How I Use Math to Solve Business Problems

I will be talking about a method that I actively use and that I think is very useful as a result of the guidance I have given to my own team. You can apply this method to your own business somehow. ☺️

When tackling complex business questions, a structured and hypothesis-driven approach to data analysis is crucial. By breaking down the problem into smaller, manageable parts, we can identify the root cause of an issue and develop actionable insights. 🚀

In this article, we will explore how to systematically address a business problem (which has happened before 👀) using data, focusing on a step by step approach that starts with the big picture and gradually narrows down to specific details.

Example Business Problem:

The problem is brought to the data team as follows:

• Observation: "Fashion category sales are down 15% compared to the previous year."
• Business Question: "Why have sales declined, and what actions can we take to reverse this trend?"

Understanding the User Funnel

Before diving into the analysis, it is essential to visualize the user's journey through the funnel. By understanding the key stages of the funnel, we can systematically ask the right questions to identify where drop-offs or issues occur.

The funnel typically includes:

1. Awareness: How are users discovering our platform/ marketng channel?
2. Consideration: How are users engaging with our products or services?
3. Conversion: How are users completing purchases or desired actions?
4. Retention: How are we retaining users and encouraging repeat actions?

By breaking down the funnel into these stages, we can focus our analysis on specific areas and uncover potential issues impacting performance.

Decision Framework: External vs. Internal Factors

When a business problem arises, such as a decline in sales, it is essential to determine whether the root cause is external or internal. This decision framework helps guide the analysis:

1. External Factors:

   1.1 → Supply Issues: Has there been a change in the availability of products?

   1.2 → Demand Changes: Has customer demand for the products decreased due to market trends, economic conditions, or seasonality?

   
   

2. Internal Factors:

   2.1 → System Changes: Have there been any changes to the platform, pricing, promotions, or algorithms?

   2.2 → User Behavior Changes: Has user behavior changed despite no internal changes? For example, users may have shifted their preferences or purchasing patterns.

If no external factors are identified, we then proceed with a deeper analysis of internal factors. This involves examining the user experience, platform performance, and the impact of any recent changes to identify potential internal issues that might be causing the observed problem.

We can follow the steps below to address whether the internal flows are system related or the user experience has changed. 🚀

Step 1: Analyze Channel-Based Traffic

Questions Answered:

• Which marketing channels are driving the most traffic?
• Has there been a decline in traffic from specific marketing channels?
• Are there any SEO or marketing issues impacting traffic?
• We break down traffic by channels:
  • Organic Search: Down 20% YoY
  • Paid Search: Flat YoY
  • Social Media: Down 10% YoY
  • Email Marketing: Up 5% YoY

Step 2: Platform-Based Breakdown

Questions Answered:

• Are there differences in performance between platforms?
• Which platform is experiencing the largest decline in traffic or conversions?
• Are there any technical or user experience issues affecting specific platforms?
• Platform breakdown:
  • Mobile App: -5% YoY
  • Desktop: -25% YoY
  • Mobile Web: Flat YoY

Step 3: Page-Level Performance

Questions Answered:

• Which key pages are experiencing a decline in engagement or conversions?
• Are specific pages causing drop-offs in the user journey?
• Are there technical issues affecting page performance?
• Page performance:
  • Homepage: Flat YoY
  • Product Listing Page: -15% YoY
  • Product Detail Page: -20% YoY
  • Cart Page: -5% YoY

Step 4: KPI Metrics for Visitors

Questions Answered:

• Are users spending less time on key pages?
• Is there an increase in bounce rates or drop-offs?
• Are there specific user behaviors indicating a lack of engagement?
• Visitor metrics:
  • Bounce Rate (Product Detail Page): Up from 40% to 55%
  • Average Time on Page (Product Detail Page): Down from 3 minutes to 2 minutes
  • Conversion Rate: Down from 3% to 2%

Step 5: Component-Level Performance

Questions Answered:

• Are specific page components causing a decline in engagement or conversions?
• Are there technical issues with components like recommendations, ads, or banners?
• Is the recommendation engine displaying relevant products?
• Component analysis:
  • Recommendations Click-Through Rate (CTR): Down from 8% to 4%
  • Banner Engagement: Flat YoY
  • Ads CTR: Down from 5% to 3%

Step 6: Payment Method and Checkout Flows

Questions Answered:

• Are there specific issues with payment methods affecting conversions?
• Is there an increase in drop-offs during the checkout process?
• Are there user complaints or feedback about payment-related issues?
• Payment performance:
  • Credit Card Payments: Flat YoY
  • Digital Wallet Payments: Down from 15% to 10%
  • Conversion Rate at Checkout: Flat YoY

Step 6: Monthly Analysis and Change Impact

Questions Answered:

• When did the decline in sales or performance metrics start?
• Were there any specific changes made during that period?
• Are there seasonal trends or external factors influencing the decline?
• Monthly breakdown of key metrics:
  • March: Flat
  • April: -5% decline
  • May: -10% decline
  • June: -15% decline

Action: Review changes made during the critical months (April to June), such as:

• Changes in pricing or promotions.
• Modifications to page design or layout.
• Updates to recommendations or ads algorithms on recommendation components.

By addressing these specific changes, we can determine which factors contributed to the decline and identify corrective actions.

Conclusion: A Systematic Approach to Addressing Business Problems

By following a structured, hypothesis-driven approach that starts with the big picture and gradually narrows down to more granular details, data analysts can effectively identify the root cause of business problems.

This method not only helps in uncovering insights but also provides a roadmap for developing actionable solutions.

Ultimately, the choice of specific tools and methods should align with the unique needs and complexity of each project, ensuring the most effective outcomes.

Thank you for your time; sharing is caring! 🌍