Unlocking a $200m market potential - Revolutionizing financial reconciliation
Nanonets is an AI-powered document processing platform that automates data extraction from various documents to reduce manual labor. Recognizing the potential for streamlining financial reconciliation, we created a solution that combined data extraction from invoices, bank statements, and other financial documents with AI that automates tedious reconciliation processes.
Responsibilities
User interviews, usability testing, user flow optimization, wireframing, mockups, prototyping, impact tracking
Duration
12 weeks
Goal
Our goal was to design an intuitive user experience that not only simplified the financial reconciliation process but also broke into the market by leveraging AI and LLMs to revolutionize the process. This innovative approach aimed to drive significant customer acquisition and increased Annual Recurring Revenue (ARR), gaining a market share and challenging the dominance of established competitors like QuickBooks, SolveXia, Xero, Oracle and Blackline.
MY PROCESS
Kick starting with research
As reconciliation was a new domain for the team, we lacked a deep understanding of the specific requirements and challenges associated with the reconciliation process. To bridge this knowledge gap and gain valuable insights into the process, its pain points, and market gaps, we initiated a series of expert interviews. These interviews helped us understand the complexities of existing solutions, and identify areas where there was a clear opportunity to improve the user experience and streamline the process.
OPPORTUNITIES IDENTIFIED
Streamlining Manual Comparisons
Automate the comparison of bank transactions on CSV files with ERP systems, eliminating the need for manual matching.
Automating Physical Document Processing
Automatically extract data from physical documents, allowing for seamless comparison with Excel sheets and eliminating the need for manual entry.
Handling Complex Transactions
Develop software capable of reconciling many-to-one transactions, addressing the limitations of existing ERP solutions.
Simplifying Credit Card Reconciliation
Create a solution that can effectively handle the varying charges and complexities associated with credit card reconciliation.
Centralized Reconciliation Platform
Offer a centralized platform that can handle complex rule-based reconciliation processes as well as seamlessly integrating with other ERP systems.
Challenges
The reconciliation-through-AI market was relatively new, with few established competitors. This limited our ability to gather detailed information about existing solutions. Additionally, most traditional reconciliation software was only accessible through sales demos, making it difficult to directly compare features and functionality. We had to rely on the insights shared by potential customers regarding the pain points they encountered with their current reconciliation software.
Overcoming Constraints: An Iterative Approach to Feature Development
To overcome the limited competitor information, we prioritized an iterative development approach. By rapidly prototyping and seeking feedback from early users, we were able to gather valuable insights and continuously improve our feature. This iterative process allowed us to identify and address specific user needs, even in the absence of extensive market data.
Who are we designing for?
We used information we got from our interviews with experts to profile an ideal persona that would use such reconciliation software on a daily basis. Having this persona in mind while designing helped us empathize with the end user and refer to any time we had blockers in decision making.
DESIGN: PHASE ONE
To validate our insights from the interviews, we initiated the first phase of feature development. This MVP (Minimum Viable Product) was designed and tested with a focus on speed to quickly gather user feedback. We integrated this version with our existing workflows applicatoin for a specific AI model to gather practical insights.
By mapping out the user flow, we established a solid foundation for the reconciliation feature. This process allowed us to visualize the user journey, organize the information architecture, anticipate potential edge cases, and ensure alignment with stakeholders from the get-go.
Designs
Since we were integrating the reconciliation feature into our existing workflow, we were able to leverage our existing design system and rapidly iterate on high-fidelity wireframes.
Testing and Impact Tracking: Post-Launch Insights
After the successful launch of the reconciliation feature, we closely monitored user interactions to identify areas for improvement. The feature automated reconciliation by 85%, significantly reducing manual effort. This efficiency gain, coupled with the enhanced user experience, attracted three new customers who are generating a modest increase in ARR.
While the feature was well-received, our analysis revealed the following areas for enhancement:
Lack of Visualization
Users expressed a desire for a more visual representation of the mapping between invoices and bank statements.
Manual Verification
While users appreciated the automatic reconciliation, they still wanted the option to manually review and correct any mismatches.
Transaction Marking
Users requested the ability to mark specific transactions as "reconciliation not required."
DESIGN: PHASE TWO
Based on the insights gained from testing, we enhanced the reconciliation feature to include a matching view, allowing users to manually approve, correct, or nullify reconciliations performed by Nanonets. To improve visibility and accessibility, we also added a dedicated reconciliation section on the sidebar, providing users with quick and easy access to this feature.
Since we were designing a feature that diverged from our existing product layouts, we created wireframes to test out different structures and information architecture (IA) options. This iterative approach allowed us to refine the design based on feedback without taking much time.
Impact
By incorporating the insights from user feedback and enhancing the feature to improve visibility, visualization, and user control, we witnessed a substantial increase in the adoption of our reconciliation solution. Over 14 customers opted for Nanonets reconciliation, contributing to a 12% increase in ARR. Moreover, we have automated over 95% of the reconciliation process and reduced financial close times by 70%.
Reflection
Empathy and understanding user needs were of high priority in this project as we didn't have the resources for a competitive audit. Therefore, I learnt to take an iterative approach to design and improve based on insights from session recording and usability tests.
I found that setting metrics and tracking performance was crucial for measuring my impact and identifying areas for improvement. Aligning product goals with those of the business ensured that I was working towards a shared objective.
Finally, I understood the importance of having a complete and deep understanding of the subject matter and industry in order to design a tailored and detailed solution.