Combining RPA and Analytics to Amplify Value from Digital Transformation Programs

The demarcations between automation and analytics are rapidly fading away

Two very important pillars of digital transformation, RPA (Robotic Process Automation) and Analytics, are drawing increasingly closer and creating complementary opportunities that have the potential to deliver powerful results.

Analytics has been around for much longer than RPA/IPA (Intelligent Process Automation). Smart enterprises, already familiar with analytics, can now leverage that maturity to run RPA programs better.

RPA programs, on the other hand, can help enterprises make analytic insights much more useful by weaving them right into the operating fabric of a company.

Using the below framework in their digital transformation programs, enterprises can leverage both RPA and Analytics seamlessly. Analytics and RPA, used in unison, can approximately double the value realized from digital transformation programs.

Investment in analytics pre-date those made in RPA by almost a decade. Analytically mature organizations have their data issues sorted and at a minimum have deployed Business Intelligence (BI) in their organization. Some have progressed to using data science tools and algorithms to solve specific business use cases (there are a rapidly growing number of organizations that have made investments in advanced analytics like NLP and AI).

Armed with that advantage, the RPA or Robotics COE (Centre of Excellence) should ideally co-opt analytics talent to function most effectively.

But, where to begin? A hierarchical representation, akin to Maslow’s pyramid – is a handy guide to the analytics opportunity presented by a future of work transformation program.

The journey of RPA COE looking to exploit analytics doesn’t have to start at the base layer, though often it is a good and solid place to start. Your enterprise may very well have the tools, talent and culture, to start at a higher level, and move quickly upwards and the choice can be based on the transformation plan, the COE, its design and relationship with business functions. The framework presented is agnostic enough to yield value whatever the model. Every layer in the hierarchy merits a deep dive of its own but a quick primer on each is a great starting point.

Level 1: Transformation intelligence
This is where organizations analyse current and past datasets and zero in on the right functions/processes for transformation. Once identified and steps to automate initiated, the same analytic techniques can closely monitor the pattern and trends of value to be unlocked. It can proactively remove blocks and track how the transformation is impacting strategic performance metrics of the firm.

Transformation intelligence has a generous BI component. Think of it as getting access to, munging and analysing all data that is impacted by or impacts a “future-of-work” transformation. Organizations that do not get this right complain of not getting “value” from their transformation program. Identification, measurement, monitoring and continuous calibration of “value” – at both operational and strategic performance level – that is what this level is all about.

Level 2: Embedded intelligence
Your enterprise now is set on the correct path to value realization. The time now is ripe for optimizing outcomes. Enterprises often have analytic-silos – a situation where departments run analytics locally to solve a specific use case. In this level of maturity, enterprises begin to collate these analytic assets and embed them into RPA’ed processes.

The cost of embedding analytical intelligence into human work flows is costly. In sharp contrast, embedding an analytic procedure into a robotic process incurs very little cost. Smart HR organizations identify flight-risk employees often as early as during onboarding. The Employee Onboarding Process, a favorite automation candidate in many organizations, can be immediately made intelligent with a robot running a segmentation algorithm that clusters employees based on their possible attrition risk, and report that back to the hiring managers.

A centralized digital COE (or an RPA COE) should make an inventory of such analytic assets and proceed to a cogent determination how these can be embedded into RPA processes (this segmentation algorithm can be reused from digital marketing, for example, or reused into an automated insurance claims settlement processes).

Level 3: Orchestrated intelligence
The true differentiator between Digital Adopters and Digital Leaders is the latter’s ability to bring together all data, analytic, RPA technology assets and human intelligence in a manner where information and decisions move seamlessly between these components. This is orchestrated intelligence – digital transformation’s equivalent of fifty musical instruments coming together to produce a melodious symphony.

The key operating word perhaps is “seamless”. By seamless, I mean a meticulously scripted and optimized state where information, data and activity-outcomes flow from one functional element to another (and often back) to deliver strategic outcomes that are of an order several multiples more than what these components would achieve by themselves (Example, conversations on a chat-bot, or from multiple service channels of which a chat-bot is one, get passed to an NLP engine to analyse conversation-sentiment while a robot checks the LTV (lifetime value) of the customer.

The results are fed into a decision matrix from where another robot generates customized offers that are sent through on the customer’s most preferred contact channel. All within less than an hour of a service episode).

Conclusion
There are a few operative principles that this overall framework must keep in mind:

* Each level has its own value potential (in other words, enterprises need not wait to get to Level 3 to unlock value). The value realization quantum drastically increases as one moves levels;

* The tools, talent and investments needed at each level can potentially be different (and they scale nicely as one shifts levels);

* This framework assumes the simultaneous presence of analytics and RPA in a digital transformation plan. It also assumes that at least one of the two – if not both – have a COE (of course, for optimal results organizations should gravitate towards a common Digital COE).

The demarcations between automation and analytics are rapidly fading away. The current trend of scalable cloud computing assets with very capable analytic assets, combined with rapidly maturing RPA tools, present an unprecedented opportunity for firms to seriously ‘up’ their digital game. While the prize is clear, crafting a roadmap and sharp execution is the key.

-- Subrata Majumdar, Director of Consulting, Symphony Ventures (India).



Tags assigned to this article:
RPA analytics Digital Transformation Programs

Advertisement

Around The World