By Madhavan Krishnan, Senior Vice President – SaaS Operations and Engineering, Epsilon
In the world of Software as a Service (SaaS), it can get tempting to focus solely on the features, functionality, and usability of software applications. These are arguably some of the most important pieces of the value proposition for a company, but these are table stakes and an entry point for companies to compete and gain market share. Yet the real barometer for business success (sustained profitability and growth) in SaaS is often eluding many companies since they do not necessarily follow through on a competitive and compelling product features strategy with an equally strong focus on operational excellence.
From experience, it is clear there is a need for SaaS businesses to execute around a few fundamental KPIs that balance sales focus with customer retention, engagement, and long-term customer lifetime value metric governance. KPIs such as (but not limited to) Customer Acquisition Costs (CAC), customer engagement with active user management, Customer Retention/Churn Ratio (CRR), Annual Recurring Revenue (ARR), Net Promoter Score (NPS), and Lifetime Customer Value help keep the balance between customer satisfaction and operational excellence, essential to sustainably grow the business.
As we can see, most of the value generation in SaaS happens after the first sale is made. Sustained success comes down to value generation from sustained client engagement and operations downstream. This is where driving continuous improvement in the ever-shrinking time-to-value cycle of software engineering and operations becomes the core of value generation and competitive differentiation.
Evolving from DevOps to a Site Reliability Engineering (SRE) mindset
DevOps and SRE are best visualised as an integrated set of practices in the continuous value chain of reliably launching and running software product features that meet an ever-evolving client need. Simply put, DevOps practices focus on enabling fast, efficient, and continuous development and deployment of software product features. SRE brings in the user-centric view of how these product features are experienced, with a singular focus on the system’s reliability impact on that experience.
This requires a cultural shift to view system reliability not just as an operations problem but as a key responsibility where all teams, including engineering, products, and operations, have a stake.
Scaling efficiency and impact with hyper-automation pipelines
Hyper-automation is a discipline of systematically infusing automation into every part of IT and business operations across all activities and processes of software products. The outcome of such automation initiatives often leads to the imagination of operational processes, resulting in new ways of working.
Hyper-automation initiatives that heavily leverage operational data to derive operational insights for preventive and predictive maintenance lead to a discipline of machine learning operations (MLOps), a process of building, testing, and maintaining ML models as an operations discipline that can lead to impactful outcomes for both the business and clients.
Deploying automation at scale requires organisations to reorient their approach to operations from a cost function to a value-generating function. Operations automation is a key driver for expanding people’s capabilities and overcoming operational technical debt. The discipline of building, growing and executing automation pipelines with the same rigor as the product feature backlog will deliver operational savings on an ongoing basis.
Scaling without growing with AI
Most companies underestimate the impact of AI on operations due to past experiences of incremental improvements achieved through earlier generations of automation technologies like Business Process Automation (BPA) and Robotic Process Automation (RPA).
With a foundation of software, SRE and AI build-to-deploy-manage pipeline foundation in place, the stage is set for a multi-fold impact on productivity AI can have on software operations. The value lock from AI in operations (AIOps) can happen at a speed and scale that most businesses are not used to.
It is estimated the productivity impact of AI-assisted operations can amplify human potential by three to five times. This is a scale that is far greater than earlier efforts of automation through technologies like RPA which tried to replicate human effort through machines. When companies get it right, human-assisted AI in operations can unlock productivity gains that can far exceed the best business case that companies have experienced so far. This would require companies to invest in technology, get the foundation of operational building blocks right, reskill their teams, and reimagine every aspect of their operational business processes. This is not an easy task to pull off, but a long-term commitment to stay invested in this roadmap appears to be the path to the promised land.