Want to know more?
Are you interested in this project? Or do you have one just like it? Get in touch. We’d love to tell you more about it.
Take only what you need at Zego
Building an insurance pricing model for ultra-flexible coverage and risk management
Zego provides flexible insurance products for individuals and businesses in the food delivery, private hire, and courier sectors.
Historically Zego had used third party insurance pricing models to establish premiums for their insurance products, but wanted to deliver a bespoke pricing and risk engine, delivered as a component to a new Zego platform. This would underpin the entire business, moving Zego towards offering contextually appropriate insurance to customers. Zego believed that understanding data and risk better (including leveraging data and learning across their datasets) would enable them to offer more compelling pricing options and reduce risk.
Equal Experts’ reputation for agile scaling solutions made us the ideal partner to support the services needed to achieve growth and sustainability whilst protecting Zego’s customer proposition. What started out as an MVP requirement immediately moved to an extended engagement to advance the solution to live production when stakeholders saw the potential business value.
By incorporating driver behavioural data into its offering, Zego can now price insurance policies far more accurately and fairly than its competitors.
Outcomes
12
weeks to MVP
6
months product live to customers
0
errors on pricing for live policies sold
About Zego
Zego provides insurance for food delivery drivers, couriers, private hire drivers, and small businesses. They offer flexible motor and commercial policies ranging from an hour to a year. Their goal is to offer fairer insurance solutions, saving their customers time and money.
Zego’s innovative insurance products are simple, flexible and usage-based. Zego has sold more than 14 million policies in the B2C insurance domain, and currently insures approximately half of the UK’s food delivery drivers and riders, providing protection for over 200,000 vehicles. People who use their vehicles to earn money now have the option to pay for exactly the insurance they need, when they need it – no more, no less.
Industry
Automotive, Finance and Insurance
Organisation size
Around 500 employees
Location
London
Challenge
Removing roadblocks in technology to pave the way for growth
As a start-up, Zego had originally adopted a traditional approach to insurance, with classic pricing and risk management that used proven pricing techniques and data sources. However, this technology was not built for experimentation and innovation, and didn’t support the business to scale. The monolith code base, mobile and website apps had accumulated technical debt over time as the drive to develop new solutions and grow market share outgrew capabilities.
Even ‘simple’ pricing changes required a huge amount of engineering effort to effect. At the start of the project Zego had siloed teams all producing their own versions of priorities in this space. As a consequence, data science activities were often blocked at the point of deployment.
Eventually, Zego realised that they weren’t able to sustain the innovation they were looking for whilst still getting the balance right between competitive pricing and loss ratio for the business.
To address this, they are taking an advanced approach by applying data science techniques, novel data sources and strong technical platforms to help them differentiate from their competition. Zego partnered with Equal Experts to build core IP technology through the Zego Pricing and Risk Platform; this will facilitate pricing excellence, and introduce better working practices across their engineering, data science and pricing teams to create a clear ownership and understanding of pricing technology.
Solution
Bite size and good enough is better than perfect for an initial MVP
Delivering the MVP for Zego was time critical. Here are some of the decisions we made that helped us deliver the agreed scope on time:
- Using Github as a pricing configuration store
- Pricing model acceptance tests helped Zego discover and eliminate bugs
- Deferred building a perfect product in favour of a usable product
- Evolving the product architecture depending on requirements
Throughout the project we fostered transparency and accountability with stakeholders using clear (high level) architecture diagrams on how we would get there. We were then able to slice the deliverables into 6 Sprints, each with a clear goal. This gave busy stakeholders a high-level, realistic understanding of how outcomes would be achieved, ensuring buy-in and realistic expectations. It also allowed the team and stakeholders to prioritise their time to ensure we met targeted timescales.
Throughout the project we fostered transparency and accountability with stakeholders using clear (high level) architecture diagrams on how we would get there. We were then able to slice the deliverables into 6 Sprints, each with a clear goal. This gave busy stakeholders a high-level, realistic understanding of how outcomes would be achieved, ensuring buy-in and realistic expectations. It also allowed the team and stakeholders to prioritise their time to ensure we met targeted timescales.
Accelerating delivery through agile ways of working
We started this project by adopting the build, measure and learn approach. This outcome-based process enabled us to measure the success of the delivery and identify the key learnings of the MVP, whilst simultaneously building and delivering the product. We pursued the following opportunities to maximise efficiencies and deliver an initial working solution rapidly:
- A consistent (and accelerating) cycle of releasing pricing improvements into production.
- Live data pricing experimentation in each insurance product line to drive improvements. We started with Zego’s VAN product then worked extensively with their TAXI product, choosing this as the launch product for the integration and release of the three new platforms (Pricing and Risk, Insurance, and Data).
