Webbo3 Data Analytics · Cohort 2 · Excel Track · Week 5, Virtual 1
Dashboard Review and Peer Feedback
You improve your own dashboard from your tutor's feedback, then learn to give feedback that actually helps someone else improve theirs.
By the end of this class you can
- Sort feedback by impact and fix the most important things first
- Review a dashboard against a checklist instead of by feel
- Give three structured observations: one that works, one that confuses, one suggestion
- Receive feedback without defending your work
There is no grade in this class. It is about a skill every analyst needs and few are taught: seeing your own work as a stranger sees it.
What you do in this class
- Read your tutor's feedback on your pod's Week 4 dashboard.
- Fix at least two things, starting with the highest-impact ones (Part 1).
- Swap dashboards with one peer from another pod and write three observations (Part 2).
- Read the observations you receive and decide what to change (Part 3).
Hand in: An improved dashboard, and three written observations for your peer.
Part 1: Improving Your Own Dashboard
Fix in Order of Impact
Not all feedback is equal. Fix in this order, because a beautiful chart of a wrong number is worse than an ugly chart of a right one:
| Priority | Kind of problem | Examples |
|---|---|---|
| 1 | Wrong: a number that is not true | Count instead of Sum; an average that includes the Missing rows as zeros; a range that stops at row 455 |
| 2 | Misleading: true, but read wrongly | A field-name title; months out of order; the leader not first; 3D |
| 3 | Hard to read | Unformatted numbers; too many colours; four slicers; text too small |
| 4 | Cosmetic | Alignment; gridlines; font choice |
The Self-Review Checklist
Go through your dashboard line by line. Every "no" is a fix.
| Check | Yes or no |
|---|---|
| Do the KPI numbers match the Analysis tab exactly? | |
| Does every PivotTable say Sum of, not Count of, where money is shown? | |
| Does every chart title state a finding with a number? | |
| Are bar charts sorted, biggest on top? | |
| Are months in calendar order? | |
| Is everything navy, with one accent used only for emphasis? | |
| Does the slicer change all three charts? | |
| Are there three insights, each comparing, explaining and suggesting? | |
| Is anything shown twice? | |
| Does it fit on one screen without scrolling? |
Show your dashboard to someone outside your pod for ten seconds, then hide it. Ask: "What is one thing you learnt about Verdant Foods?" If they cannot answer, your titles and insights need work.
Part 2: Giving Feedback
Definition
Structured feedback is feedback with a fixed shape, so it is always specific and always usable. In this course the shape is three observations: one thing that works, one thing that is confusing, and one suggestion.
Real-Life Scenario
A good football coach after a match: "Your passing in the first half was sharp. In the second half I could not tell who was marking their number 9. Next week, call the name of the player you are marking." One praise, one problem, one fix, each about something specific that happened.
What Makes an Observation Useful
| Weak | Strong |
|---|---|
| "Nice dashboard!" | "Works: the zone title, 'South West brings in 30% of revenue', told me the main point before I looked at the bars." |
| "The charts are confusing." | "Confusing: the category chart and the zone chart use different colours, so I wondered if the orange bar meant the same thing in both." |
| "Make it better." | "Suggestion: put the weekly line under the two bar charts, full width, so the trend is easier to read." |
Every strong observation points at one specific thing, says what it did to you as a reader, and (for the suggestion) proposes something they could do in five minutes.
Practise on Dashboard A
Write three observations for Dashboard A in the strong style: one that works (there is one!), one that confuses, one suggestion.
Check your answer
For example. Works: "The two KPI boxes are at the top, where my eye started." Confusing: "Revenue by zone appears twice, as columns and as bars, so I looked for a difference that is not there." Suggestion: "Delete the second zone chart and use the space for a weekly trend line."
Part 3: Receiving Feedback
- Read it twice before replying. The first read is defensive; the second is useful.
- Do not explain why you did it. The reader did not have your explanation when they read the dashboard, and neither will a manager.
- Thank the person, then decide: fix now, fix later, or disagree. Disagreeing is allowed, with a reason.
- Check every suggestion against the brief. If it contradicts the brief, the brief wins.
Picture handing your dashboard to a stranger on a bus. You cannot lean over and explain it. Whatever they understand on their own is what your dashboard actually says.
Checkpoint
Q1. Which do you fix first: a wrong average or a grey gridline?
The wrong average. Wrong numbers first, then misleading, then hard to read, then cosmetic.
Q2. What are the three structured observations?
One that works, one that confuses, one suggestion.
Q3. Why not explain your design choices when you receive feedback?
The real reader will not have your explanation. If it needs explaining, it needs fixing.
Tick Before You Move On
- ☐ I fixed at least two things, starting with the highest impact
- ☐ I went through the self-review checklist
- ☐ I tried the ten-second test on someone outside my pod
- ☐ I gave a peer three specific, structured observations
- ☐ I read my own feedback twice and decided what to change
Quick Recap
Fix in order: wrong, misleading, hard to read, cosmetic. Review against a checklist, not by feel. Feedback has a shape: one that works, one that confuses, one suggestion, each specific. Receive feedback by reading twice, not by explaining.
Next, Week 5, Live 2: all 22 topics in one review, the gaps fixed live, and a first look at SQL doing what SUMIFS does.