Commercialized products and platforms often over-simplify user expectations in order to deliver a product that is good fit for the majority target user group solving for a generic set of use cases. Consequently, the marketplace is flooded with off-the-shelf digital products that enterprises and users adopt and use almost immediately as compared to custom made software solutions. This works well until its effectiveness is tested in an environment where user needs are fundamentally diverse. 

India is a sharp illustration of this challenge because it operates almost as two parallel realities — India and Bharat. The former refers to the urban, global-facing, English-speaking population, while the latter is more local, traditional and non-English-speaking, typically somewhat disconnected from the global economy. Each has its own behaviours, aspirations and level of digital maturity which only amplifies variability and diversity.​

We faced this challenge in a project context, through a concrete problem statement: “How might we design a solution that is sensitive to existing diversity while still achieving a modern, unified enterprise platform for a leading Indian recruitment company? (For instance, the recruitment process for white collar worker like a Java architect is fundamentally different than a gig worker for food delivery platform like Uber Eats) ” 

We further attempted to generalize our learnings to allow for repeatability, in terms of a framework that outlines a deterministic process to design for a fundamentally diverse environment, abstracted beyond a specific industry or geography. 

Resolving the tension between standardization and customization

The chart below illustrates the challenge of India’s diversity through a set of parameters relevant to the hiring industry context that we worked on.

Whether you’re building a healthcare app, a retail store or a government program, the challenge is always the same: efficiency vs. empathy. On one side, you need standardized rules so the system can grow and stay organized. On the other side, you need personalized features so the system actually works for real people in their local context. This is a perpetual tension, on one hand building a "skeleton" that is strong enough to scale, but "muscles" that are flexible enough to adapt to the unique needs of the user.

What are the challenges to be addressed to resolve this tension?

Ethnographic diversity (e.g., India/Bharat):

• How might we design one platform that successfully caters to distinct and diverse user sensibilities and behaviors without fragmenting the experience? 

The product scale paradox:

• How might we balance between the need for standardization for scalability with the need for localization for segmental requirements?

Transformation planning:

• How do we thin-slice this transformation to ensure that value is delivered and received incrementally, without the risk of a "big-bang" failure?

While there is a plethora of AI tools readily available in the market, the answer isn’t a monolithic system or a complex mosaic of disparate and asynchronous point solutions. 

From first principles thinking, we framed the idea of an optimally unified platform (OUP): a stable common set of capabilities across the core value stream, integrated with additional capability modules that are pulled into play depending on the specific needs presented by a scenario. Such an approach is meant to deliver value, upgrade independently and perform measurably.

The solution framework: Optimally unified platform (OUP) 

Let us assume we want to build a platform for the recruitment industry for 3 different business verticals catering to 3 different employment segments with very different hiring processes. How do we define the common solution that works for all in such an environment? Do we try to solve every segment’s problem at one go or is there another way?

The goal of this OUP framework is to start simple, from a priority business with a set of needs that apply across verticals (despite their differences) and add specific features that are unique to this primary vertical. Then the platform goes through a series of versions where the common features are incrementally customised as more businesses or user types are onboarded and their specific needs are uncovered through continuous user research and discovery. This requires planning at the outset to embrace the complexity and scale of the future, by provisioning it in the architecture and functional structure at the outset.

This allows for a seamless way to manage the chaos coming from conflicting requirements from different business verticals or user groups, in a systematic and harmonious process. 

We’re moving beyond the traditional MVP approach by weaving in more empathy than just adding features and also being more systematic about it. Instead of guessing which requirements matter most, we use continuous research to resolve conflicting needs with real-world data. This allows us to build a platform incrementally, that stays flexible for unique users while remaining stable and scalable at its core.

Progressive development phases of the OUP

The progressive shaping process to build a MUP with ‘Continuous Discovery’ informing the level of Customisation on the platform evolution journey

Development methodology for OUP

Unity without uniformity

The outcomes achieved through this approach are:

• For the enterprise: Faster response to market shifts, lower cost of change, shared reusable services and consistent compliance and insight across regions or markets.

• For the user: Experiences that feel personal—fostering language compatibility, supporting channel choice and guided journeys—thereby increasing trust, engagement and conversions.

Quantifiable results from the pilot project demonstrated a significant lift in match quality and achieved 100% internal user adoption during the organization's transformative shift to its new AI-enabled recruitment platform. Expectation is a 20–40% improvement in shortlist relevance and a 15–30% reduction in time-to-hire within the first year of implementation.

In retrospect, designing for diversity requires embracing its inherent differences instead of retro-fitting ready solutions. The optimally unified platform (OUP) approach helped us do exactly that in a predictable process.