Power BI dashboard examples

These Power BI dashboard examples are three report pages on example data. A regional manager sees their own sales and their real share of the company, a reader drills from a category's returns to the products behind them, and a KPI card judges orders against a target phased by last year's pattern.

Last reviewed · Virgo Analytics

Row-level security, example

One report page: £700,000 for the director, £184,200 for the North manager

Example data
Viewed as: North regional manager. Share of company sales
100.0%
The same cell, with the measure's company total read from a summary table the role doesn't filter:
26.3%

Viewed as: Sales director

Role: All regions (no filter)

Sales by region, October 2026

Sales

£700,000

Target

£700,000

Attainment

100.0%

RegionSalesTargetAttainmentShare of company sales
North£184,200£176,000104.7%
Midlands£151,300£160,00094.6%
South East£238,500£230,000103.7%
South West£126,000£134,00094.0%
Total£700,000£700,000100.0%100.0%

Viewed as: North regional manager

Role: Regional manager. Share measure written over the sales table.

Role filter on Region: [Manager email] = USERPRINCIPALNAME()

Sales by region, October 2026

Sales

£184,200

Target

£176,000

Attainment

104.7%

RegionSalesTargetAttainmentShare of company sales
North£184,200£176,000104.7%100.0%
Total£184,200£176,000104.7%100.0%
100.0%: the company total comes from the same filtered sales table, so the measure divides North by North. Read from a summary table, as Microsoft's guidance shows, it divides £184,200 by £700,000.

The same cell, with the measure's company total read from a summary table the role doesn't filter: 26.3%

Example data: an invented October for four sales regions. The director's view totals £700,000 against a £700,000 target. Under the Regional manager role the same page totals £184,200. North's share of company sales then reads 100.0% if the measure takes its company total from the filtered sales table, and 26.3%, the director's figure, if it reads the total from a summary table.
Show the numbers
Sales by region, October 2026, example data
RegionSalesTargetAttainmentShare of company sales
North£184,200£176,000104.7%26.3%
Midlands£151,300£160,00094.6%21.6%
South East£238,500£230,000103.7%34.1%
South West£126,000£134,00094.0%18.0%
Total£700,000£700,000100.0%100.0%

Two more Power BI dashboard examples

A drillthrough from category to product

Drillthrough, example

One duvet is 21.0% of bedding sold and 60.0% of bedding returns

Example data

Returns rate by category, October 2026

Units returned ÷ units sold

  1. Bedding7.0%Right-click › Drill through › Category detail
  2. Rugs4.1%
  3. Towels3.2%
  4. Curtains2.6%
BackCategory is Bedding
Drillthrough filter: Category = Bedding, taken from the bar right-clicked

Category detail

ProductUnits soldUnits returnedReturns rateShare of category returns
Duvet, 13.5 tog, double4208420.0%
Pillow pair610213.4%
Fitted sheet, king380143.7%
Mattress topper, double300144.7%
Duvet cover set, double29072.4%
Bedding total2,0001407.0%100.0%
Without the duvet: 56 returned ÷ 1,580 sold = 3.5%
Example data: an invented October of homeware sales, counted in units on orders placed that month. Bedding returned 140 of 2,000 units, 7.0%. The double duvet sold 420 and came back 84 times. Take it out and bedding's rate is 3.5%.
Show the numbers
Returns rate by category, October 2026, example data
CategoryUnits soldUnits returnedReturns rate
Bedding2,0001407.0%
Rugs800334.1%
Towels2,500803.2%
Curtains1,500392.6%

A reader who right-clicks Bedding lands on a page already filtered to bedding, and Power BI adds a back button. Both pages count returns against the month the order was placed, so the five products add up to the bar the reader clicked.

A Power BI KPI dashboard with a phased target

Power BI KPI dashboard, example

Orders are 2.5% behind an even target and 4.2% ahead of the seasonal one

Example data

Orders year to date, even target

97,500

Target
100,000
Distance to goal
−2.5%

Orders year to date, seasonal target

97,500

Target
93,600
Distance to goal
+4.2%

Annual target 120,000 orders. Even: 10 of 12 months = 100,000. Seasonal: last year's January to October share, 78.0% = 93,600.

Seasonal target: 120,000 × 78.0% = 93,600Last year, January to October: 89,700 of 115,000 orders
Example data: an invented online shop with an annual target of 120,000 orders and a December peak. By 31 October it had taken 97,500. Last year the same ten months held 78.0% of the year's orders, so the phased target to date is 93,600.

The KPIs behind these Power BI dashboard examples

These are the seven measures on the three pages, grouped by page, with each formula in words ready to turn into DAX.

