Summary of "Clase 1 - Tu primer Dashboard profesional en solo 1 clase - Acelerador de carrera con Power BI"

Main ideas & lessons (Class 1: First “professional” Power BI dashboard)

1) Purpose of the immersion + who/why it’s for

The session is the start of a “career accelerator” immersion focused on building a professional Power BI dashboard in one class.

It targets multiple motivations:

The course is presented as free, with a step-by-step structure and ongoing practice.

2) Daily schedule preview (what comes after Class 1)

Methodology / instructions taught (detailed checklist)

A) Accountability + certification rules (to receive the free certificate)

To receive the certificate:

B) How to build the first Power BI dashboard (end-to-end workflow)

  1. Install Power BI Desktop

    • Use the official Microsoft download link.
    • Avoid downloading immediately during the demo if it may take too long (they want students to follow along live).
    • For Mac users, they mention a separate YouTube installation video.

  2. Understand the Power BI Desktop interface

    • Three key sections:
      • Report view
      • Table view
      • Model view
    • Add/remove pages (e.g., delete Page 1, create/use Page 2).

  3. Adopt the correct mindset

    • You are not “just a student”; you are a data analyst hired to create a dashboard (a “control panel”).
    • Start with the most important first step:
      • Exploratory analysis (write this down; always do it first)

  4. Do exploratory analysis in Excel

    • Open the provided Excel file.
    • Inspect:
      • Data size/structure (exclude the header row from “count of points”)
      • Time coverage (it’s only one month: March)
      • Meaning of fields (e.g., answered vs solved/not solved are different)
      • Key metrics:
        • Call outcomes (e.g., cancellations, not answered, etc.)
        • Duration
        • Satisfaction Index (NPS)

  5. Import data into Power BI

    • In Power BI: Add data to report → Import from Excel
    • Choose/load the dataset using the preview screen.

  6. Transform data with Power Query (critical step)

    • Do not “Load” immediately if the dataset has issues.
    • Emphasis: Transform first so the dataset is cleaned/standardized before modeling.
    • Concept taught:
      • Power Query edits the data pipeline without permanently damaging the original source (compared to a cloud/process layer).
      • Changes occur in Power Query, not by overwriting the original file.

  7. Apply concrete transformations (example used in the class)

    • Fix misspellings in category values:
      • Example: replace “cancellations” with the wrong letter variant (Z vs C) using:
        • Right-click → Replace Values
    • Standardize text formatting:
      • Use Power Query transform options like:
        • lowercase/uppercase controls
        • Capitalize Each Word (to make visuals consistent)

  8. Split columns to extract meaningful parts

    • Use:
      • Split Column
      • Choose delimiter-based splitting:
        • By delimiter (example delimiter: hyphen “-”)
        • If spaces exist around delimiters, use an approach that supports that (a “custom” space delimiter approach).

  9. Remove unnecessary columns

    • After splitting, if a column/field doesn’t add value to the report, remove it:
      • Remove Columns
    • This enforces critical thinking: decide what’s important vs not important.

  10. Finalize transformations

    • Close & Apply to load the cleaned dataset into the model.

  11. Create the first visuals (starting simple)

    • Make a Card visual for total number of calls.
    • Common issue addressed:
      • If the card shows only an ID, switch to a proper aggregation like:
        • Count (instead of showing a raw ID)

  12. Format and improve clarity (data storytelling)

    • Use formatting carefully:
      • Avoid memorizing every slider; proceed step-by-step.
    • Apply:
      • consistent colors
      • clear titles
      • readable contrast
    • Add a meaningful title (example: “Total Calls” instead of “Call ID count”).

  13. Fix wrong aggregations for NPS/satisfaction

    • Demonstrated error:
      • NPS/satisfaction should not be shown as a raw sum (which produced impossible values like 343).
    • Correct aggregation:
      • Average satisfaction index (Power BI provides suggestions like Avg).

  14. Add more KPIs using appropriate aggregation

    • Convert “satisfaction index” to Average.
    • Add Waiting time / response speed (again fix aggregation to average to avoid huge sums).

  15. Use the donut/pie chart carefully

    • Rule:
      • Donut/pie works only if categories are limited to ~4 items.
      • If there are too many categories, use a different chart type.
    • Use fields like:
      • Legend (e.g., resolved vs not resolved)
      • Values (e.g., call counts)
    • Demonstrate interactivity:
      • Clicking segments filters other visuals.

  16. Use interactivity + edit interactions to automate the “story”

    • Building charts isn’t enough; you must guide the viewer.
    • Use Edit Interactions (e.g., Do nothing, Highlight, Filter, etc.).
    • Example goal:
      • When a person (assistant/agent) is selected in one visual, other visuals filter to show that person’s performance.

  17. Build a stacked bar chart for “by assistant” comparisons

    • Conceptually:
      • Y-axis = vertical
      • X-axis = horizontal
    • Then:
      • Put assistants on the proper axis and call counts/metrics on the other.

  18. Add a treemap for category composition

    • Use when you want to show “slices” of categories.
    • Emphasis on clarity:
      • they demonstrate how small differences can become unclear without removing axes (they switch back/forth based on readability).

  19. Handle time visuals correctly

    • Strong rule:
      • Never graph time vertically (don’t put time on the Y-axis).
    • Time should read horizontally.
    • They add a grouped column chart:
      • Date on the horizontal axis
      • Call count as the values
    • Drill down based on selected hierarchy level (day/quarter).

  20. Add slicers for segmentation

    • Add a date slicer to filter the dashboard for:
      • a range of days
      • specific windows (e.g., day 8 to day 13, later end-of-month days)
    • Use results to identify patterns like cancellation spikes.

  21. Add a text box title

    • Name sections clearly (example: “Service Tracking”).

  22. Improve the dashboard canvas/theme (visual finishing)

    • Use a template library (background/design assets prepared by designers).
    • Custom background workflow:
      • Export background design from tools like PowerPoint/Canva/Figma.
      • Save as SVG if possible; otherwise JPG or PNG.
        • Recommendation order:
          • SVG first
          • JPG second
          • PNG third
      • Upload as the canvas background in Power BI.
    • Transparency check:
      • If the background doesn’t appear, it may be at 100% transparency—remove transparency.

Three “takeaways” they insist you cannot leave without

  1. Make your report unique (don’t make it identical to everyone else; visuals + structure matter).
  2. Generate hypotheses (don’t stop at one explanation; verify patterns and consider context—e.g., changes over time, new teams, probation periods, etc.).
  3. Power BI is “literal” (don’t fear it; learn it—it unlocks potential).

Additional market/job emphasis + “surprise” material

Final engagement challenge and how participation is verified

Certification reminder

They reiterate certification requirements and the phrase of the day.

Extra activity

Speakers / sources featured

Speakers

Sources / references mentioned

Category ?

Educational

Share this summary

Is the summary off?

If you think the summary is inaccurate, you can reprocess it with the latest model.

Summarize another video

Video