Part 1 gave you the blueprints—the strategic thinking most organizations skip. Now we're stepping into the workshop.
This is where Adam Hopewell's frameworks shift from philosophy to practice. From "here's how to think" to "here's exactly what to do."
His central claim: hiring is a prediction problem. You're not filling vacancies—you're forecasting whether this specific person, with their specific wiring, will thrive in your specific environment and make your business measurably better.
That's a hard problem. But it's solvable if you stop winging it and start treating evaluation like a science experiment with clear hypotheses, data collection methods, and feedback loops.
What follows are the frameworks, metaphors, and systems that make prediction structured, scientific, and surprisingly human all at once.
The Talent Seed Metaphor: Beyond “Culture Fit” Most conversations about hiring still orbit around the fuzzy idea of culture fit. Hopewell finds that language too vague to be useful.
“I believe that there are three concepts working in parallel here. There is a talent seed, and every candidate is a talent seed. There is organizational soil, which is how an organization gets stuff done. And then there are the business skies, which is basically the business dynamic in which the organization is working and what the talent seed is going to be a part of.”
Every candidate brings a genetic makeup of sorts. Skills, motivations, habits, and behavioral tendencies make up the seed. The soil is how work actually happens inside your company. This includes the formal systems and the informal rules. It includes values, decision rights, pace, and how conflict gets resolved. The skies are the external conditions. These involve market cycles, competition, regulatory pressure, and your specific growth stage.
Hopewell presses the metaphor further.
“What we are trying to do in talent acquisition is reliably and validly predict if that talent seed put inside this organizational soil will germinate, sprout, grow, thrive, and will do some of the best work in its life and add to the entirety of the farm that we are creating under the business skies.”
This reframe matters because it makes evaluation concrete. Instead of vague judgments about whether someone fits, the question becomes: Will this seed thrive in this soil under these skies. In practice, that means asking specific, predictive questions that map to real conditions of work.
How does this candidate handle ambiguity when strategy pivots at short notice? How do they communicate in distributed or cross time zone teams? Will they stay resilient when storms hit, such as a sudden downturn, a competitive shock, or a resource squeeze? Do they have a history of learning new tools or domains quickly when the environment demands it? Can they influence without authority when decision making is consensus driven? The seed, soil, skies frame also clarifies failure analysis. If a hire struggles, you can ask whether the mismatch was in the seed, the soil, or the skies. Perhaps the seed was strong but the soil was rocky, meaning the way work happens in your company did not support the person’s strengths. Perhaps the skies changed and the role needed different capabilities six months later. This gives leaders language and signals for post mortems that are better than blame.
How to operationalize the metaphor? Create a one page seed profile for the role. List the must have competencies, motivational drivers, and non negotiables. Create a soil brief for the team. Describe pace, feedback norms, decision authority, and the two or three cultural traits that are most predictive of success. Write a skies snapshot that names the market conditions that will shape the work over the next year. Bring these three documents to the intake meeting. Now everyone is working from the same model.
Solve for X, Y, Z: Structured Prediction
The seed metaphor sets the stage, but prediction needs structure. Most interview loops jumble everything together—technical questions bleed into behavioral prompts, "culture fit" gets assessed alongside coding ability, and nobody can agree on what they're even evaluating.
When someone fails six months later, no one can say why. Was it technical capability? Work style? Bad timing? The data's too muddled to learn from.
Adam solves this by separating evaluation into three sequential stages: Solve for X, Solve for Y, and Solve for Z . Each stage answers a different predictive question and uses different instruments. When someone succeeds or fails, you can trace back to see which signals were predictive and which were noise.
Solve for X: Foundational Competencies X is about the basics: Do they have the core skills and experiences to even start the job?
Remember the Five Competency Rule from Part 1? This is where it kicks in. Those five criteria guide resume reviews and early screens. If a candidate doesn't meet a majority of the five, you stop there. No point investing interview hours in someone who lacks foundational requirements.
For a senior product manager, X might include discovery techniques, roadmap ownership at scale, data literacy, stakeholder management, and launch execution. For a staff engineer: systems design, production ownership, performance tuning, code quality at scale, and mentorship.
The screening questions map directly to these five. Simple. Focused. No surprises.
Solve for Y: Applied Capability Y asks a harder question: Can they actually do the work in practice?
