Fish, Diapers, Grapes and Data: The RedMart connect

Pratima H

SINGAPORE: It is a Singapore-based online grocery marketplace. But there’s more to it beyond the simple nameplate of 'just another online grocer'. It is young, offers on-time guarantees, intends to deliver fresh goods in the trickiest of sun-kissed alleys, offers competitive prices on a wide range of groceries and household essentials, and analyzes customer behavior with a different personalisation lens.

It was founded in 2011 by Roger Egan, Vikram Rupani, and Rajesh Lingappa, and has come up to the level of a 100,000 square foot warehouse, regular stocks of over 16,000 products, and understandably, banks a lot on fast, updated insight into inventory and shipment data.

How much can data do though when one wants to promise on-time guarantees or when complex algorithms and contact demand numbers have to be orchestrated to give a personalised yet on-time experience for online deliveries?

Rajesh Lingappa, Chief Technology Officer, RedMart ships out some hands-on insights. Don’t worry; he also unscrambles some hot industry areas like fluid data, contextual data, pricing precision, personalised notifications and more.

Tell us something about the role of technology, specially leveraging data and analytics in areas like Customer experience, guarantees, same-day/last-mile/two-hour/incentivised deliveries, customer service cycle, and operational edge.

Data and analytics have played a key role at RedMart, empowering employees and business users alike to gain insights into their operations and make strategic decisions based on real-time data from their business verticals. More importantly, by increasing visibility, each team has more accountability, driving more productivity and better decision-making.

For example, our Customer service team relies on Skyline Reports to understand different customer touch points across channels and understand the behaviors and pattern. It allows the business owners to understand the ‘contact demand’ which shows the volume of each contact category in each period, each category having an owner among the executive team. This brought in accountability and ownership among different departments in delivering the best service to our customers. This also allows us to simplify, leverage or eliminate certain processes across departments to minimise the touch points.

And how does that bring an edge?

We have built our own clickstream data pipeline to understand the in-app and website behaviour to understand the detailed path/interaction sequences that users take on our website and apps. This dramatically increases our ability to mine insights from customer interactions, and create true competitive advantage. We are in the process of leveraging this data to build personalised shopping experience to our customers.

We monitor real time GPS data from our last mile delivery fleet to deliver goods to our customers in the two- hour delivery window. At RedMart, we aim to save time and money for the important things in life for our customers. And when we don’t deliver your groceries on time, we’re not fulfilling that mission. That’s why we have an on-time delivery guarantee. This means that if we do not deliver your groceries in your chosen two-hour time slot, we will automatically credit $5 to your RedMart account. We’ve used complex vehicle routing algorithms to ensure optimal set of routes for a fleet of vehicles to traverse in order to deliver to a given set of customers.

How does this work on the warehouse side?

At our warehouse, we’ve used real time data on picking to create a leader board. This allows the shift managers at our Fulfilment centres to have the instant knowledge of what’s happening and also what is likely to happen. Real time data analytics gives them an option to have that power, and can plan their operations smoothly.

Anything on the fronts of CRM, retention and pricing strategy?

We also use data and analytics extensively to keep our prices of products low and attractive for our customers. Price is one component of the retail experience that is in a constant state of change. The product prices are lower than our competitors for selected categories and offer a price-match guarantee in case of a price difference. We use quantitative approaches on our data-sets for measuring and monitoring discounting policy. Price-match guarantee is one of the approaches of creating a unique customer experience where customers perceive the pricing to be fair.

The performance of our suppliers, and our relationships with them, has a direct impact on the performance of our business. At RedMart we monitor the health of these relationships regularly by collecting and managing the data we need for a successful Supplier Relationship Management program. We create and share ‘supplier scorecards’ and set of KPIs to monitor and record supplier performance. This has helped us in negotiating better contracts and passing the benefits to our end customers.

You recently cut analytics time by 70 to 80 per cent with Tableau & Amazon Redshift. What were the drivers and the outcomes of going with this solution? How well does it align with the strategic big picture? Any challenges?

Building a Data warehouse and pipelines internally allowed us a near-real-time access to data. By aggregating data from several disparate systems in the organization and allowing it to be reported and visualised, we never had access to as much data as we do now, allowing cross-functional analysis and decision-making. Earlier the SaaS-based solution allowed the data to remain in silos and didn’t allow real time access to data, not to mention the lack of flexibility in ad-hoc analysis and reporting.

We constantly culled data from the web and mobile channels. But business users didn't have easy access to the data. Report generation often took days and what was more worrisome was the part that the information was usually stale by the time they were delivered. There was a clear need for something that was easy to implement, but powerful enough to handle their needs.

