SINGAPORE: With the help of data and analytics, online food delivery provider Deliveroo is able to predict what consumers are most likely to order throughout the week.
According to General Manager Tristan Torres, Singaporeans tend to go for salads at the start of a new week. “There is usually a spike of six to 10 times more salad orders from Mondays to Wednesdays because people want to eat healthy after all the partying over the weekend.”
By mid-week, orders for pizzas and burgers start emerging before indulgent food items form the bulk of orders on Fridays. On weekends, the average size of each order is usually double that of weekdays “as people are usually with friends and families, and that perhaps explains why they like to have salads on Mondays,” joked the Spaniard who started the Singapore arm of London-headquartered Deliveroo last November.
This insight allows the on-demand premium food delivery service to tweak its advertising and marketing strategies with personalised content.
“Based on this demand, we arrange our display ads to be different from Mondays to Fridays. Because we know that on Mondays you’ll be more open to ordering a salad or healthy food, we wouldn’t display burgers because you wouldn’t be enticed,” noted Mr Torres, who was speaking to Channel NewsAsia on the sidelines of the inaugural Millennial 20/20 Summit held in Singapore last week.
Deliveroo Singapore's General Manager Tristan Torres. (Photo: Deliveroo)
HOW DOES BIG DATA HELP?
Big data, which involves taking both structured and unstructured data and figuring out how to incorporate them into a company’s planning, has been a common practice at tech firms. But in recent years, other sectors such as the food and beverage (F&B) industry have started to climb on the bandwagon.
For Deliveroo, mining data also plays a key role in keeping its delivery machine efficient.
Apart from constant testing of its algorithm that assigns orders to the nearest available driver, its team of data scientists evaluate orders received on its website, delivery timings and experiences to better understand various geographies.
That is why in central areas of Singapore such as the Central Business District (CBD), Tanjong Pagar and City Hall, Deliveroo deploys a team of cyclists for deliveries, instead of its usual fleet of bikes and scooters.
“What we realise in Singapore’s city areas is that the way the streets are done sometimes means longer routes,” said Mr Torres. “For example, getting from Cecil Street to Robinson Road (means) going up to One Raffles Place and coming back, which means you’re losing 2 to 3 minutes.”
Cyclists, on the other hand, can be nimble when it comes to choosing routes and that helps to shorten the delivery time by nearly 30 per cent, he added.
This finding “is a product of using data,” according to Mr Torres. “We needed to find out why Woodlands, which is much bigger than the CBD, is one of the fastest delivery zones. After working on all the data, we realise it’s the structure of the streets.”
Food start-up Grain also turns to quantitative analysis of copious amounts of data, usually customer feedback for its food and delivery services gathered online, to improve business operations.
Founded in 2014 by a group of friends, the Singapore-based start-up prides itself on being a “full-stack” company that cooks its own food and delivers it. While it started off with a daily-changing menu with an emphasis on healthy food, data analysis has enabled Grain to nail down six best-sellers that it now offers on a daily basis, as well as steer its move to serve more than just health-conscious customers.
“We started off with a ‘eat clean’ focus before starting to introduce hearty dishes with a healthier spin like our Samsui Chicken Rice and Dried Laksa… from our periodic review, we realised that customers love these dishes so that has helped us to shape our menu design,” said co-founder and director Gao Rifeng.
Grain also leverages on data and analytics to estimate demand and manage its delivery speed. According to Mr Gao, food items are usually prepared in batches in its central kitchen before being sent out to its distribution hub every hour, where orders are then delivered on bikes to customers. This minimises delivery time and enables the lean start-up to overcome chronic manpower shortages in the industry.
“Cooking only when orders come in would make delivery so slow. So drawing on our past data, we cook our food ahead in batches,” said 27-year-old Mr Gao. “When we cook in batches hourly like this, we can also be efficient in terms of manpower and that also lowers costs.”
HUGE DATA APPETITE AMONG TRADITIONAL F&B BUSINESSES
Traditional players in the F&B industry are also not far behind when it comes to using data to drive business decisions and change.
American confectioner Mondelez International, whose portfolio includes various household names such as Oreo, Chips Ahoy! and Cadbury, started work on big data in Asia-Pacific over the last two years as it worked towards its goal of generating e-commerce revenues of at least US$1 billion by 2020. The multinational company is counting on Asia to account for at least one-third of the e-commerce target.
According to Mr Ganesh Kashyap, director of e-commerce for Asia-Pacific, Mondelez is merging information from its traditional research sources such as retail stores and focus groups, with data gathered from social media platforms and e-retailers. This has helped the firm to decide how it can display its banner advertisements on online platforms – something that was previously unachievable with data from offline sources.
“With the data that we have now, we can target our ads to the profile of the shopper,” said Mr Kashyap. “In the offline world, there is no way of knowing the profile of shoppers that’s looking at my supermarket aisle right now but in the online world, we have that.”
Data is also helping the snack food giant to stay relevant amid changing demands of the millennial generation. In particular, it salvaged the problem of a waning appetite for its signature crème-filled sandwich cookie Oreo in one of its most important markets, China.
Said Mr Kashyap: “From our data, we found out that millennials between the age of 18 to 30 were familiar with Oreo, but they didn't have a connection with the brand when they make purchases online.”
Apart from teaming up with e-commerce behemoth Alibaba, Mondelez rolled out a campaign last year, which allowed customers to select the artwork and customise personal messages on their Oreo purchases. Despite having to pay triple the price, the campaign was a big hit among Chinese millennials.
An example of a customisable Oreo box. (Photo: Tang See Kit)
“Not only did it drive brand engagement levels with millennials for the franchise, it grew the sales of Oreo in the country,” said Mr Kashyap. “For a millennial consumer, if you want them to buy online, it must be a good deal but at the same time, they are also willing to pay three times the price for an experience to customise their Oreo box.
“I think that's the dichotomy that exists in the minds of the millennials and data allowed us to uncover and solve it,” he added.
Meanwhile, Pizza Hut Asia is also using data to help it predict purchasing habits based on customer characteristics and other behavioural indicators. On top of the order and delivery website that it rolled out in 1994, the pizza chain also has two other proprietary tools to measure customer feedback.
One is a online survey for in-house diners, while the other is a net-based social listening tool called Social Hive, which monitors what social media users are saying about Pizza Hut all over the world. Apart from written comments, the tool can also pick up responses based on “emojis”.
“For us, data, be it customer or transactional, comes from many sources. We know when our customers buy their pizzas, their favourite flavours and all these allow us to predict what the customer will order next time,” said Ms Karen Chan, Chief Digital Officer at Pizza Hut Asia.
NO PIECE OF CAKE
However, businesses acknowledge that relying solely on big data is no winning recipe as there could be instances where teams become overloaded with massive amounts of data, resulting in misinterpretations.
Then, there are also markets where data is not readily available.
“Not all markets in the world have the same level of maturity when it comes to data… In India, we do have social media feeds where we monitor but the actual transactional data from online retailers is mixed,” said Mondelez’s Mr Kashyap.
When that happens, some businesses turn to local business partners and employees to find its way around a foreign market.
Deliveroo’s Mr Torres said: “When we first started in Singapore, we had little information about how the people here choose their food so we relied on the knowledge of the locals we hired.
"And that's what we do whenever we enter a new market, we work with local teams and businesses, at least for the first 5 to 6 months, as it doesn't quite make sense to bring in an account manager from London."
Follow See Kit on Twitter @SeeKitCNA