🚧 MVP (Under Development): This dashboard is a working first version and is still evolving.
Expect ongoing improvements to the data model, measures, drilldowns, and UI polish.
The Product Performance Analysis Report evaluates how the Campaigns product is performing across core business metrics, including revenue, entries, clients, and campaigns.
The report is designed for quick executive visibility (overview KPIs) and deeper operational insights (participation performance + demographics) for product, marketing, operations, and management.
- Google Sheets — lightweight staging, validation checks, and quick QA during development
- Power BI — data transformation, modeling, measures, and interactive reporting
- Figma — UI/UX wireframing and layout design (report look & feel)
- Problem Statement
- Data Source
- Workflow Overview
- Report Pages
- Live Power BI Report
- Analysis & Insights
- Recommendations
- Contact
This dashboard answers questions like:
- 💰 How much revenue is the Campaigns product generating overall and by campaign type?
- 🧾 How many entries are received, and what portion is valid?
- 🏆 What is the win rate, and how many unique winners are participating?
- 🎁 How much reward value is allocated, and what is disbursed vs pending?
- ⏰ What time-of-day shows peak participant activity?
- 🗺️ Where are participants coming from (regional distribution)?
- 👥 What do demographics (e.g., gender) look like, and what was/wasn’t captured by campaign?
- Source: Internal company database (web-based connection)
- Format: Web (confidential/internal)
- Privacy Note: Due to confidentiality, the dataset and direct source link cannot be shared publicly.
- Data extraction from internal company systems (web connection).
- Staging + QA in Google Sheets (spot checks, reconciliation, and early sanity tests).
- Power Query transformations (type enforcement, cleaning, and shaping for reporting).
- Measures/KPIs in Power BI for:
- Revenue, Entries, Valid Entries, Wins
- Rewards (total/disbursed/pending)
- Rates (Validity Rate, Win Rate)
- Participation breakdowns (time, campaign, demographics)
- UI/UX design in Figma, then implemented in Power BI for a clean, app-like layout.
Purpose: Executive snapshot of Campaigns product performance (high-level KPIs + trends).
- Total Revenue, Total Entries, Total Clients, Total Campaigns
- Revenue split by Campaign Type
- Monthly revenue trend (Month/Quarter toggle)
Purpose: Participation funnel + reward operations + activity patterns.
- Rewards overview (Total / Disbursed / Pending)
- Entries overview (Max Daily / Avg Daily)
- Validity + win performance (Validity Rate, Win Rate, Unique Winners)
- Reward amount distribution
- Valid entries & rewards by campaign
- Peak traffic analysis (entries by hour)
Purpose: Who/where participants are (and what fields were captured).
- Regional entries distribution (Ghana shapemap)
- Super participant leaderboard (high-activity participants)
- Entries per gender (including “No Gender Data” where gender wasn’t collected)
Open the interactive report here:
Demo Report
Note: GitHub README pages typically do not render iframes.
The metrics below reflect the snapshot shown in the attached report images and may vary depending on refresh timing.
- Total Revenue: GH¢470K
- Total Entries: 133K
- Total Clients: 2
- Total Campaigns: 2
| KPI | Value |
|---|---|
| Total Entries | 133K |
| Valid Entries | 105K (~79% validity) |
| Invalid Entries | 28K (~21%) |
| Wins | 93K |
| Win Rate (Wins / Valid Entries) | ~88.6% |
| Unique Winners | 38K |
| Avg Wins per Winner | ~2.45 |
Invalid entries are mainly driven by two major causes:
- Wrong code submission — customers enter an incorrect/invalid code.
- Agent discretion — a client’s agent can mark an entry as invalid based on internal rules or review outcomes.
Insight: Reducing wrong-code submissions is a product/UX opportunity, while agent-driven invalidation needs clear governance and auditability.
| Campaign Type | Revenue | Share |
|---|---|---|
| Sure Win (Single) | GH¢300K | ~63.8% |
| Basic Random Win | GH¢170K | ~36.2% |
Insight: Revenue is currently weighted toward Sure Win (Single), contributing nearly two-thirds of total revenue.
- Total Rewards: GH¢70.52K
- Disbursed: GH¢60.12K (~85%)
- Pending: GH¢10.40K (~15%)
What “Pending” means: Some rewards remain pending because clients did not load/fund their payout accounts, so payouts could not be completed to winners at the time.