Further, RedMart’s business intelligence (BI) team wanted to keep close track of customer behavior on the web and mobile channels and they had a sense that data could empower them with insights necessary to deliver personalised experiences, determine optimal pricing, and improve customer service.

So what was the 'but'?

On the ease and fluidity parts. Irrespective of the part that the team had the data they needed, they had no mechanism in place to get timely insights that could guide business decisions. Most of the data was in different silos too and that meant that the reporting mechanism allowed for only one report at a time from each data source. Even worse, business users had access issues when it came to grabbing the data themselves. Instead, the IT department was required to handle report generation.

But that has changed now?

Yes, we have cut down the time spent on analysis to about one fourth of the original, gained real-time visibility into data, facilitating quicker decision making and reduced report creation timelines from days to hours. Tableau allows us to create reports, visualisations and dashboards in an intuitive way and Tableau’s known ability to put actionable information in the hands of line-of-business users quickly, helps the business teams. By providing self-service access to data and the ability to perform custom analysis on-the-fly translates business users into empowered decision makers.

The gains for start-ups which are growing at a rapid pace are better decisions at a faster pace. With timely data available directly and regularly and shared with everyone at RedMart, we have aligned our goals and not restricted it to the executive who required data to effectively manage the business; or a data specialist—a business analyst who gathered, analysed, and reported the numbers for management.

We managed to decentralize decision-making and increase responsiveness, thus empowering more business stakeholders by putting meaningful data at their fingertips and essentially democratizing the data.

How well has technology been used with what you have attempted with Marketplace, SPRING, and new expansion plans?

RedMart and Spring, a governmental agency responsible for enabling enterprises growth, are working together to help SMEs expand their businesses to online. As part of this program we decided that the best way for companies to grow it is to have access to self-serve tools and real time sales data so that they could take action instantly.

RedMart is providing these tools by building an in-house platform which we call ‘Partner Portal’ and simplifying store management to a point where companies can have relevant growth in topline without the need to increase headcount. So technology at RedMart is enabling sellers to scale their businesses without increasing headcount at the same proportion they would have to in their offline businesses.

How has India shaped up so far for your goals?

Well, 35 per cent of the tech team is in Bangalore focusing on building next generation e-commerce platform. The Bangalore team is an extension of our Singapore office, as opposed to being merely an offshore development centre. We have a talented team of engineers to scale our operations and continue to innovate along the way. We also have several back office operations setup in Bangalore giving us the leverage to access wide talent.

What's your view of predictive analytics, unstructured data, streaming/fluid data etc?

Managing high-speed data streams generated in real time is an integral part of today’s applications. Real time streams maximise the data value when processed and acted sooner. We’ve implemented streaming data architecture to collect events from our website and apps, internal apps used in warehouse, last mile delivery apps and sensors.

The data world has been revolutionised with the new age tools supporting unstructured data and real time stream processing. At RedMart, we have been leveraging AWS products like Kinesis, S3 and Redshift to handle large volumes of variety of data and using it for real time alerting and dashboarding. Some of the data is fed back to our applications to build contextual applications.

Is contextualizing data a bigger challenge than collecting it?

You've probably heard this before, content is king but context is God. Contextualization is crucial in transforming senseless data into real information – information that can be used as actionable insights that enable intelligent corporate decision-making.

Any gleanings from the app, My List features etc.

‘My List’ is an important feature in grocery world as the average number of items purchased for each order comes to over 20. It allows users to choose based on frequency, category or when it is on sale or on the basis of recency. Our aim is to make shopping easy. Just click, click, done! We also have a handy feature on RedMart apps and website where the customer is notified the moment their favorite household essentials are on discount or back in stock, So that they never miss a sale again!

Are personalisation approaches, targeted offerings, insight-based-modelling contradictory to privacy and intrusion concerns that shoppers have?

Personalisation algorithms can use customer id which is a unique identifier and doesn’t involve any personally identifiable information and is not sensitive. Targeted offerings and user profiling with big data techniques is advantageous for providing better services to customers with more relevant products and offers. Personalisation also improves the impact of a given service. Ethical and trustworthy service prevents the disclosure of private information maintaining anonymity and also designs systems preventing data breach.

What else is on your radar in terms of future plans? More personalisation, Targeted offerings, apps on the anvil?

We are in the process of leveraging clickstream data to build a more personalised shopping experience across platforms and features like personalised homepage, targeted campaigns, targeted promotions etc. We are also working on changing the In-app and Website search to a more personalized search, helping users to more quickly find the information they are looking for based on their historical purchase, dietary preferences and brand affinity.