​Moving from assumption based to consumption based

An element of our TBM Office roadmap is to become a Trusted Advisor which initiated a review of our allocation metrics to ensure clarity and fairness are being maintained. We created a tooling team, collaborating with the Software Asset Management team, to identify licence consumption data using Flexnet.

Problem Statement

As part of this review, it was identified that certain Services are grouped together and use total headcount numbers as the metric when allocating to Business Units. Although this initially seemed like the fairest allocation method at the time, with the introduction of licensing information from Flexnet we have been able to establish that this is no longer the case.

The project specifically focuses the consumption data of Micro Focus

Method for Proof of Concept

The Micro Focus data contained in Flexnet is grouped into several categories such as employee IDs, emails, domains, and division; this data helped to identify the consumer of the licensing.

The first step was to extract consumption data and identify the users of each Apptio Business Units

The Second step was to compare the original headcount allocation with the user licence consumption allocation

Final step, if allocations differ greatly, is to complete the development work to use Flexnet consumption.

Data standardisation and other limitations

As with the introduction of new data sets there is a requirement to standardised and clean up the data. Here we outline some of these and the limitations in the data.

When grouping the Flexnet users to one of the Apptio domains, we faced a few challenges. One of which was differing naming conventions between Apptio and Flexnet, so had to drop a lower level using the domain or even their division. Any users which were not within the Apptio domains were not included in the cost allocations.

In addition, several users had no other data associated with them, for these, we used Workday employee data to identify their Business Unit
and/or division. Those that had emails or employee IDs that could not be found in the current Workday employee data were deemed to be inactive users and therefore not included in the allocations.

Outcome

After comparing the percentage change headcount allocation with the Micro Focus licence consumption that are based on the Flexnet consumption data, we can see in the table below that the majority of the Business Units/divisions are being charged even though they are not consuming the licences. With division 1, 2, 5 and 7 receiving an increase suggesting that their costs were previously being under-estimated.

For some divisions such as 2 and 4 they show little movement in allocations so will not be majorly affected by the change to start using consumption-based allocations.

Changes Made

The user licence consumption allocation has now been implemented by reallocating the contract costs to an individual ITRT then to the specific application and updating the allocation metric to reflect the licence user consumption.

As part of this project a methodology has been created for the standardisation of future datasets from Flexnet so will only need to be done once saving time in future as well as triggering a project to standardise the naming conventions across the various tools used.

An outcome of this project was that only 21% of the licences were actually being used so has enabled the SAM team with the data to decide on the appropriate number of licences needed and enabling cost savings.

Another powerful impact of Flexnet is that with data driven consumption we can provide more clearer and fairer allocations and has given us the springboard to look at larger contracts such as Microsoft and McAfee which will have more of an impact on cost allocations for software assets.