Sourcing Subscription Services KPI Data — Considerations & Recommendations

For cloud and communications service providers — such as those offering cloud infrastructure, data centers, SaaS, Internet and other services — Key Performance Indicators (KPIs) are vital for measuring success and guiding decision-making.

While agreeing on and defining what should be measured, as well as determining the appropriate calculation methods (see this blog for insights on top cloud and communications metrics and KPIs), is a critical first step, it is only the beginning of creating a successful KPI system.

The next step involves deciding where the KPI data should be sourced from. This can be challenging, as the relevant information is often scattered across multiple systems. For instance, even a seemingly straightforward metric as ‘new sales’ can be measured in various ways — such as quoting (sold), orders (in progress, canceled), billing (billed), or even a combination of these sources.

This article explores key considerations and actionable recommendations for sourcing data to ensure accurate and reliable KPI tracking.

Challenges and Caveats: An Example

Let’s begin with a simple example in which we want to calculate churn based on monthly recurring revenue (MRR) using the service inventory.

The formula is:

Churn (%) = [Recurring Revenue Lost in a Month] / [Total Recurring Revenue at the Beginning of the Month]

Let’s assume that there are three active services at the beginning of the month:

SERVICE	MRR
SERVICE‑1	$120.00
SERVICE‑2	$130.00
SERVICE‑3	$150.00
Total Monthly MRR	$400.00

In the middle of the month, a customer decided to upgrade SERVICE-3, which required rip-and-replace, and caused cancellation of SERVICE-3 and building SERVICE-4 with $200.00 of MRR instead:

SERVICE	MRR
SERVICE‑1	$120.00
SERVICE‑2	$130.00
SERVICE‑3	($150.00)
SERVICE‑4	$200.00
Total Monthly MRR	$450.00

From the service inventory perspective one service was lost, so the calculation looks like follows:

Churn (%)= $150.00 / $400.00 = 37.5%

But this is misleading as no business was lost whatsoever! Instead, zero churn and $50.00 of revenue uptick should be reported.

So, where does the problem come from? Well, the service inventory on its own does not know about the relationship between SERVICE-3 and SERVICE-4.

Let’s take a look how this could look on an order:

ORDER	LINE ITEM/OPERATION	SERVICE	SERVICE MRR
ORDER-1	LI-1 / REPLACE
		SERVICE-3 (REPLACED SERVICE)	($150.00)
		SERVICE-4 (NEW SERVICE)	$200.00
		Net MRR	$50.00

This is a much more usable information structure for churn calculation — looking at a list of active services month-to-month alone (without orders), it is impossible to obtain accurate churn numbers.

Recommendations

As illustrated in the example above, sourcing KPI data — even for straightforward metrics — can be challenging. Below are a few recommendations to avoid major problems that we developed during our customer engagements.

Look for Natural Data Sources

Most KPI definitions emphasize that a KPI should be both ‘available’ and ‘measurable’. This principle also extends to systems, making it important to identify the most ‘natural’ system for sourcing a particular KPI, such as:

System	Example KPIs
CRM	Sales pipeline, sales bookings, etc.
Order Management	Churn, compression, order cancellations, etc.
Service Inventory	Total MRR, MRR by product, etc.
ERP or Billing	Gross revenue, EBIDTA, etc.

Re-Examine KPIs That Are Difficult to Obtain

If a KPI doesn’t have a “natural” data source, it’s a potential red flag that could indicate issues with the current data model or systems.

In some cases, adjustments may be necessary. For instance, in Salesforce, this might involve creating a new formula field or implementing a trigger-calculated field to capture the required data effectively.

Another sign of a problematic KPI is a calculation that relies on numerous if/then/else conditions or exclusions. A well-defined data source should allow for straightforward data retrieval — akin to a simple SQL SELECT statement — without excessive complexity.

Prioritize Value Over Ease of Implementation

Sometimes, ease of implementation drives the choice — a few KPIs are sourced from a system already and it is easy to add one more, even though it is generally agreed that it should be sourced from somewhere else.

This usually leads to problems down the line with more conditions and data transformations, usually leading to eventual re-design & re-implementation.

Be Open to Adjustments

Organizations and their systems naturally evolve over time. Startups, for example, often begin by tracking their operations with Microsoft Excel spreadsheets but eventually transition to more sophisticated IT systems as they grow and require tools better suited to their scale and processes.

While a perfect data source might not exist initially, this can change as new systems are implemented. Leveraging new systems as data sources as they become available typically pays off quickly.

Conclusions

Building good KPI reporting can be challenging but is essential for any organization to be successful. Many decisions need to be made along the way, with data sourcing being one of the most critical ones.

Even when most data resides on a single platform — such as Salesforce CPQ, Orders, Service Cloud, Billing, or FinancialForce — the process still requires careful preparation. While fewer systems typically reduce reporting implementation and maintenance costs, the need for proper planning remains vital.

Nextian has extensive experience building KPI, reports and dashboards for cloud and communications providers, using Salesforce, Tableau and data warehouses.

Our Reporting & Analytics module contains 80+ pre-configured dashboards and reports featuring real time updates and pre-built, best-practice quote-to-cash metrics for subscription services.