Scratchpay Saved Time and Efforts by Automating their Data Preparation Process Using Hevo

Chris Lockhart, Data Science Manager at Scratchpay uses Hevo to streamline their
data pipeline process and improve data reliability.

Get a Demo

The credit risk team at Scratchpay is comprised of data scientists, analysts, and operations experts. Chris Lockhart works as a data science manager in the credit risk team. He and his team look after all the risk-related aspects of loan management, including loan pricing, approvals/denials, and operations processes including manual review. The overall objective of his team is to build predictive models for consumer risk and monitor the evolving loan portfolio.

Chris Lockhart

Data Science Manager

Data Challenges

Scratchpay was founded in 2016 with a goal to offer pet parents the care they needed using fair, affordable, and transparent payment plans. So far, the company has offered financing through its service to over 100,000 pet owners.

Scratchpay makes borrowing extremely simple through its digital platform and it takes just 60 seconds for anyone to sign up on their platform, choose the plan, and get the care that they need. Their underwriting method uses machine learning to provide on-the-spot approvals for pet owners who sign up online usually through a phone or desktop. Their ML algorithms are heavily dependent on the proprietary pet data that they capture at the vet and in the application.

Chris says, Data is embedded into Scratchpay’s core and is used across all teams in varying capacities. Most of our data is scattered across multiple systems and, historically, we used custom in-house pipelines to integrate this data. The biggest challenge with our pipeline solution was that it wasn’t streamlined and involved a lot of ad-hoc and manual processes.

The credit risk team at Scratchpay is completely dependent on the data to build consumer risk models. And these models are the backbone of their entire lending business. Due to the manual process involved in the data preparation stage, manual errors and delays were the obvious things. But the team couldn’t afford to make even a single error in their algorithm due to unreliable data. So they decided to look for an alternate solution that can enhance their data pipelines and be extremely reliable.

The Solution

Chris and his team came to know about Hevo while searching for the most popular data pipeline tools available in the market. They tried Hevo along with several other tools and finally decided to go ahead with Hevo because of the wide range of integrations, ease of usability, Python integration, and the quality of data output.

Chris says, For a data pipeline tool, we expect to be able to seamlessly integrate our data sources with custom scheduling, logging, and alerts. Hevo provides all these features and allows us to get access to data when we critically need it.

Chris describes the Before Hevo situation as The Dark Era. Prior to using Hevo, their workflow was best described as ad hoc because of a lot of aperiodic data dumps and uploads. After switching to Hevo, the credit risk team at Scratchpay was able to eliminate the manual efforts in creating the pipelines and drastically reduced the data errors.

Using Hevo, Chris, and his team built their advanced data pipelines that aggregate their data across multiple sources such as MySQL, Postgres, Google Analytics, their payment service, and Zendesk into BigQuery. Using Google Cloud as a single source of truth has worked greatly for Scratchpay and it allows members across all the teams to access data through BigQuery, Google Sheets, and Data Studio.

Hevo Scratchpay Data Stack

We love how easy it is to get started with Hevo, the vast amount of built-in integrations and the visibility we get into how data is flowing through our pipelines is simply amazing. Also, Hevo’s integration with Python is very exciting for us for the credit risk team as this allows us to queue up custom scripts and retrain our predictive risk models on the fly,  says Chris.

Key Results

Using Hevo, Scratchpay dramatically improved data integration. Automation has greatly freed up employee hours to work more on the analytic part of data as compared to its preparatory steps. Hevo has become integral to the core of Scratchpay and has easily improved its throughput by over 100%.

Chris says, Our customers feel the impact of all over these data-driven efforts, from stream-lined application flows to proprietary credit models to loan management and servicing. This modern data stack and the aggregated data has also helped our internal teams in multiple ways - the finance team has been able to link loan performance to the conversion funnel to understand consumer price sensitivity and increase loan originations marketing and partnerships team has used payment data to attach lifetime value to customers acquired through various marketing channels; engineering has saved countless hours per week by eliminating manual efforts.

After establishing its strong presence in the pet-care market, the company is now expanding its portfolio by offering lending services for dental, optical, and chiropractic care for humans. The expansion into the human health and wellness segment accelerated because of the cresting wave of the COVID-19 pandemic. The company now boasts over 1000 human health and wellness practitioners on the company’s financing network and is also constantly growing its pet care segment. Hevo is proud to be a part of their journey and continues to be a catalyst for all their data-driven projects.

Excited to see Hevo in action and understand how a modern data stack can help your business grow? Sign up for our 14-day free trial or register for a personalized demo with our product expert.