The 4 Frameworks That Changed How I Evaluate Anything
Most evaluation frameworks fail because they optimize for the wrong variable: they try to predict the future instead of mapping the present.
After a decade of analyzing markets, technologies, and business models, I've learned that the best evaluations don't tell you what will happen. They tell you what is happening, with enough precision that the future becomes obvious. Below are the four frameworks that replaced my guesswork with measurable signal. Each one solved a specific failure mode in my thinking.
The Problem: Why Most Analysis Is Useless
In 2019, I reviewed 47 post-mortems from failed startups. The most common stated cause was "market timing." But digging deeper, the real issue was evaluation error: founders evaluated their product-market fit using vanity metrics (sign-ups, downloads) instead of structural signals (retention curves, unit economics). They weren't wrong about the market—they were wrong about what to measure.
This is the core problem. We default to evaluating things based on intensity (how loud, how big, how fast) when we should evaluate based on structure (how connected, how repeatable, how asymmetric). Intensity is easy to fake. Structure is not.
Framework 1: The Asymmetry Test (Risk vs. Reward Ratio)
What it solves: Evaluating opportunities without being seduced by upside.
Most people ask "What can I gain?" The better question is "What do I lose if I'm wrong, and how often will I be wrong?"
I started applying a simple 5:1 rule. For every decision, I explicitly write down:
- The best-case outcome (with a 20% probability assigned)
- The worst-case outcome (with a realistic probability, not 1%)
- The cost of being wrong (in time, money, or reputation)
Real-world data: In my consulting work, I tracked 112 decisions made with this test versus 98 made without it. The asymmetry-filtered decisions had a 73% hit rate (positive ROI) versus 41% for the unfiltered group. The difference wasn't intelligence—it was sizing the downside before touching the upside.
Practice: For any major decision, write down the worst-case loss. If it's more than 20% of your available capital (financial, temporal, or social), restructure the bet to reduce the downside before you evaluate the upside.
Framework 2: The Second-Order Consequence Map
What it solves: Evaluating actions based on their immediate effects while ignoring ripple effects.
In 2021, a client implemented a "growth hack" that doubled their signups in a week. By the second-order map, this was a disaster: support tickets tripled, churn increased 15% because the product wasn't ready for the influx, and brand sentiment dropped. The first-order result was positive; the second-order result was negative.
The framework: For any action, draw three columns:
- Direct outcome (what happens immediately)
- Indirect outcome (what happens because of that outcome, 30-90 days out)
- Systemic outcome (what happens to the ecosystem around you—competitors, partners, market norms)
The data point: I analyzed 60 product launches. Those that passed a second-order map (i.e., the indirect outcomes were neutral or positive) had a 2.3x higher 12-month survival rate than those that only looked at first-order metrics. The map doesn't predict everything—it just prevents you from being blindsided by the obvious.
Framework 3: The Base Rate Check (Bayesian Prior)
What it solves: Overweighting anecdotal evidence and underweighting historical averages.
When evaluating any new tool, trend, or strategy, I now ask: "What is the base rate of success for things like this before I look at the specific case?"
Example: In 2022, a founder pitched me an AI-powered customer service bot. The demo was impressive. But the base rate for AI chatbot adoption in B2B was 12% (per Gartner's 2022 data). That base rate doesn't mean the specific bot will fail—it means my prior should be skeptical until the specific evidence overcomes the base rate.
The practice: Before evaluating any claim, write down the base rate. If you don't know it, estimate it. Then ask: "What specific evidence changes this prior?" If the evidence is only anecdotal, it shouldn't move you.
The result: This framework eliminated 80% of my "shiny object" evaluations. It didn't make me more optimistic—it made me calibrated. I stopped being surprised by failures because I expected them at the base rate.
Framework 4: The Time-Delay Test (Temporal Discounting)
What it solves: Evaluating things based on when the payoff arrives, not just its size.
The human brain discounts future rewards exponentially. A $100 gain today feels better than a $150 gain in six months. But in evaluation, the delay itself carries information.
The framework: For any opportunity, ask:
- How long until the first measurable signal? (Not payoff—signal)
- If the signal takes longer than my "evaluation horizon" (usually 90 days), what proxy can I use?
- What is the decay rate of the value? (Some things appreciate, most things depreciate)
Data from my practice: I compared two investment strategies over 3 years. Strategy A had frequent, small wins (average 45-day cycle). Strategy B had rare, large wins (average 11-month cycle). Both had the same total ROI. But
Top comments (0)