This is where cognitive tests, technical assignments, and structured behavioral interviews come in. The goal isn't checking if someone knows the right buzzwords. It's watching them apply judgment under realistic constraints.
Good Y exercises share three traits:
Relevant to actual work, not abstract puzzlesTime-boxed and focused on one clear challengeTransparent about their thinking processFor a sales role, run a discovery call role-play with a messy customer brief. For a designer, critique two flawed flows and improve one in thirty minutes. For engineering, provide a small refactoring challenge with a failing test suite—not a whiteboard brain teaser.
Solve for Z: Behavioral and Cultural Integration Z is the hardest part. It's not about what the person can do—it's about how they do it.
Z examines pace, collaboration style, resilience under pressure, learning agility, and ethical judgment. This is where the seed-and-soil metaphor comes back. Will this person's working style thrive in your environment under your business pressures?
Z interviews should be structured and specific. Skip the generic "tell me about your strengths and weaknesses" garbage. Prompt for episodes that reveal habits and principles:
Describe a time when priorities shifted late and you had to renegotiate scope Walk me through a situation where a stakeholder disagreed with your recommendation and how you maintained trust while pushing back Give me an example of a significant mistake and the repair steps that followed Why Segmentation Matters Here's the payoff: segmenting evaluation creates a diagnostic feedback loop.
When a hire struggles, you can trace whether the issue came from X, Y, or Z. Weak X screens? Many candidates pass early stages but fail technical tasks. Weak Y? People clear exercises but underperform at 60 days. Weak Z? You see interpersonal friction, misalignment rework, or early attrition.
This granularity turns hiring outcomes into data rather than mystery. You're not just making hires—you're running experiments and learning which evaluation methods actually predict success.
Use the same 0-3-5 scale across X, Y, and Z. Zero = clearly absent. Three = standard or can't tell. Five = clearly exceptional with cited evidence (artifact, portfolio link, direct quote).
This reduces score inflation and forces interviewers to anchor judgments in observations rather than vibes. With structured evaluation in place, the next question becomes: how many candidates should move through each stage? That's where the funnel comes in.
The 12, 6, 3, 1 Funnel: Pressure Testing for Outliers Even with structured evaluation, prediction falters without scale. Here's the uncomfortable truth: great hires are statistical outliers. They're rare by definition. You can't find them by interviewing three people and hoping for the best.
Adam's solution is the 12-6-3-1 funnel:
"If you've got a single hire to make, I want to see the 12 best candidates possible at the sourcing stage. Out of a thousand applicants, my team finds the twelve best. From that twelve, we narrow to six through screening. From those six, three go to deep technical and behavioral interviews. And from those three, one hire gets made."
The specific numbers matter less than the discipline. Each stage raises the bar and shrinks the pool with intention. You're not processing applications—you're systematically hunting for the outlier.
The funnel also creates a built-in measurement framework:
Which sourcing channels produced the best twelve? Which screen questions best predicted who'd make the six? Which exercises best predicted who'd make the final three? Which signals correlated with the eventual standout? Adam uses a fishing metaphor: you cast nets in the right waters for a set time, bring in the catch, then evaluate against criteria. The funnel pressure-tests that catch until you identify the true outlier.
Running the Funnel Without Breaking It Time-box each stage and publish dates at intake. Commit to specific days for screening calls and panel interviews. No "we'll get back to you when we can" vagueness.
Share the five competencies with candidates before deep stages. Clarity about what you're evaluating isn't giving away answers—it's respecting their time and getting better signal.
Keep a running scorecard with X, Y, and Z subtotals. Use a tie-breaker rule: when two candidates are close, the one with stronger, evidence-backed 5s wins.
The Traps That Kill Your Data Don't over-index on brand names at the 12 stage. You want signal, not prestige. That ex-FAANG candidate might look great on paper but score 3s across the board in evaluation.
Don't let hiring managers skip stages because "this referral is really strong." Break the discipline, break the data. If you can't learn from exceptions, you can't improve the system.
Don't drag the process beyond your time box. The cost of delay is usually higher than the risk of deciding with the information you have. Speed is a feature, not a bug—if your evaluation is actually measuring the right things.
Why the Numbers Matter The 12-6-3-1 ratio isn't arbitrary—it's based on conversion math that accounts for signal degradation at each stage.
