Brent Hofacker - Fotolia

Pizza Hut optimises pizza delivery platform during lockdown with data analytics

Pizza Hut Digital Ventures has been improving the company’s e-commerce performance before and during the coronavirus crisis

Pizza Hut is using advanced analytics to optimise online sales during the Covid-19 public health crisis, when restaurants are closed.

Speaking at last week’s Retail Expo virtual event, Tristan Burns, analytics lead for Pizza Hut Digital Ventures (PHDV), described how the company’s e-commerce arm interrogates data to reduce friction for customers without decreasing the value of their online orders.

The digital ventures arm of the company works across Pizza Hut, which is organised on a national franchise basis.

He cautioned peers not to overdraw lessons from the current coronavirus crisis, and said customer behaviour at present might indicate some trends that might hold validity for the future, but it was too hard to be sure of that.

Burns said the current coronavirus public health crisis, from a conversion optimisation perspective, had created an abnormal situation that must be carefully watched. In the UK, the restaurants are closed and customer collection is not available, but pizza delivery is still happening. All transactions are therefore online, and all are distributed by contact-free delivery, not collection.

“It is hard to say, but it is unlikely that things will go back to exactly normal, with the same ratios of collection versus delivery as before or as now,” said Burns.

His advice for other optimisation teams was: “If you have to optimise in the short term for coronavirus, then do use your current data, but if you are thinking more long term, I would be careful not to look too deeply into the current data about customer behaviours.”  

Acting on pizza’s online potential

Pizza Hut was one of the first food retailers to see the potential of e-commerce in the 1990s. On Burns’ account, Pizza Hut has seen rivals like Domino’s Pizza stealing a march since those early pioneering days of selling pizzas over the internet.

“If you don’t have the resources to optimise for every market you are in, use your larger markets’ datasets to optimise and then scale those out to the smaller ones”
Tristan Burns, Pizza Hut Digital Ventures

“Pizza Hut is widely credited with being the first company to sell a physical good over the internet,” he told attendees of the virtual conference. That was in 1994 in Santa Cruz, with the back-end servers being in Wichita, Kansas, and was done over the Mosaic web browser, cash on delivery for the pizzas, so no online monetary transactions at that time.

The national franchise model – and Pizza Hut operates in over 100 countries – has meant the development of “an inconsistent online brand and a sub-optimal customer experience”, said Burns.

When Domino’s Pizza’s then CEO, Patrick Doyle, began describing his firm as a tech company that sold pizzas, Pizza Hut in the US, China and internationally decided to up its game – Pizza Hut Digital Ventures is part of that effort. It set out to build a “scalable e-commerce solution that could be rolled out across the globe to deliver a consistent customer experience wherever your hunger for pizza might strike”.

It started, two-and-a-half years ago, with Pizza Hut UK and then France. Today, PHDV has offices in London, Ho Chi Minh City in Vietnam, and Dallas, Texas, each serving their region.

Maximising sales without alienating customers

In the UK, the vast majority of orders are made via the Pizza Hut website or smartphone app. The goal of the website optimisers across the company is to make sure as much traffic as possible converts to sales and that the value of each transaction is maximised, but to do that in such a way that the customer journey is as frictionless as possible.

A high volume of low transactions does not necessarily lead to high profitability. However, prompting customers to add more to their baskets, on the model of budget airlines, can have a negative impact on conversions. “The challenge for us in optimisation is how we can use customer insight to not only increase basket value, but to do so without negatively impacting conversion,” said Burns.

General marketing and ensuring fast website load speed play a role here, but the data analytics team focuses on granular metrics, such as average order value (AOV). In the case of AOV, deal types, such as a certain percentage off above a certain order value, are important. They use heat map visualisations laying out AOV against conversion rate by time.

“This is a great thing [any data analytics team] can do, without looking at features of your website,” he said. “You can get an understanding of customer behaviour at particular times, and see opportunities you would not have seen before.”

Abandoned basket analysis is also an important area of data analysis and experimentation. What extra offers will reduce instances of abandoned baskets, for example. Decreasing friction while not decreasing order value is a constant balancing act.

Pizza Hut’s approach, as a worldwide food service, is to test e-commerce optimisations in large markets, such as the UK, Japan and India, and then scale out the ones that bear value to the smaller markets. 

As advice for others operating in global markets, Burns said: “If you don’t have the resources to optimise for every market you are in, use your larger markets’ datasets to optimise and then scale those out to the smaller ones.”

He concluded with the controversial proposition that “pineapple on pizza is fine”.

Read more about data analytics in e-commerce retail

Content Continues Below

Read more on IT for retail and logistics

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.

SearchCIO

SearchSecurity

SearchNetworking

SearchDataCenter

SearchDataManagement

Close