Keboola helps set up data mesh architecture that can handle large amounts of analytical use cases at once.
Česká spořitelna is the biggest Czech retail bank with 4.5 million clients across 400 branches.
Running a bank of this size brings its own data challenges from strict regulatory compliance via a wide range of data management needs, to almost limitless product possibilities within the data-rich environment.
To better understand how the largest Czech retail bank uses data to run its operations and grow its business, we spoke with data architects from the engineering and platforms department at Česká spořitelna.
Banks are highly regulated companies. So, data security and privacy are of paramount priority.
Any data storage, access, sharing, and collection from third parties has to be highly regulated, monitored, and follow strict data governance policies.
Before Keboola, Česká spořitelna managed the entire data lifecycle by hand on-premise. They self-managed their Oracle database instance and all integrations to website data and external data sources.
This setup caused a lot of technical overhead and delayed product developments and deliveries.
Reporting, too, must follow the same data governance policies as data storage.
In this mindset, the largest retail bank was battling against what is called “Excel kingdoms” — each department was a data silo, creating its own Excel reports with varying standards, data definitions, and unaligned business metrics. Cooperation between the Excel kingdoms was rare. Too often, the big picture was lost in the email chains of Excel files.
In this wide and fragmented data environment, people with good ideas rarely acted on them.
Implementing a data experiment, trialing a new data-driven product, or simply running unconventional analyses required both perseverance and technical know-how, which deterred a lot of people from following through on their ideas.
Česká spořitelna decided to change and standardize their data operations with Keboola.
Keboola is the data platform that unifies data operations across the data lifecycle. The platform’s components are used to:
By using Keboola, Česká spořitelna has streamlined long-tail analytic use cases.
Česká spořitelna runs on a conservative company model. Any change in the existing data operations involves a lot of red tape, checks, and balances.
Just integrating a new data source (for example, credit card information for risk modeling) requires multiple departments to get involved and a lot of bureaucracy to make sure regulatory compliances are met.
Ultimately, any new addition or change to the data assets can take weeks, even months, to come through.
By relying on Keboola’s easy-to-use components, their support teams relies on the platform doing the heavy lifting. Components are configured only once and data is automatically extracted, transformed, or saved in a secured and governed manner.
Before Keboola, retrieving data took weeks, maybe months. With Keboola, the team gets the same results in hours or days.
By streamlining data operations, the data teams at Česká spořitelna can move faster and deliver business value within a fraction of the previous time.
Before Keboola, product development was difficult.
Česká spořitelna houses some of the brightest minds in the field. But, being a financial expert doesn’t mean you’re also an expert on data.
Some people at Česká spořitelna had a business idea in mind but couldn’t validate it. Ultimately, they didn’t know how to share their idea with the company in a tangible way. Data was managed by the IT department for the IT department and was not widely accessible.
To empower the employees, the data architecture of Česká spořitelna was redesigned as a fully-governed data mesh.
The company now has a central data warehouse that contains the data various departments need to excel at their jobs.
With this setup, every department can take care of its own data needs without enlisting the help of a centralized data team.
Keboola offers drivers that empower users to access data themselves. Need data that isn’t based on the centralized platform? Simply use one of Keboola’s Extract components. For data experiments, users can whip up a secured sandbox and not worry about security and data-sharing issues.
The self-service nature of Keboola allows Česká spořitelna to re-organize how work is achieved in the bank. Experts from non-data fields can use Keboola to answer their business-based questions.
So, let’s look at a couple of concrete business problems Česká spořitelna solved or is in the process of solving since introducing Keboola.
Česká spořitelna uses multiple levers to run its marketing from paid advertising (Facebook, Sklik) and search advertising (Google, Seznam) to acquisition mechanisms on owned websites (tracked with Google Analytics 4).
All the channels work together to provide Česká spořitelna with leads they can later contact and turn into paying customers.
Before Keboola, the company experienced a massive problem with lead generation. Different reporting standards of each platform, missing data, and other operational hurdles prevented Česká spořitelna from understanding which channels bring in the most affordable and qualified leads.
Keboola automatically generates daily reports in Oracle to help organizations understand where their marketing investments generate the most leads.
The platform also helps with the entire integration pipeline — extracting data from marketing platforms (Google, Facebook, Seznam, Sklik, and Google Analytics 4), cleaning it, aggregating it, merging it across data sources, and formatting it into an easily-digestible report.
With this report, Česká spořitelna can determine which channels outperform others and by how much they outperform them. This data helps the company reallocate its budgets to better impact lead generation.
But, the data isn’t just used to better understand lead generation and the company budget — it’s also used to personalize marketing.
By collecting customer data from multiple resources, Česká spořitelna can build more accurate customer and lead profiles.
For example, Česká spořitelna can tell if a customer reacts more to a blog on marketing travel insurance or is spending time comparing loans on their website for first-time home buyers.
The data doesn’t only come from owned digital assets, but also from credit card providers, marketing platforms, and other third parties — sources that previously had to be integrated manually.
Thanks to Keboola, company data can be collected almost seamlessly to build customer-enriched profiles in Oracle. The same data is piped into the Salesforce cloud, where marketing campaigns are enhanced with personalized customer data.
The entire pipeline helps Česká spořitelna automate personalization at scale.
The customer data from multiple touchpoints alongside the customer journey bestows the team with laser-focus to target the right audience and personalize communication with each customer. As a result, the company can fine-tune its targeting, offering banking products that are relevant to individuals.
Automated personalized marketing at scale is performed against the backdrop of customer privacy.
Data collection and advertising have to follow strict privacy rules. Sure, personalized communication can entice your customers, but not when it infringes on their consent.
Keboola helps Česká spořitelna apply rules that comply with privacy regulations. On one hand, Česká spořitelna uses Keboola to navigate fine-grained data-collection rules — the platform only collects information for which explicit consent was given. Additionally, any personally identifiable information (PII) is masked, complying with data governance and privacy regulations.
Česká spořitelna will be extending its banking products on its digital assets in a more digitized way.
To prevent any wrongdoing or harm to small-loan applications, Česká spořitelna will automate credit risk scoring, streamlining the process of loan approval or rejection.
Keboola’s role will be to feed Česká spořitelna customer data into the credit risk scoring algorithms that are used to scale their back office operations.
The Keboola data stack as a service platform helps you automate end-to-end data operations.
Streamline your data collection from various sources. Use transformations to clean, aggregate, filter, and reformat your data. Send your data to data warehouses and data lakes. Automate all your data operations end-to-end while keeping the data secure and applying the strictest data governance policies.
In the words of the Česká spořitelna team,
“We would recommend Keboola mainly because, if you've got an idea, you don't have to worry about the technicalities around it. Keboola will give you the technology to test your idea easily.”