Frameworks
Metrics questions
Product rational (<5 min) Measuring Impact (15 min) 5 metrics in a dashboard - a basic pulse check of the Specific + time frame e.g. orders per day Counter metrics / guardrail Evaluating tradeoffs (10 min) Set Goal (company / user goal) User actions (user journey) Acquisition & Activation: prioritise actions based on goals turn each action to quant metric Individual or Aggregate - Can you group many actions for an overview? Ratio (percentages, rates, or per-unit measures) Quantity/Absolute Numbers (raw counts or amounts) Does it matter more for you to measure the overall magnitudes?
Intrinsic vs. Heuristic (direct vs. proxy) Can you derive more knowledge from the measure? Or do you have to rely on heuristics? Leading indicators help predict future performance Session duration, Sessions/user, Activation Facebook - seven friends in the first ten days Lagging indicators give insight into past outcomes Trade-offs and limitations of these metrics In this case of measuring success of {company_X}, I believe there are many parts of the product we can focus on, for example we can talk about measure the success of {product_Y}. Are you ok with that assumption?
2 out of 3: mission, product summary, user journey North star (tracking the goal) evidence that we are moving the north star Make sure we are not hurting the biz Look back and check if we’re tracking the right things Describe the product (e.g. marketplace) Mission / differentiation Users, when are they ‘successful’? The intersection of value between users and company Levers to impact the metric e.g. views, orders, cancellations, Business (how they make money) Audience (light segmentation) Adoption, Activation, Engagement, Monetisation Thinking through extremes Closest proxy to the goal Try not to use ratios or avg. Numerator / Denominator - can be gamed What is the north star missing? Will you change the north star? What levers effect the north star? RCA
Dianna Yau 1 (tradeoff + debug) Describe the product (e.g. marketplace) Mission / differentiation Users, when are they ‘successful’? Channels (e.g. facebook is down) new product, cannibalisation 3x Solutions, prioritise impact/effort/confidence acquisition, activation, engagement, retention Evaluate success (of the solution) Business (how they make money) Audience (light segmentation) Is it a critical issue? (unlikely that it isn’t) debug the metric - ask questions Internal (self inflicted) Messy bit that needs cleanup
Context
Understand the product, business model, use cases - ask or state your assumptions.
Understand the issue - clarify the metric.
Timeline - when, slope, trend
Extent of the problem - geo, platform, demographic, are other metrics dropping?
Where in the user journey (e.g. users starting session from notification, new, power, resurrected)
- Technical
- Value
- Experience
- Awareness
- Attention
- acquisition
- onboarding
- active use
- resurrection
RCA Factors
Internal
Technical - Engineering - deployment, infra, bug reports, crashlytics Technical - Instrumentation Value & Experience - Product - release, UX change, test (start or end), other teams or products in ecosystem Value - Pricing and incentives (start or end) Awareness - Marketing, announcements External
Provider (e.g. AWS, library, platform) Attention - Seasonality (e.g. Christmas) Event and news (e.g. power outage, weather, big game) Solutions