foodpanda CTO: Why autonomy is important for developing agile tech teams

Gone were the days where deliveries took days to fulfil. Today, speed is of the essence with advancements in technology to streamline delivery processes.

Enter quick commerce (q-commerce), the third evolution of e-commerce.

foodpanda has grown from its humble beginnings in 2012 to be among the largest food delivery platforms in Asia with presence in 12 markets, the latest being its expansion into Japan in September.

Managing an equally diverse tech team is Benjamin Mann, CTO of foodpanda. In the first part of the interview with e27, Mann shares his experience in dealing with the multiple challenges thrown at his tech teams during the pandemic, how he sets them up to deal with them and the importance of balancing customer feedback and data.

Edited excerpts:

The evolution of q-commerce (Photo credits: Delivery Hero)

How has foodpanda responded to changes brought about by the pandemic from a tech perspective?

Food delivery platforms are incredibly tech-intensive and a lot goes on behind the scenes — from placing your order to receiving it. If you look at the volume of steps that need to be taken in perfect sequence for someone to get food delivered within 25 to 30 minutes on a global scale as we operate, it’s a beautiful and elegant puzzle.

What we did prior to the pandemic which we are doing more now is that we’re constantly running many A/B tests to aid us in adjusting our tech solutions to the constant micro-changes in user behaviour.

With the volatile situation of the pandemic, it often requires us to make micro-adjustments in a particular vertical, framework or country. So we run numerous tests and evaluate the metrics that come out of it.

We are also trying to see how we can take an idea which was successful in country A to implement in country B by conducting small controlled tests on the local population.

Has the pandemic presented any new challenges for the tech team to deal with?
The pandemic has certainly thrown us numerous curveballs. For example, more customers are preferring contactless payments due to hygiene concerns triggered by the pandemic. Therefore, COVID-19 has certainly accelerated trends like that and we were forced to quickly adapt.

Other issues involved are the onboarding of merchants onto our platform virtually and ensuring this remained a smooth process for them. We have dedicated a whole team to work on making the process as seamless as possible by exploring how we can introduce certain features to reduce the steps required.

Our riders were also facing issues. Overnight, places were getting shut down and curfews were imposed. The remote working trend has resulted in certain areas having a higher demand now as they have become residential areas. On the other hand, office areas are seeing a drop in demand.

How do you set up your teams to deal with the large magnitude of changes?
Food delivery is a complex business consisting of many moving parts that are often out of your control. For example, if it’s raining in Singapore or there is a parade somewhere in Thailand that blocks all the roads of the restaurant that you want to order, how does the rider make an on-time delivery in such circumstances?

Therefore, even before the pandemic, we had operated in a volatile and constantly changing environment. What the pandemic has done is that it has put some of these changes on steroids!

As a team, we are set out in a way that allows us to quickly move people from one focus area to another. From a technical point of view, what we needed to do is quickly adjust certain parts of our systems to deal with the increased demand.

For example, you would be seeing things like high spikes in demand right before the curfew starts because everyone wants their food delivered before riders can no longer go out. I often refer to such spikes as “micro-seasonality”, which had existed before the pandemic however they are greater now.

How has the multitude of changes altered how your teams operate?
It required our engineering and product teams to rethink how can we scale up and down quickly to cope with the increase and fall in demand. It has also forced us to fundamentally rethink how we are building our architecture.

Fortunately, we didn’t have to overhaul our architecture overnight or do rewrites as we were fast to adapt to the changing demands as our teams operate in a relatively autonomous way and changes are responded to quickly.

Our engineering and product teams have the liberty to amend their road map and prioritise what they feel is more important. This way, we can stay agile and make small changes and adapt instead making delayed decisions that would have necessitated an overhaul of our systems.

How do you balance data and customer feedback when building a product?
At foodpanda, we are extremely obsessed with metrics. We have dozens of dashboards in real-time or near real-time where teams can see the impact of smallest changes to the customer behaviour of a certain segment.

Therefore, customers sharing their feedback manually through our channels means we are too late in identifying their concerns. Instead, we should observe these metrics and identify changes in customer behaviour and decide on how we should respond.

If the metrics are moving in the right direction, it signals that we are doing the right thing and should continue. However, if it’s moving in the wrong direction, that’s when we need to analyse the data, run tests within control groups to identify the root cause of the issue and resolve it.

We try to focus more on data instead of relying purely on customer feedback. Feedback remains important because it does come from areas where we don’t have the right metrics to measure them. This is the space where we would value feedback instead.

In areas where we have the right metrics, we want to know that something is negatively impacting the customer before they tell us.

Taken from e27: https://e27.co/foodpanda-cto-why-autonomy-is-important-for-developing-agile-tech-teams-20201218/