If your top-of-funnel is working (good sourcing, clear job specs), you should see roughly 50% conversion at each evaluation stage. The 12 candidates who passed X screening should yield 6 strong Y performers. Those 6 should produce 3 candidates where you'd genuinely be happy with any of them at the Z stage.
When this ratio breaks down, it tells you something:
Too many people advancing? Your evaluation criteria are too loose or your interviewers aren't calibratedToo few people advancing? Your sourcing is weak, your job specs are unrealistic, or your bar is miscalibratedRandom conversion rates? Your evaluation stages aren't actually measuring different things—you're just running the same test multiple timesThe funnel isn't just about finding one great hire. It's about creating a repeatable system that gets better every cycle because you can see exactly where a signal is generated and where it's lost. The funnel gets you to the right candidate. But how do you know if your prediction was actually correct? That's where the feedback loop comes in.
The 30-60-90 Loop: Teaching Your System to Learn
Here's the problem with most hiring systems: they're one-way streets. You make predictions, hire people, then... nothing. No feedback. No learning. Just rinse and repeat with the same broken assumptions.
Adam's 30-60-90 loop turns hiring into a two-way street. Every hire becomes a test of your prediction accuracy. Every success or struggle teaches you something about what actually matters.
How It Works Every new hire starts with clear, observable expectations tied to the five competencies. Set one or two concrete outcomes for each milestone—30, 60, and 90 days.
For a customer success manager: 30 days might mean running two supervised calls and documenting a playbook update. For engineering: 60 days could be owning a bug class and shipping a small feature behind a flag. For sales: 90 days might include hitting pipeline coverage and delivering a practice demo that earns a 5 from a senior peer.
At each checkpoint, the manager scores performance using the same 0-3-5 scale from interviews. A 5 requires evidence—not vibes. A 0 requires describing the gap in plain terms and naming the support offered.
This isn't a gotcha system. It's a learning loop for both the hire and the hiring process.
The Feedback That Actually Matters Those scores flow directly back into your evaluation design. If Solve for Y consistently predicts 60-day success, double down on that exercise. If Solve for Z isn't filtering out cultural misfits, recalibrate the prompts and retrain interviewers. If X screens correlate poorly with 90-day outcomes, revisit whether those five competencies are actually predictive for this role.
Make it visible : Publish a lightweight dashboard showing rolling 30-60-90 outcomes by cohort. Keep it anonymous outside the core team—you're tuning the system, not profiling individuals. Share one insight each quarter with leadership about what changed in your process as a result.
Close the loop with onboarding : Use those same five competencies to design onboarding. Assign buddies who are strong in different competencies. Align manager one-on-ones to the same language. When hiring and onboarding speak the same dialect, new hires climb faster.
Why This Actually Changes Things Talent density compounds. Every exceptional hire raises the bar and accelerates the team. Every weak hire pulls the average down. Adam's frameworks systematically tilt the scales toward density.
The Talent Seed gives you shared language for fit. X-Y-Z creates structure for prediction. The 12-6-3-1 funnel finds outliers at scale. The 30-60-90 loop makes the whole thing learn and improve.
This isn't intuition dressed up as science. It's science made usable. It respects the messy reality of organizations—asking for evidence where evidence is possible, judgment where judgment is needed, but always within shared structure.
That's the difference between companies that stumble through hiring and companies that use hiring as a competitive advantage.
What Changes When You Run This Hiring managers get their time back because early stages filter more effectively. They're not wasting hours on candidates who were never going to make it.
Recruiters build credibility because their process actually predicts back-end performance. They can point to data, not gut feel.
Candidates have clearer experiences because expectations are explicit. No more mystery-meat interview processes where nobody knows what's being evaluated.
Finance and Legal engage as partners because evaluation is explainable and decisions are auditable. No more black box hiring that legal can't defend.
Teams feel the lift. Work gets better because people are better matched to the work and to each other. That's not fluffy culture talk—that's measurable in velocity, quality, and retention.
So far, we've covered how to build a talent system that learns and compounds. But there's a bigger shift happening that changes everything: AI is entering the game on both sides of the table.
In Part 3, we'll look at how AI is reshaping this landscape, why the market is heading toward an optimization arms race, and why your competitive advantage won't come from better resume screening—it'll come from the structured, human-centered systems deeper in your funnel.
Read part 3 now
.svg)
.svg)
