August 13, 2021

Scaling through data, without engineers

  • Industry B2B hospitality
  • Location London, UK
  • Funding Round Private
  • Data Stack kleene, Snowflake, Metabase
OrderPay Case Study | kleene.ai

OrderPay is an order and payment app for the hospitality sector. Concepted prior to the outbreak of Covid-19, the app launched on the 4th July 2020 when hospitality venues initially re-opened following the pandemic and ordering at the table, via an app, became ubiquitous. The multi-tenanted app re-skins to the branding of the venue, using geolocation and beacon technology. As a growing business producing vast amounts of data, the team looked to utilise this to make product and user improvements and manage internal reporting to the executive management and board. Data was also necessary for due diligence and investor conversations. With only production databases in place, there was no way to connect the data and utilise it for BI and reporting. The team decided to implement a data analytics infrastructure with a cloud warehouse and ELT tool. 

Success criteria

Speed of setup, ease of use and no requirement to utilise the app development engineering team for the set up or maintenance of ELT.

Challenge

Deriving value from data both internally and externally, optimising the product to meet the needs of customers, vendors and brands, without a data team.

Solution

Implementing a modern data analytics infrastructure without engineers.

Outcome  

Data driven team, utilising data insights to fuel activity and growth.

Results 

  • Data analytics without engineering 

The Chief Data Officer joined OrderPay as the only data professional in the business, with limited resources in place. Whilst the production databases were populated with app data, it was impossible to utilise this data in an efficient manner for effective reporting. Establishing a data warehouse infrastructure was the top priority, however with the development team dedicated to building the app, limited resources were available to build a data analytics infrastructure. 

“The first consideration was build versus buy. It quickly became apparent that if we did it in house, anything that I needed would always be competing with the developers time for app development. I’m not a data engineer, so I wanted a tool that was easy to use single handedly and by future data hires. Coming from an analyst background, the fact that kleene is SQL based meant that I didn’t need to spend time learning a new coding language.”

  • Predictable cost of ownership as the business scales

As a scaling business quickly acquiring new customers and high volumes of data, cost of ownership was a key concern for OrderPay. kleene’s pricing structure doesn’t penalise growth; charging per connector, not row of data. It’s far easier to forecast the future cost of a tool based on additional data sources, than to predict how many rows of data may be produced.

“Keeping a lid on the costs was a major factor when reviewing our ELT options. As our customer and user numbers rise, the amount of data that we produce will increase. As a result, I wanted a tool that I could predict the costs of accurately. A lot of other ELT tools are priced per transaction, or row of data. The fixed fee per connector was really appealing – it’s a much easier way for us to be able to predict how much we will be spending on this tool in 6,12 or 24 months’ time, which is important as we scale.”Understanding business metrics 

  • Providing business insight at speed 

Prior to establishing a data warehouse, the CDO spent hours carrying out manual tasks in order to have a view on customers and partners. Without a data analytics infrastructure in place, it was difficult and time consuming to answer ad hoc business requests. Each time a question was raised, a cost benefit analysis of whether the question was worth the manual labour was raised. Now with all of the data centralised, reports can be built with ease and business questions answered quickly. 

“Whilst we had production databases in place, those databases couldn’t talk to each other. As a result, reporting was limited – if I were to try to report on a specific venue, I would have to pull data from separate databases. This led to a horrible process where I created a list in Excel of LOOKUPS and then copy and pasted 90 pieces of code into our reporting tool, to be able to report on venue level sales. Every time we added a new venue, I had to amend the 90 pieces of code manually. We had new venues onboarding every day, resulting in my spending 5-10 hours a week, repeating the same mind numbing exercise, just to make sure that we were keeping on top of our reporting.” 

  • Data democratisation driving business activity 

kleene enables business users to access and utilise their data to drive value and growth. Empowered by data and insight, business functions can plan, strategise and ensure they are proactive in their activities, as well as reactive. With dashboards built in Metabase BI tool, the CDO could provide insight to teams and give them the flexibility to play around with the data themselves. 

“With the infrastructure in place, I can respond to requests and provide value to teams. It now takes 5-10 minutes to complete what would have taken hours to get the same result. This has given the business better access to data, so now we’re able to make use of it and be data driven. We’re much more nimble in our approach – responding to trends in marketing, using the data for PR activity and sharing opportunities with operators.  

OrderPay Testimonial

“I’ve loved every interaction that I’ve had with kleene. I don’t feel like it’s a transactional relationship, I feel like there’s a genuine understanding of each other’s business. We trialled kleene and some other vendors and it was chalk and cheese in terms of the level of engagement, customer care and attention. Before starting this project, I was quite daunted, so having the guidance and support from the team made a huge difference. The consultancy provided throughout has made everything so much easier to manage.”
kleene.ai