Insight: Pending rewards are not just an ops backlog—this is also a dependency risk tied to client funding readiness.
From the “Valid Entries and Rewards per Campaign” visual:
- Vivo Always-On 2025: 74K valid entries (~70% of valid entries)
- Indomie Hero Pack Promo 2025: 31K valid entries (~30% of valid entries) and ~GH¢71K total rewards (rounded on the chart)
Insight: One campaign drives most engagement, while another carries most of the reward value—useful for comparing campaign mechanics and ROI.
Observed pattern (from the peak traffic chart):
- Very low activity overnight (roughly 00:00–04:00)
- Strong ramp-up from early morning
- Sustained activity through the day
- Highest activity appears in the late afternoon / early evening, then declines after ~20:00
Insight: Schedule push messages and operational monitoring around peak engagement hours.
Below is the regional distribution dataset used for the map visual (regional totals may not always match overall entries depending on coverage/time window and availability of region fields).
| Region | Total Entries | Share % | Popular Location | Number of Locations |
|---|---|---|---|---|
| Accra | 38840 | 45.63 | Accra | 119 |
| Ashanti | 21628 | 25.41 | Adum | 71 |
| Central | 8009 | 9.41 | Kasoa | 37 |
| Western | 7104 | 8.35 | Sekondi-Takoradi | 36 |
| Eastern | 3207 | 3.77 | Koforidua Town Center | 41 |
| Volta | 2085 | 2.45 | Ho | 19 |
| Bono | 1771 | 2.08 | Sunyani | 13 |
| Ahafo | 724 | 0.85 | Goaso | 6 |
| Northern | 704 | 0.83 | Tamale | 20 |
| Bono East | 429 | 0.50 | Kintampo | 7 |
| Savannah | 152 | 0.18 | Kpandu | 6 |
| Upper East | 150 | 0.18 | Bolgatanga | 10 |
| North East | 140 | 0.16 | Walewale | 5 |
| Oti | 99 | 0.12 | Nkwanta | 8 |
| Upper West | 69 | 0.08 | Wa | 4 |
Insights:
- Accra (45.63%) and Ashanti (25.41%) together account for ~71% of captured regional entries.
- The top 4 regions (Accra, Ashanti, Central, Western) contribute ~88.8% of captured regional entries.
- This indicates strong regional concentration—useful for targeted marketing and operational planning.
| Gender | Entries | Share |
|---|---|---|
| No Gender Data | 74K | 70.41% |
| Female | 20K | 18.82% |
| Male | 11K | 10.76% |
What “No Gender Data” means: Not all campaigns collect gender, so “No Gender Data” often reflects gender not captured by campaign design, not necessarily a data quality issue.
Insight: Where gender is collected, it can support segmentation; where it isn’t, reporting should treat it as “Not Collected” rather than “Missing”.
-
Reduce wrong-code invalid entries (Product + Marketing)
- Improve code entry UX (format hints, validation feedback, clearer error states)
- Where possible, implement code structure validation (length/pattern checks) before submission
-
Govern agent-driven invalidations (Ops + Management + CTO)
- Document the invalidation policy and standardize reasons
- Add audit-friendly logging (reason codes, agent ID, timestamp) to support transparency and dispute resolution
-
De-risk pending payouts with client funding readiness (Ops + Management)
- Set a client funding/prefund SLA before campaign launch
- Implement alerts when payout accounts are not loaded (to reduce pending accumulation)
- Track “Pending Rewards Aging” as an operational KPI
-
Improve demographic strategy without forcing it (Product + Marketing)
- Treat “No Gender Data” as “Not Collected” when campaigns don’t request it
- If demographic segmentation is a goal, align campaign setup to include gender capture consistently (where appropriate)
-
Use regional concentration for targeted execution (Marketing + Ops)
- Prioritize engagement initiatives in Accra and Ashanti where participation is most concentrated
- Use high-entry regions for A/B testing campaign mechanics and messaging
-
Monitor fairness/fraud signals (Ops + CTO)
- Review super participant patterns for abnormal behavior (high frequency, unusual win rates)
- Add rules/thresholds for automated flagging (MVP → v2 improvement)
- Phone: +233 54 583 2352
- LinkedIn: Click here
- Email: Click here
© 2025 Douglas Boateng | GitHub