Power BI for Data Analysts — Complete Guide

Power BI Workflow — How It All Connects

Power Query — Loading & Transforming Data

// Step 1: Remove top rows (Excel exports often have title rows at top)
Table.Skip(Source, 2)

// Step 2: Promote first row to headers
Table.PromoteHeaders(PreviousStep)

// Step 3: Rename columns
Table.RenameColumns(Source, {{"Amt", "amount_inr"}, {"Dt", "order_date"}})

// Step 4: Change data types
Table.TransformColumnTypes(Source, {
    {"order_date",  type date},
    {"amount_inr",  type number},
    {"customer_id", type text}
})

// Step 5: Filter out test orders
Table.SelectRows(Source, each [order_id] <> "TEST")

// Step 6: Add a conditional column (like CASE WHEN in SQL)
Table.AddColumn(Source, "value_tier", each
    if [amount_inr] >= 2000 then "High Value"
    else if [amount_inr] >= 500 then "Mid Value"
    else "Low Value")

// Step 7: Replace city name inconsistencies
Table.ReplaceValue(Source, "New Delhi", "Delhi", Replacer.ReplaceText, {"city"})

Data Model & Relationships — The Star Schema

DAX — Writing Measures for Business Metrics

// Total Revenue
Total Revenue =
SUM(fact_orders[amount_inr])

// Total Orders
Total Orders =
COUNTROWS(fact_orders)

// Average Order Value
AOV =
DIVIDE([Total Revenue], [Total Orders], 0)

// Delivered Revenue only
Delivered Revenue =
CALCULATE(
    SUM(fact_orders[amount_inr]),
    fact_orders[status] = "delivered"
)

// Return Rate %
Return Rate % =
DIVIDE(
    COUNTROWS(FILTER(fact_orders, fact_orders[status] = "returned")),
    [Total Orders],
    0
) * 100

// Previous Month Revenue
PM Revenue =
CALCULATE(
    [Total Revenue],
    DATEADD(dim_date[date], -1, MONTH)
)

// Month-over-Month growth %
MoM Growth % =
DIVIDE(
    [Total Revenue] - [PM Revenue],
    [PM Revenue],
    0
) * 100

// Year-to-Date Revenue (resets each financial year)
YTD Revenue =
TOTALYTD(
    [Total Revenue],
    dim_date[date],
    "31-03"           // Indian financial year end
)

// Same period last year
SPLY Revenue =
CALCULATE(
    [Total Revenue],
    SAMEPERIODLASTYEAR(dim_date[date])
)

// YoY Growth %
YoY Growth % =
DIVIDE([Total Revenue] - [SPLY Revenue], [SPLY Revenue], 0) * 100

// CALCULATE overrides the current filter context
// Syntax: CALCULATE(expression, filter1, filter2, ...)

// Electronics revenue — regardless of what category slicer is set to
Electronics Revenue =
CALCULATE(
    [Total Revenue],
    dim_products[category] = "Electronics"
)

// % of Total — Electronics as % of company total
Electronics % of Total =
DIVIDE(
    CALCULATE([Total Revenue], dim_products[category] = "Electronics"),
    CALCULATE([Total Revenue], ALL(dim_products)),  // ALL removes category filter
    0
) * 100

// Revenue excluding returns
Net Revenue =
CALCULATE(
    [Total Revenue],
    fact_orders[status] <> "returned",
    fact_orders[status] <> "cancelled"
)

// Noida customer count — ignoring city filter on a visual
All Cities Customer Count =
CALCULATE(COUNTROWS(dim_customers), ALL(dim_customers[city]))

// Rank cities by revenue (1 = highest)
City Revenue Rank =
RANKX(
    ALL(dim_customers[city]),
    [Total Revenue],
    ,
    DESC
)

// Flag top 5 cities
Is Top 5 City =
IF([City Revenue Rank] <= 5, "Top 5", "Others")

// Running total (cumulative revenue over time)
Cumulative Revenue =
CALCULATE(
    [Total Revenue],
    FILTER(
        ALL(dim_date[date]),
        dim_date[date] <= MAX(dim_date[date])
    )
)

Visuals — Choosing & Configuring Charts

Building a Complete Power BI Dashboard

Is Power BI important for data analyst jobs in India in 2026?

Power BI is one of the most in-demand tools for data analyst roles in India in 2026. It is required or preferred in the majority of data analyst job postings across industries — FMCG, banking, e-commerce, logistics, healthcare, and IT. The reason is Microsoft's dominance in Indian corporate environments: most companies already have Microsoft 365 subscriptions, which include Power BI Pro. Hiring managers who cannot do complex analysis themselves want analysts who can build dashboards that update automatically and are shareable within the organisation. The ability to build a clean, interactive Power BI dashboard from raw data — including correct data modelling and meaningful DAX measures — is the skill that separates strong candidates in interviews. At EVIKA ACADEMY, Power BI is part of the core curriculum for this reason.

What is the difference between a DAX measure and a calculated column in Power BI?

A calculated column is computed row by row when the data is loaded and is stored in the data model. It adds a new column to your table with a fixed value for each row — like adding a "Profit" column that is Revenue minus Cost for every row. A DAX measure is computed dynamically at query time, based on the current filter context (what slicers, filters, and row/column headers are active in the visual). Measures do not add columns to the table — they produce a single aggregated value for whatever context the visual applies. As a rule of thumb: use a calculated column when you need a value that belongs to a specific row (category, classification, a fixed calculation per transaction). Use a measure when you need an aggregation that should change based on filters — total revenue, average order value, percentage of total, YTD comparisons. Measures are always preferred for analytical aggregations because they are filter-aware and more efficient for large datasets.

What is a star schema in Power BI and why does it matter?

A star schema is the recommended way to organise tables in a Power BI data model. It has one central fact table (containing the transactions — orders, sales, events) surrounded by dimension tables (customers, products, dates, locations) connected by relationships. The fact table has many rows and contains numeric measures (amount, quantity) and foreign keys. Dimension tables have fewer rows and contain descriptive attributes (customer name, city, product category, date information). A star schema matters because: (1) it makes DAX measures simpler to write — most measures just SUM or AVERAGE a column in the fact table; (2) Power BI's query engine (VertiPaq) is optimised for this structure and performs significantly better than a flat denormalised table; (3) relationships between tables propagate filters automatically — selecting "Electronics" in a slicer connected to the Products dimension automatically filters the fact table. Building a correct star schema before writing DAX is the most important habit to develop in Power BI.

What Power BI skills are tested in data analyst interviews in India?

Common Power BI assessment tasks in Indian data analyst interviews: (1) Build a report from a provided CSV or Excel file — demonstrating data import, cleaning in Power Query, and a basic dashboard with 3–4 visuals; (2) Create a measure for Total Revenue, Previous Month Revenue, and MoM% change using DAX; (3) Explain the difference between a measure and a calculated column; (4) Set up a relationship between two tables and explain the filter direction; (5) Create a date table and explain why it is needed for time intelligence functions; (6) Use CALCULATE with a filter to compute a conditional total (e.g. revenue for Electronics only); (7) Add Row Level Security (RLS) so different regions only see their own data. Companies also ask conceptual questions: what is a star schema, when to use Power BI vs Excel, how does CALCULATE work. Practical build tasks are more common at analytics firms; conceptual questions are more common at IT companies and MNCs.