- Engaging with capacity providers using experimentation and analysis to drive down approval times and build trust.
- Shared cross-functional priorities in the business units and data platform teams.
An intensive two weeks discovery phase prior to initiating inception by the delivery team helped us to define and refine the problem statements, and understand the business and technical landscape so we could scope the inception. It also established close collaboration and transparency with the senior stakeholders at Zego.
In this way we were able to deliver an MVP in just 12 weeks, launching a single prototype pricing capability into production to ensure our foundational capabilities delivered the following outcomes:
- Internal user benefit: a new product configuration store allowing pricing models to be updated quickly and accurately by the pricing teams, reducing engineering effort.
- Benefit to the business: foundation components in place to provide feature parity of the existing engine.
Stakeholders immediately saw the business value and the engagement was extended to advance the solution to live production. This involved radical changes to the existing setup, with enhancements made that would dramatically improve Zego’s customer offering.
We tested the components built in MVP in shadow production, then continued iterating the MVP and building production-ready services to unlock new capabilities like underwriting checks. This ensured scalability for more of Zego’s insurance products to start serving live B2C customers, whilst guaranteeing accuracy and performance.
Results
A bespoke pricing platform that leverages data for revenue and cost optimisation
A key outcome of putting the pricing and risk engine into live production was the ability for Zego to maximise the value of the vast quantities of data they had in their platforms. As part of the MVP we used a tactical solution for data enrichment, to establish data as the key priority for the platform and simplify data dependencies to achieve faster delivery.
Our enhancement work as we built the live product enabled the use of that data by a separate platform in the business to create smarter pricing options, both for the customer and to protect Zego against loss. Our underwriting engine assesses the risk profile of individual drivers to create personalised, behavioural based pricing options that are unique in the marketplace; it also minimises the possibility of competitive prices attracting high-risk customers, reducing potential loss.
We built a product configuration store that allows pricing models to be updated accurately, by providing acceptance test suites to the pricing teams for validation. Changes can be tested before release and as released, enabling faster and safer updates to pricing models. Zego’s time to release new features has reduced, with minimal to no engineering effort required.
An audit store provides a clear record of quotes and a consistent view of risk data for all users; 100% of the pricing model changes were version controlled and recorded in the audit store, which satisfies regulatory requirements by retaining a complete record of pricing decisions. It also serves as a foundation for explainable pricing.
Data scientists are now able to use Machine Learning in pipelines to compare historic data and test different configurations and pricing strategies without affecting live customers.
The solution is build on AWS Cloud, using EKS, DynamoDB, Kinesis data Streams and FireHose.
Adding business value through improved ways of working
This initiative was technology driven, but as we started to take the product towards live production we hit some roadblocks, and it became clear that there was a lack of alignment between business units within Zego and the various platforms they were using.
This was a pivotal moment for the business as we worked to help define a unified roadmap for all dependent services; it triggered a milestone project for Zego which unblocked challenges in serving live traffic and reducing dependencies, and successfully catalysed a change in culture and ways of working to bridge the gap, creating collaboration and alignment they had not seen previously. Senior leadership discovered that product and engineering teams needed to work together seamlessly, and Zego is now a product-driven business.
The result was a pricing and risk engine that provides a consistent view of clearly defined risk data which enables all users – from data scientists and data enrichment teams to pricing actuaries – to work cohesively with the same goals and tools.
The team (on both client and consultant sides) was universally acknowledged as exceptional by key stakeholders.
The Zego project lead said that the EE team was setting a high bar and was the “gold standard” for how teams should operate in Zego.
When the Zego pricing and risk engine went into live production, it was with zero pricing errors on live policies sold, thanks to the novel acceptance test suite now used by all Zego’s pricing analysts. We also saw an efficiency improvement, by reducing the manual steps required of pricing analysts, saving them considerable time.
Equal Experts is the only team in Technology delivering on time and operating with transparency on progress… There’s nothing we can’t throw at you that Equal Experts doesn’t respond to in a positive way.
CTO
, Zego
Conclusion
The new pricing and risk platform supports Zego to apply industry leading pricing techniques to their products, enabling experimentation to fuel innovation and continuous improvement. It means Zego can now compete in larger markets and continue to scale its business for sustainable innovation.
Zego can now take advantage of faster times to market thanks to increased speed, simplicity, and governance of pricing changes. They are able to constantly improve on existing products to stay ahead of the curve, and give pricing quotes in minimal time through their lightning fast calculation engine. The priority given to data and its integration in the new platform enables more consistent decision making across the business, putting Zego in a position to continually innovate on their products whilst guaranteeing the best prices to their customers.