KPIFormulaA good readingThe common mistake
Sales by region
Target attainment by regionRegion sales ÷ region target × 100, for the month The total row equals total sales ÷ total target, whoever is viewing Storing the target as a column on the sales table. It then adds up once per order line, so keep targets in their own table at region and month grain.
Share of company salesRegion sales ÷ sales across all regions × 100 The regions add up to 100% on the director's view, and a manager sees their region's real share Writing it over the region table and expecting it to hold for a manager. A DAX expression can't override row-level security, so the company total has to come from a table the role doesn't filter, such as a summary table.
Returns
Returns rate, unitsUnits returned ÷ units sold, for orders placed in the month Steady per category, with any category that jumps opened on the drillthrough page Building the drillthrough page on the return date while the summary page uses the order date. The category then shows one rate and its products add up to another.
Share of category returnsProduct units returned ÷ category units returned × 100 Close to the product's share of category units sold, or with a reason written beside it Reading it alone. A best seller returns the most units at an ordinary rate, so set it next to the product's share of units sold.
Orders against target
Orders year to dateOrders placed from the first day of the financial year to the last complete day Moving in step with the phased target Counting today's orders so far against a target that includes all of today. The card then reads behind every morning.
Phased target to dateAnnual target × (last year's orders in the same months ÷ last year's full-year orders) Agreed with whoever set the annual target, before the year starts Phasing evenly, as annual target ÷ 12 × months gone. A business that peaks in December then reads behind all autumn.
Distance to goal(Orders to date − phased target to date) ÷ phased target to date × 100 At or above zero, or with an agreed plan for the gap Handing the KPI visual the annual target. Distance to goal then shows a large shortfall until the last day of the year.

How to build these Power BI examples

  • Data sources

    Order lines from the order system, the targets file by region and month, a region table holding each manager's sign-in email, and a returns table where each returned line carries the order line it came from.

  • Grain

    Sales at one row per order line. Targets at one row per region per month. Returns at one row per returned line, dated by the order it came from, so the drillthrough page and the summary page count the same month.

  • Refresh cadence

    Sales and returns overnight. Targets only when finance issues a new version, with the version's name shown on the KPI page.

  • Who sees which rows

    Each role filters the Region table, and the filter reaches sales and targets through their relationships. Microsoft notes it is often more efficient to put these filters on dimension tables, and they travel only through active relationships. Roles apply to people with Viewer permission in the workspace. A regional manager added as a Member sees every region, so check workspace access before blaming the role, then try each role with Test as role.

  • The model's shape

    Sales, targets and returns are three fact tables. Region, Product and Date sit on the one side of one-to-many relationships that filter in one direction. Attainment works because Region and Date filter both sales and targets, and the two fact tables never join each other. Microsoft's guidance is to keep two-way relationships to a minimum, because they can slow queries and give report users confusing results.

Questions these dashboards should answer

“Am I ahead of target in my region?”Target attainment, under the manager's own role.
“How big is my region next to the rest?”Share of company sales, with the company total taken from a table the role doesn't filter.
“Which product is behind the bedding returns?”Share of category returns, on the drillthrough page, next to share of units sold.
“Are we on track for the year?”Distance to goal against the phased target, with the phasing shown under the card.

The pipeline KPIs behind a regional sales page are on the sales dashboard page, and the monthly board pack, on the same kind of model, is the finance dashboard example.

These examples in other tools

  • Power BI

    Roles are written in Power BI Desktop, and people are added to them in the Power BI service. Test as role then opens the report as that role sees it. If your team will keep the roles up after handover, our Power BI training on your own model covers writing one and testing it.

  • Data Studio (formerly Looker Studio)

    Before rebuilding the regional page in Data Studio, decide how each manager will be kept to their own region, and test it by opening the report as a manager. Share of company sales still needs its company total from a source the manager's restriction doesn't reach, as in Power BI.

  • Excel

    A workbook carries whatever rows were put in it. If the regional page goes out as Excel, build one file per region from the model, holding that region's rows and the company total as a single figure.

  • Google Sheets

    The phasing in the KPI example works well in a sheet, with one row per month and last year's orders beside the target, so whoever sets the target can check the sum. Power BI Desktop can connect to Google Sheets, so the model reads that sheet as its targets table.

North's share reads 26.3% or 100.0% depending on one measure, so Power BI consultant work on your own model, part of our dashboards and reporting service, starts with a review of its relationships and measures. We test every role before you sign the pages off.

Questions about row-level security and targets

Can you build a Power BI sales dashboard with row-level security?

Yes. Each manager's role filters the region table by their sign-in email, so one report serves every region. A DAX expression can't override row-level security, so a region-against-company measure takes its total from a table the role doesn't filter, such as a summary or aggregation table. Before sharing, Test as role opens the report as each role sees it.

What makes a good Power BI KPI dashboard?

Each KPI needs a value, a target and a written reason for the target's size. Power BI's KPI visual takes a target and shows the distance to it, so the target you hand it decides whether the card reads behind. Phase an annual target by last year's monthly pattern, and say on the page which phasing it uses.

When should a report use drillthrough?

Use it when a summary page raises a question one level down, such as which product drives a category's returns. A right-click on the bar opens a detail page filtered to that value, with a back button. Build the detail page on the same date field as the summary, or its products won't add up to the bar.

Why does a total change when someone else opens the report?

If the model has row-level security, each viewer's role removes rows before any measure runs, so a total covers only what that person may see. Check workspace roles too, because row-level security applies to Viewers, and anyone added as Admin, Member or Contributor sees every row.

Can these be added to our existing Power BI model?

Yes, if region, product and date are dimension tables with one-way relationships to the facts. The roles, the drillthrough page and the phased target then go on top of what you have. If targets are a column on the sales table, or relationships filter both ways, the model needs fixing first.

Different totals for different viewers?

On the call we open the report that gives two people two totals under each role and show you whether a role filter or a measure makes the gap.

Book a call

Free 30 min call