Let's Connect
aura menu

Role of Design-led AI in Powering Predictive Analytics

Naveen Puttagunta

TreeImage
TreeImage

For Founders, staying ahead of the curve demands foresight.

To achieve this foresight, it’s critical to understand ‘what’s next’ for the product, business and the goals to gain a competitive edge.

Predictive Analytics is one of the keys to unlocking this.

And AI has been a complete game-changer!

With ever-evolving AI / ML models that can include an exponential number of variables and process vast amounts of data in record time, these AI / ML models are becoming sharper and more scalable, enabling predictive analytics to become more accurate over time.

However, incorporating predictive analytics into the applications and making them more contextual and usable for end users is still complex.

The end users are yet to realize the complete value of predictive analytics in empowering them to analyze, comprehend and make informed decisions.

And here’s where Design comes in.

Design Strategy and UX UI designs function as catalysts and ensure that the platforms with predictive analytics, with all their complexity are still easy to use.

Not only that, design enables the users to derive the true value of the predictive analytics and help them achieve their end goals.

Design also enhances user experience by providing interactions and visualizations to provide real-time feedback to the user and reduce the cognitive load.

Being Design-led Makes Predictive Analytics More Consumable

Let’s take a look at how we leveraged design to minimize the complexity of a pricing intelligence platform enabling stadiums to optimize one or more of these below in different combinations and achieve their goals through predictive analytics:

  • Profitability

  • Revenue

  • Occupancy

In this pricing intelligence platform, we primarily looked at two aspects of predictive analytics:

  1. Real-time dynamic ticket pricing

  2. Marketing budget optimization

As with any such predictive model, there were many influencing variables such as stadium occupancy, other demographics, and geographics, game season, game jeopardy, game popularity, timings, pricing, on-sale window timing, teams, location, game popularity, marketing platforms that would impact their business goals and business growth.

The pricing analysts have to consider these influencing factors in many combinations to identify the most desirable impact on their end goals – profitability, occupancy and revenue.

The key challenges that we had to address with design were:

  • Help the pricing analysts identify the key influencing factors in the given context of the game

  • Enable them to dynamically change the values of these influencing factors in an easy and simple manner

  • Evaluate the impact of these changes through dynamic computing on the business goals

Along the journey of this workflow, we focused on reducing cognitive load for the user and enabling it through design patterns, streamlining the flow of selection, and output.

To achieve this, we focused on 2 things:

1. Understanding the personas of pricing analysts, along with their business objectives and product roadmap.

2. Understanding of their domain expertise and business strategy, ensuring our design strategy aligned seamlessly.

In addition to this, we needed to understand the key influencing factors for the pricing intelligence platform. This included understanding their significance, why they mattered, and how they would affect the pricing analysts’ goals.

We also identified where these factors were applicable and most influential, as well as areas where they might not be effective.

This comprehensive understanding of the user persona and the variables affecting the accuracy of the predictive analytics platform enabled us to build a robust design & functionalities that mapped with their business goals.

We layered this understanding with our expertise in creating what-if scenarios in analyses to design the platform in a way that users would easily pick essential variables, lock the non-critical variables, adjust the key variables and dynamically change them to predict the outcomes and their impact on the end goals.

The designed application enabled the analysts to:

  • To experiment with different pricing strategies in a simulated environment before applying them in real-world situations.

  • Make informed decisions based on the outcomes of these simulations for maximizing profitability, occupancy and revenue.

As a result, Design increased adoption.

butterfly
Let'sTalk
butterfly
Thanks for the submission.