Optimizing Your Pricing Strategy with Markov's Data Analytics Tool

Pricing can make or break your product's success. Whether you are running a store or strategizing for your brand's next big sale, understanding how discounts, ratings, and customer behavior influence sales is crucial. But analyzing data often feels overwhelming—especially if you are not a data analyst. That’s where MarkovML's Data Analytics Tool steps in.

In this tutorial, you will learn how to make smarter pricing decisions using real Amazon product data. With its three modes—Text, Text-SQL, and SQL—Markov's Data Analytics empowers anyone, regardless of technical skills, to analyze data effectively and make faster, data-driven decisions.

📘

Note

This tutorial isn’t about building the best product strategies. Instead, it’s written to help you get comfortable with the product, so you can find your way around and use it with ease.

Dataset Overview

Here’s a quick look at the Amazon product dataset you will analyze. It includes columns like:

• product_id: Unique identifier for the product
• product_name: Name of the product
• category: Product category (e.g., electronics, books, clothing)
• discounted_price: The product's price after discount
• actual_price: The product's original price before any discount
• discount_percentage: Discount percentage applied
• rating: Average customer rating (typically out of 5)
• rating_count: Number of customer ratings or reviews
• about_product: Description or details about the product
• user_id: User who submitted the review
• user_name: Name of the customer who submitted the review
• review_id: Unique identifier for the review
• review_title: Title of the review
• review_content: Content of the review
• img_link: Link to the product image
• product_link: Link to the product page on Amazon

This dataset is rich with insights, ready to help you make informed pricing decisions.

Quick Overview of Markov's Data Analytics Tool

Markov’s Data Analytics Tool makes working with data effortless, no matter your skill level. Whether you are exploring trends, creating reports, or making informed decisions, it’s as easy as asking questions in plain English.

What sets MarkovML’s Data Analytics Tool apart? It offers three powerful query modes, so you can choose the level of control and flexibility that suits you best:

• Text Mode: Perfect If you want quick insights without writing a single line of code.
• Text-SQL Mode: This is for those who want more control over the query but might not want to write SQL from scratch.
• SQL Mode: For advanced users who are comfortable with SQL, you can write and execute your own SQL queries, giving you the power to drill into the data as deeply as needed.

No matter which mode you choose, Markov’s tool helps you quickly find valuable insights from your data in just a few minutes.

Let’s dive into the steps!

Start Simple with Text Mode

Text Mode: Ask Simple Questions for Quick Insights

The Text Mode is your go-to for quick insights. You can type simple questions in plain English, and the tool gives you instant answers—no technical skills required.

Say you are a marketer looking to adjust your product pricing to boost sales. You want to know how discounts affect sales and whether high-rated products can handle a price increase. You may start by asking clear, straightforward questions like:

• "What is the average discount percentage across all products?" - The answer can help you identify your baseline discount strategy.

  

• "Which categories have the highest average ratings?" - This will help you understand pricing spreads within a category.

  

• "What’s the price range for electronics?"- This could help you understand pricing spreads within a category.

  

• "How many products have ratings above 3.5 and discounts above 10%?" - This can find highly-rated, high-discount products that drive conversions.

  

Dive Deeper with Text-SQL Mode

Once you are comfortable, use Text-SQL Mode to refine your analysis. In this mode, you ask detailed questions, and the tool generates SQL queries for you to check and run.

• "Show me the total revenue for each category, considering only products with ratings above 4."- This could find which high-rated categories generate the most revenue.

  MarkovML provides "Explain with AI" to improve your query in case of error, as seen in this case. It looks like the 'discounted_price' column contained unexpected characters like '|,' which is not a number and is causing an error.

  

  After some attempts, it's known that the pricing data needs some tweaking. Here's the updated AI suggested query: "Can you show me the total revenue for each category, considering only products with ratings above 4, and ensure all non-numeric characters are removed from the price data? Also, remove commas from the price data before converting it to a number."

  

  

  📘

  Note

  1. If your query fails, you can use the Explain with AI to understand what went wrong and look at the suggestion to improve your query as seen in the above example.
  2. These errors also help you understand if your data is clean and ready to use. If not, then what is wrong with it.

• "Which products have the largest gap between actual price and discounted price?" - Highlight products with the most aggressive discounts for further marketing pushes.

  

• "How many products have more than 100 reviews, discounts above 20%, and ratings above 4." - Find popular, well-reviewed products with attractive discounts.

  For this data, it looks like there are zero product with above 100 reviews and discount above 20% with rating above 4.

  

• "What’s the revenue impact of increasing the price of all 3-star products by 10%?" - Predict the effect of a pricing adjustment on revenue.

  

Go Pro with SQL Mode

For those comfortable with SQL, SQL Mode gives you the power to write custom queries for in-depth analysis. This mode is perfect for uncovering complex trends and patterns.

You may try some sample queries like:

• SQL to find the top 5 categories with the highest average discounts and their total revenue. - Pinpoint high-discount, high-revenue categories for strategic focus.

  SELECT category, AVG(discount_percentage) AS avg_discount, SUM(discounted_price \* rating_count) AS total_revenue  
  FROM amazon (1)  
  GROUP BY category  
  ORDER BY total_revenue DESC  
  LIMIT 5;
  

  
  

  It looks like the data needs some fixing, and the AI suggested a better SQL:

  SELECT "category", 
  AVG(CAST(REPLACE(REPLACE("discount_percentage", '%', ''), '₹', '') AS FLOAT)) AS avg_discount, 
  SUM(CAST(REPLACE(REPLACE("discounted_price", '₹', ''), ',', '') AS FLOAT) * CAST(REPLACE("rating_count", ',', '') AS INT)) AS total_revenue
  FROM "amazon (1)"
  GROUP BY "category"
  ORDER BY total_revenue DESC
  LIMIT 5;
  

  

• SQL to list products where the discounted price is less than 70% of the actual price, sorted by rating. -Target highly discounted, top-rated products for promotional campaigns.

  SELECT * 
  FROM "amazon (1)"
  WHERE CAST(REPLACE(REPLACE("discounted_price", '₹', ''), ',', '') AS FLOAT) < 0.7 * CAST(REPLACE(REPLACE("actual_price", '₹', ''), ',', '') AS FLOAT)
  ORDER BY CAST("rating" AS FLOAT) DESC;
  

  
  

📘

Note

1. You may also download the table (results) by clicking the Downloadbutton on the top-right corner of the table.
2. You may switch between the table view and chart view by clicking on table viewand chart view toggle buttons.

• SQL to calculate the average revenue per product for each category. - Calculate the average revenue per product for each category.

  

Making Smarter Pricing Decisions

Now that you have explored the data, here’s how you can act on it:

• High Ratings, Low Discounts: Increase prices for products with consistent 4-star ratings.
• Low Ratings, High Discounts: Offer deeper discounts or improve product features.
• Top Categories: Focus marketing efforts on high-revenue categories.
• Price Adjustments: Use the "what-if" analysis (like increasing 4-star product prices) to predict and plan effectively.

That's it!

With Markov’s Data Analytics Tool, pricing optimization is no longer a guessing game. By analyzing key patterns like discounts, ratings, and revenue, you can make smarter, faster pricing decisions that drive growth.

Start exploring your data today—and watch your pricing strategy transform!

Need more help exploring? Talk to our expert today!