As click and collect becomes a more ubiquitous fulfilment option in the world that Amazon built, there is a growing need for parcel pickup locations. The challenge for collection and delivery service Doddle has been to ensure that staff at collection points dotted across the UK have up to date customer order data on their mobile apps - requiring a stable internet connection.
Founded in 2014, Doddle originated as a way to utilise train station space better and to facilitate click and collect for time-poor commuters. Now Doddle has 84 locations, allowing customers to collect and return online shopping from its stores at convenient times. Thanks to a host of partnerships with key retailers like Amazon and Asos, shoppers can select a Doddle location as their delivery address straight from the checkout.
The company wouldn't have been able to achieve this sort of location growth in such a short amount of time without the technical help of the Couchbase's NoSQL database software, according to Doddle CTO Gary O'Connor. He told Computerworld UK that setting up a reliable internet connection was vital to serving customers and setting up new stores. Patchy network connections could severely limit its speed to market.
However, O'Connor had found that the units Doddle used in train stations and supermarkets such as Morrisons didn't have a BT phone line installed. To get the internet connectivity required to keep operations going, Doddle was installing expensive ad-hoc solutions such as 4G routers. The problem with 4G, especially in train stations, is that it pinches at peak times when commuters are flowing through the station, which is also peak time for click and collect.
O'Connor identified the agility a NoSQL database brings, alongside the handy offline capabilities, as key differentiators when it came to building mobile apps on top of Couchbase.
This drove O'Connor to completely re-engineer its two staff-facing mobile apps, one for store staff and one for logistics, so that it "writes to the local system. We have a Couchbase database on device, namely an Android app" installed on Zebra mobile devices.
Now staff can scan hundreds of parcels an hour at peak times, which will update local records on the device and Couchbase with automate the sync back to Doddle's cloud backend when it has a connection. If there is no connection it simply queues the data.
This sounds like pretty simple piece of development work, so why did it need to be an off-the-shelf commercial solution?
"Traditional databases didn't really have that sync logic built in," O'Connor explained. "We could do it, but Couchbase had solved that already."
Doddle was finally convinced to opt for Couchbase when it saw that an airline customer had built a similar capability for mid-air transactions that would sync when the aircraft returned to the ground.
Now, instead of expensive internet connectivity hardware, Doddle used a software solution, reducing not only the capital expenditure of opening new locations by up to half, but also the deployment time from six weeks to two weeks or less, as it doesn't have to wait for broadband lines to be installed before opening.
O'Connor said that the NoSQL database also allows the organisation to feed in various data sources from a plethora of logistics and retail partners without having to amend data models all the time.
Doddle is now looking to leverage Couchbase analytics to give store managers quicker access to key KPIs around customer's walking away with their parcel within two minutes.
Instead of maintaining a secondary analytics tool on top of another database, doing the analytics within Couchbase will drive better real-time insight. O'Connor explained: "What they have done with [programming language] Nickle is make it easier to allow us to pull those real time stats to know day to day how we are working. Amazon really drives that operational reporting."