that drives the metric, you need to identify the levers at your disposal. What data is available? Are the datasets internal or external? Do we have enough information to build predictive models that relate to the insight? Who do we involve? 

Through this process, you’ll identify many important datasets and also some bottlenecks in the dataflow. You’ll gain a better understanding of the change you are driving and what levers you need to pull. Your driving force becomes more sticky when you work backwards in this fashion. 

Your platform is an enabler of this process. At Google, there is a concept of ‘zero ops’. Instead of creating a lot of infrastructure, organisations should focus on activating the one key service in the Cloud that drives value. Our solution Google BigQuery conforms to zero ops by allowing clients to connect to a data warehouse or lake, and also to connect other sources and end-user tools. It is perfect for an analytical purpose like this. 

We’ve noticed in the data and analytics space that retail players are increasingly consuming services that focus on price benchmarking, elasticity of price, assortment recommendations, and product trains. They are seeking ‘off the shelf’ solutions and artificial intelligence models for forecasting recommendations and products. 

Having access to such a service can quickly industrialise data. Instead of building data models from scratch, you can merge your datasets with existing services on the Cloud. In a matter of weeks, you can start testing and validating the insights with marketing teams and customer teams and start feeding back in through the cycle. 

When you get results that are high in accuracy or assertiveness, you can return to the two-speed approach and start industrialising the insight and translating it into the fabric of how you do business. That’s how you incrementally grow and become an insight-driven organisation. 

There is work to build a foundation to get there. But once Cloud as a platform is enabled organisations can stand up instances in hours or days instead of months as you usually would with on-premise solutions. 

Once solid data foundations are in place, retailers can maximise the use of  advanced data analytics to drive personalisation and better understand their customers . Combining disparate datasets enables a single view of the customer and new trends in consumer behaviour start to emerge. By leveraging these data sources in strategic analytic platforms, retailers can optimise their in-store and online experiences to meet the needs of customers.
 

Carlos Aggio is an Analytics and Cognitive Partner at Deloitte Australia. He specialises in the intersection of data, Cloud, and analytics. In his role, he helps clients move towards modern analytic platforms and drive operational improvement and Cloud-led digital transformation.
 

Authors:

James Allan, Partner, Consulting, Deloitte Australia

Carlos Aggio, Partner, Consulting, Deloitte Australia

Sameer Dhingra, Director, Retail & Consumer | Industry Verticals | Google Cloud

More about our authors

James Allan

James Allan

Partner, Consulting

James is a Partner in Deloitte’s Consulting practice. He has played an instrumental role in the development and leadership of the firm’s Cloud and Infrastructure team. Several years ago, he had a vision to have a more focused technical team in the IT infrastructure space, as a catalyst to technology transformation work. James has been successful in leveraging the intersection of the team’s deep technical capabilities, and an understanding of client issues.