Assembling, Documenting, and Presenting Your Final Deliverable

Structuring Your Final Deliverable

Your capstone should read as one connected piece of work, not four disconnected artifacts loosely related by dataset. Organize it in the same order someone would need to follow it: the question you set out to answer, how you cleaned and prepared the data, what the exploratory analysis revealed, the dashboard built around your core question, and the narrative explaining what it means and what should happen next. Each section should reference the ones before it — the dashboard should visibly answer the question stated at the start, and the narrative should point back to specific findings from the exploration, not introduce a conclusion that seems to come from nowhere.

Consistency across sections matters more than most people expect. If your exploration flags three customer segments and your final dashboard only shows two, or your narrative's headline number doesn't match what's on the dashboard, a reviewer will notice the gap before they notice the quality of your analysis. Do a final pass specifically checking that numbers, segment definitions, and terminology are identical everywhere they appear.

Documenting Your Cleaning and Analysis Decisions

Every nontrivial decision from earlier in the project — how you handled a specific set of missing values, why you excluded a segment, how you defined an outlier — should be documented somewhere a reviewer can find it, even if it's a short appendix rather than part of the main narrative. This isn't about proving you followed a checklist; it's about making your conclusion reproducible and defensible if someone questions a specific number.

A simple, effective pattern is a running decision log kept alongside the analysis itself: one line per decision, what you chose, and why. It takes minutes to maintain as you go and can take hours to reconstruct from memory afterward, so keep it updated throughout the project rather than trying to recall your reasoning at the end.

Building the Dashboard and the Narrative Together

The dashboard and the narrative should be built as companions, not in sequence with no connection between them. As you draft the narrative, check that every claim it makes is something a viewer could actually verify by looking at the dashboard — a narrative claiming a trend that isn't visible on any dashboard view is a sign one of the two needs revising. As you build the dashboard, keep asking whether each view earns its place by supporting a specific point in the narrative, rather than existing just because the underlying data was available.

This back-and-forth, rather than a strict "analyze, then dashboard, then narrative" sequence, is what past cohorts have found produces the most coherent final result — treat the two as iterating on each other, not as separate stages.

Presenting Your Capstone for Review

When you present, apply everything from the storytelling module directly: open with your so-what, walk the reviewer through the dashboard as evidence rather than as a standalone tour, and close with a specific recommendation. Anticipate the two or three questions a skeptical reviewer would ask about your data or method, and address them before they're raised rather than hoping they don't come up.

Treat the review conversation as useful signal, not just an evaluation to get through. Questions you couldn't fully answer point to exactly the follow-up analysis that would make the next version of this project stronger — that feedback loop is the same one you'll use on real analytics work after this course.

Practical Review Checklist

Before moving on, confirm that you can:

• Structure your deliverable so each section references the ones before it
• Check for consistency in numbers and terminology across all sections
• Maintain a running decision log for cleaning and analysis choices
• Verify every narrative claim is supported by something visible on the dashboard
• Anticipate and prepare for likely reviewer questions before presenting

Conclusion

A strong capstone isn't four separate assignments submitted together — it's one coherent piece of work where the cleaning, analysis, dashboard, and narrative all visibly support each other. That coherence is the real skill this course has been building toward.