IT systems generate a plethora of information; much of which remains untapped from the power of deep insight analysis. I’m a huge proponent of measuring what you manage. As such I’ve compiled a list of five common IT data sources that can be analyzed to save money, unlock headcount, drive sales, and improve operations.
1. SaaS install base savings
Install base measures how much of a particular product you have in use. For commercial software as a service (SaaS), evaluating your SaaS install base yields an understanding of which applications you have installed, how many licenses are being used, and detailed application usage patterns.
The average enterprise today hosts hundreds of SaaS applications; many of which are far from optimized. Examples include overpaying for features not used, artificially high license seat costs due to inactive users, and in many cases failing to consolidate multiple purchases of the same product under a single contract. Understanding these insights can save tens or even hundreds of thousands of dollars per year!
There are many ways to analyze SaaS install base, including do-it-yourself finance spend reports to procurement of a more robust software asset management tool. The common denominator, however, is getting access to your finance system(s). With access to purchase orders and expense items, one can immediately get eyes on the SaaS spend footprint.
Failing to examine your SaaS install base is akin to forgetting about a massive investment portfolio. If not managed properly, enterprise SaaS usage will inevitably become a bloated rats’ nest of redundancy; slowly and painfully draining the corporate coffers.
2. Infrastructure as a service usage
Similar to the way software as a service has spiraled into an all-you-can-eat buffet of budgetary binging, cloud infrastructure spend has also spun out of control. Exorbitant AWS or GCP bills, poor bill-back visibility, and lacking demand forecasts make IaaS spend an expensive enigma. This is especially true for Internet companies whereby IaaS is part of product delivery; in other words, a cost of revenue or cost-of-goods-sold direct cost.
There is a silver lining with IaaS, however. Unlike SaaS solutions whose API maturity levels vary widely, all of the major IaaS players (Amazon, Microsoft, and Google) have impeccable APIs. This means IaaS components are instrumented from day-one with robust cost accounting and usage telemetry. All you have to do is consume the data; ideally from an easy to manage, off-the-shelf commercial costing tool.
As with SaaS install base, understanding your IaaS usage insights can lead to huge cost savings as well as reduction of information security risks.
3. Program management office summaries
Most IT program management offices crank out an unholy number of projects and programs each year. Yet at the end of the fiscal year, there are no celebrations. No parades. And in some cases, not even a “thanks” for all the hard work. IT is only rewarded with more work.
As we all know – IT supply is tightly coupled with business demand. As business demand on IT increases, the deeper the bench of IT resources required. Yet simply asking for additional IT headcount without some numbers to back up the need falls flat almost every time.
Get your PMO team into Jira, Smartsheet, or whatever PPM tool du jour and slice and dice the raw data. Create reports to convey the work IT has done as well as the (typically insane) demand on the horizon. Socialize these metrics relentlessly so that business stakeholders understand the criticality of the IT supply chain. Cold hard facts get head count. Crying about workload gets you Kleenex.
4. Support and community forums
Product managers need to understand how customers perceive the functionality of their offerings. This is true for external products, but also for IT product owners who manage internal platforms such as Salesforce, SAP, or Workday.
Ticketing tools such as Jira Service Desk or ServiceNow use input forms which allow for structured input (think drop-down menus) and free-form text fields. Drop-down menus have a finite set of text to analyze, whereas free-form text fields allow the user to convey their experience using a more natural style of human dialog.
Imagine a simple help desk input form with three fields: a pre-populated requestor name, the type of request in a drop-down (software request, outage, etc) and a free-form text field to describe the request. Any reporting tool can easily count the number of request types, but deep analysis of problematic issues requires natural language processing (NLP). Tools like AWS Comprehend or Python NLTK (for the roll-your-own data science types) can quickly tease out topics and custom entities like “Workday” or “Cisco VPN.”
These tools are essentially like hiring an army of analysts to pour through massive corpora (bodies of text) to examine key insights that would otherwise remain buried, and hence unactionable, within your operational systems. Getting ahead of issues or capitalizing on opportunities through insights is a much more proactive approach to enterprise service management.
5. Salesforce opportunity summaries
Continuing with the natural language processing paradigm – Salesforce is another prime candidate for text mining. Many sales opportunities are closed as either “won” or “lost” with a simple drop-down consisting of a finite set of reasons. While it’s interesting to see a report of canned reasons such as “security compliance” or “price point,” sales leaders must continuously analyze free-form text fields to discover new reasons for deals winning or losing.
Deals may be lost because the product’s user interface is now perceived as outdated. There could also be a new competitor siphoning deals out from under you. Likewise, you may be winning deals for reasons which aren’t so obvious; such as achieving FedRAMP compliance or providing a freemium offering for light-touch users. Either way, getting into the weeds of free-form text is the only way to tease out these insights. Understanding these insights may ultimately help you sell more products and lose fewer deals.
