Jonathan Filbert - OKR Presentation for PMB Fakultas Ilmu Komputer, Universitas Indonesia

Deck I used as an OKR and KPI presentation for the super-app and machine learning app we build for the orientation programme.

Transcript

1. RISTEK in PMB-Land: Through the Matching Glass RISTEK 2021
2. OKRs (Objective & Key Results) • Improve the enthusiasm of “Kenalan” from Maba ◦ User Count ◦ Page Views -> 1k/day ◦ Interaction Count ◦ Match / User • Help Maba in maximizing their time in the website during PMB ◦ Match / users ◦ Kenalan maba ◦ AI Prediction
3. • Key Results (30 days): Assume that: 1. There are (400 students x 4 + 400x1/2) = 1800 students KPI = 1800 users Actual = 1297 users KPI = 72% succeeded Actionable Item =>PMB Website announcement / Guerilla Marketing User Count
4. • Key Results (30 days): Assume that: 1. Every users view 1 page per day 2. There are (400 students x 4 + 400x1/2) = 1800 students 3. For 90 days KPI = 30 x 1800 = 54k page views Actual = 494,708 page views KPI = 914% exceeded Page Views
5. Interaction Count • Key Results (30 days): Assume that: 1. 1800 students 2. For 30 days 3. 1 day = 4 View + 1 scrolls + 5 click KPI = 540k Interaction Actual = 800k Interaction KPI = 148% Exceeded
6. Engagement Time • Key Results (30 days): Assume that: 1. 1800 students 2. For 30 days 3. 1 day = 10-30 mins of PANSOS KPI =20 minutes avg Actual = 49 mins avg. KPI = 245% Exceeded
7. Matches Made • Key Results (30 days): Assume that: 1. 1800 students 2. For 30 days 3. 1 maba = 10 kating (takes 3days for 1 kenalan - write notes, revision, etc) KPI = 18k matches Actual = 38741 matches KPI = 200% Exceeded
8. Kenalan Maba • Key Results (30 days): Assume that: 1. 500 Maba 2. For 30 days 3. 1 Maba = 10-100 kenalan (50 avg.) KPI = 25k kenalan between maba Actual = 25100 matches KPI = 101% Exceeded
9. Prediction • Key Results (30 days): Assume that: 1. 1800 students (500 Maba + 1300 Non) 2. 38741 matches (prev slide) 3. 1 student = 30 prediction 4. Optimum prediction = 54,000 matches KPI = (38741/54000) = 72% accuracy Actual = 87% accuracy (less quantity but more quality prediction) KPI = 120% exceeded
10. Cost •
11. Cost •
12. Cost Total = $282 ($282 AWS + $20 Netlify) = Rp 4.300.944,09 Web (Frontend) = $20 (Netlify) + $207 = $227 = Rp 3.236.622,75 Data (RecSys Engine) = $76.15 = Rp 1.085.765,74
13. Tengs! Email@ristek.cs.ui.ac.id Email2@ristek.cs.ui.ac.id