Building data cubes across subscriptions, usage and marketing for a SaaS business

A private-equity-backed software-as-a-service business asked QuantSpark to solve a familiar problem: plenty of data, no shared picture of it. The result was three purpose-built data cubes and a suite of self-service dashboards that replaced manual board reporting with an automated, always-current view of subscriptions, product usage and marketing performance. Board reporting is now automated rather than manually compiled, the process no longer depends on one person's institutional knowledge, and teams across the business can query the same numbers themselves instead of waiting on an analyst.

The problem QuantSpark was brought in to fix was not a shortage of data. The business held extensive product and customer information, but it sat in disconnected systems: marketing platforms, sales records and product-usage logs that never met in one place. Nobody could see, in one view, which customers were the most valuable, which marketing channels actually drove profitable growth, or which usage patterns preceded churn. Board reporting was assembled by hand each cycle, a slow process that concentrated knowledge in whoever built the numbers and left the business exposed if that person left. The company wanted to democratise data access across teams and use it to sharpen acquisition, cut churn and grow market share, but without a consolidated data layer that ambition had nowhere to run.

QuantSpark's response was to build the data layer first and the analysis on top of it. The engagement started with a discovery phase mapping the client's existing data infrastructure: what was collected, where it lived and in what form. Much of it sat in non-relational stores unsuited to the cross-cutting reporting the board and product teams needed, so the next step was data modelling, relationalising that data into structures that could support consistent, repeatable analysis. QuantSpark then built the underlying data engineering: a cost-effective cloud storage environment and a pipeline connecting the client's digital-advertising and web-analytics platforms, so marketing performance data could flow into the same environment as subscription and product data. An open-source SQL transformation tool sat at the centre of the build, chosen specifically to keep the ongoing cost and maintenance burden low.

On top of that foundation, QuantSpark built three data cubes, each answering a question the business had previously had to answer by instinct. A subscriptions-and-contracts cube gave finance and the board a reliable, automated source for financial reporting. A product-usage cube gave the business a way to identify and predict churn from behavioural signals rather than after the fact. A marketing-attribution cube let the business see which customers were most profitable and which channels were actually delivering them, rather than relying on guesswork. Self-service dashboards sat on top of all three, so finance, product and marketing teams could each interrogate the data relevant to them without raising a ticket with an analyst.

The system in place today is best understood as three layers working together: a cloud data-storage environment holding the modelled data, an open-source SQL transformation tool doing the build and maintenance work in between, and a pipeline drawing in data from the client's digital-advertising and web-analytics platforms, all surfaced through self-service dashboards for finance, product and marketing.

The value delivered splits into two kinds. The first is already realised: board reporting is automated, which the client reports as saving significant analyst hours and removing the key-person risk that came with a manually compiled report, and data access is genuinely democratised, with teams pulling their own numbers instead of queuing behind an analyst. The second is a foundation rather than a finished number: the marketing-attribution and product-usage cubes give the business, for the first time, a measurable basis for calculating return on the customers it acquires and retains, by making high-value customers and churn signals visible rather than invisible. The client itself frames this second category as a return that will grow over time, a forward-looking expectation rather than an audited figure, and this case is reported on that honest basis rather than converted into a number the source does not provide.