Business Intelligence – Intelligent Terms to Never Forget!!

blogging
jupyter
Deep Learning
Author

Kashish Mukheja

Published

Friday, 22 September 2023

KPI.jpg

Gentle Introduction

Business Intelligence (BI) is a field that plays a pivotal role in helping organizations make data-driven decisions. Business Intelligence involves the process of collecting, analyzing, and transforming raw data into meaningful insights that can be used by businesses to make informed decisions. Now, for someone who is reading this blog, I am pretty confident that you arrived here in order to understand the intelligent terms I’m referring. Well, for any Business Analyst or Business Intelligence Engineer, here I’ll be describing about various jargons used in the industry and what each of these mean in detail. This will be done on a few blog series. Under each blog, I will focus on 1 business and 1 technical jargon.

Also, comment down below if you know why I chose to write 1 business and 1 technical terms as part of the blog.

KPI

I wouldn’t be doing justice to the article and the Analysts here, if I don’t commence with possibly the most fundamental terminology in the BI community, viz., KPI.

KPI stands for Key Performance Indicators. Well, as the phrase suggests, these are the indicators which directly assess the business objective. KPI help businesses measure their success of failure in achieving their objectives. Some key characteristics for KPIs include:

  1. Quantifiable: KPIs are quantifiable measures that reflect the performance of an organization, department, team, or individual in relation to their goals and objectives.
  2. Timeliness: KPIs should provide information in a timely manner, allowing for timely interventions or adjustments.
  3. Consistency: KPIs should be consistent over time and comparable across different periods. Why not, If you can’t compare over time, how would they lead to actions!
  4. Actionability: KPIs should provide insights that can lead to actionable decisions and improvements.

These can be high-level at a company’s stand point of view, or lower level at various departments level. Let’s consider a scenario where you’re running an online clothing brand company with your in-house application. You’re the Executive of the company and you have Directors reporting to you for various departments such as Sales, Logistics and Software. At the C-Suite level, we can define the following KPIs:

[Universe]

WAIT BOOMER!!! How can you define KPI just like that ?? Don’t you wanna define your business objective first ?

[Author]

Ohhh I forgot about that part! Forgive me Universe!

[Universe]

Forgiven! But don’t you dare forget such primer details in your blog!

[The Author Continues]…

So let’s brainstorm on what the Executive would care about. You can stop for a bit, think and maybe then come again to check how many of the goals intersected. For me, If I were the C-suite leve, I would want to know the following information:

  1. How much revenue has my company generated this month Vs the other months?
  2. How has the new feature launch impacted our sales?
  3. How has our application performed? (There can be various aspects to consider such as latency, click through rate, recommendation engine’s performance, etc.)
  4. What has been the in-stock and out-stock trend for the past year?
  5. Which marketing strategy impacted our sales by the highest percentage?

While the above are not the only goals, but these are definitely the ones I would be interested. Once, you have got your goals established, let’s proceed towards creating KPIs and metrics for each of the goals:

How much revenue has my company generated ?

KPIs: - Monthly Gross Revenue trend (growth or fall)

Metrics: 1. Monthly Sales Revenue 2. Average Order Value 3. Customer Lifetime Value (CLV) - Calculated using RMF score - Recency, Frequency and Monetary. 4. Conversion rate (percentage of visitors who make a purchase)

How has our application performed?

KPIs: - Application conversion Rate

Metrics: - Application Load time or latency spikes - Application Vs Website Traffic share - Mobile Cart Abandonment Rate

How has the new feature launch impacted our sales?

KPIs: - Feature-Driven Sales Growth Rate : Percentage Increase in Sales Attributable to the New Feature

Metrics: - Sales conversion rate (Conversion rate of users who engaged with the new feature compared to those who did not) - New Feature Adoption Rate (Percentage of customers who have used the new feature) - Churn Rate (Reduction in customer churn rate after the feature launch)

You can perform your own practise now to define various Goals, KPI and metrics for your business use-case!!

Data Silo

Imagine you’re a big company consisting of different departments namely Sales, Finance and Software. Each of your department contains plethora of data concerning their respective departments. For E.g., Sales department might have number of purchases made per week, amount of units sold, profits or losses incurred, etc. This data kind is very different when we compare data from Software department which can be: latency observed per week, percentage increase in API hits, etc. Hence, it makes logical sense that the data for each of these departments is stored in a isolated systems, albiet it also would lead to data redundancy as multiple departments (Such as Sales and Marketing) may store similar or identical data independently. The Data Silo scenario happens all the time in the industry. Here are a few characteristics of Data Silo scenario:

  1. Data Redundancy: Data duplication can occur as a result of data silos. Multiple departments may store similar or identical data independently, leading to inefficiencies and potential data inconsistencies.

  2. Limited Data Sharing: Data sets are isolated and not easily accessible or visible to other parts of the organization. This lack of data sharing can hinder collaboration and decision-making.

  3. Inefficient Processes: Data silos can lead to redundant processes. For instance, if different departments maintain their customer data separately, it may require duplication of efforts when it comes to customer interactions and analysis.

  4. Barriers to Insights: Business Intelligence and Data Science rely on the availability of comprehensive and integrated data. Data silos can be a significant barrier to gaining holistic insights from an organization’s data.

  5. Compliance and Security Risks: Data silos can also raise security and compliance concerns, as different data repositories may have varying levels of access controls and security measures.

There are various methodologies to tackle the Data Silo effects such as 1. Data Integration (through ETL - Extract Transform and Load processes); 2. Data warehousing; 3. Building API (Application Program Interfaces) and data connectors to create connections between disparate systems and enable real or near real time data sharing … and so on.

The details of these are out of scope of this current article.

Next Up

I would end part 1 of the series here! Coming up next are 2 more important concepts viz., Data Dashboard and Data Warehouse.

Stay Tuned.. Until then!

Back to top