A full case is a terrible daily workout. It is the right test, and a poor way to grow the specific muscle that just failed the test.
I score cases the way firms score them: structure, judgment, math, exhibits, synthesis, communication. When a candidate leaks on charts, the fix is not another 45-minute mock that also demands a MECE tree, a market size, and a recommendation. The fix is ten chart reps with a score on each one. That is what the CoachNed drill engine is for. I built it because I watched candidates burn weeks on "one more full case" and still miss the same move.
This piece is a product explainer, not a sales brochure. You will see what the six rooms actually cover, how the practice design maps to deliberate-practice research, a worked math drill you can reproduce on paper, and an honest list of what isolated drills cannot train. Then you can decide whether a week of open access is worth your time.
What interviewers score, and what candidates practice
In the room, I am not grading "case ability" as a single blob. I am listening for discrete moves. Did the tree cover the profit bridge without double-counting? Did the math land in the right units? Did the candidate read the chart before inventing a story? Did the recommendation open with a decision, then evidence, then a risk?
Most candidates practice the opposite of how scoring works. They sit one full case, receive a foggy debrief ("be more structured"), and schedule another full case. Volume without isolation feels productive. It rarely is.
The research name for the better pattern is deliberate practice: a specific goal, a task designed for that goal, immediate informative feedback, and enough repetition to change the next attempt. Ericsson, Krampe, and Tesch-Römer laid that out in their 1993 Psychological Review paper. Hours of "doing cases" do not equal that regimen. A drill is closer: one prompt, one skill, a score you can act on before the next rep.
What the drill engine is (and is not)
CoachNed's drill surface is six skill rooms plus a free-forever quick-math track. As of the public drills page in 2026, the bank sits at 550+ drills (the page currently shows 554 plus quick-math sprints). You can speak or type. You can compare your answer to the model answer as often as you want. Official graded reps are capped at three a day for your first three days; after that, paid plans open unlimited grading on the same bank.
| Room | Path | What you practice | What I listen for in a live case |
|---|---|---|---|
| Structure | /drills/structure | Issue trees from a short prompt | MECE coverage, prioritization, no kitchen-sink lists |
| Math | /drills/math | Profitability, growth, unit economics | Formula first, staged calc, units, a one-line so-what |
| Charts | /drills/charts | Reading a figure under time | Correct read before implication; no invented axes |
| Market sizing | /drills/market-sizing | Top-down or bottom-up estimates | Stated assumptions, sanity checks, business translation |
| Synthesis | /drills/synthesis | Decision + evidence + risk | Recommendation-first order; no dump of analysis |
| Brainstorming | /drills/brainstorming | Drivers and ideas under MECE pressure | Breadth without chaos; buckets that do not overlap |
| Quick math | /drills/quick-math | Timed mental-math sprints | Speed without losing the business meaning |
The public rooms also show rough bank sizes where listed (for example, about 90 structure and brainstorming reps, about 120 math). Treat those as inventory, not as a claim that you need every item. The point is enough variety that you stop memorizing one prompt.
Ned built CoachNed as an independent practice product. The drill engine is the daily layer; full spoken cases and assessment simulators are the integration layer. Drills are not a substitute for sitting a case out loud.
A worked math drill (reproduce it on paper)
Here is the kind of prompt a math drill throws. Invented company, real arithmetic.
Prompt. HarborLine Logistics runs 40 delivery vans. Each van averages 22 deliveries per weekday. Contribution per delivery is $18 after variable costs. Fixed costs for the fleet are $48,000 per weekday. What is daily contribution after fixed costs, and what happens to that number if deliveries per van fall 10%?
A clean path.
- Deliveries per day: 40 × 22 = 880.
- Contribution before fixed: 880 × $18 = $15,840.
- After fixed: $15,840 − $48,000 = −$32,160 (losing money).
- At −10% deliveries: 19.8 deliveries per van → 40 × 19.8 = 792 deliveries.
- 792 × $18 = $14,256; after fixed: $14,256 − $48,000 = −$33,744.
- So-what: the fleet is already underwater on contribution after fixed; a 10% volume miss widens the hole by about $1,600 per day, so volume alone does not fix the cost base.
What the grade cares about is not speed-for-its-own-sake. It is the equation, the staged numbers, and whether you stopped with a business sentence. That is the same habit I want inside a full case when the interviewer hands you a P&L extract.
Why isolation beats another full mock (for the skill that failed)
After a weak chart turn in a live interview, candidates often schedule another full case "to rebuild confidence." They then spend twenty minutes structuring and ten minutes math before they touch a chart again. The leak gets maybe one exposure.
Deliberate practice argues for the opposite schedule: design a task that attacks the weak mechanism, get feedback, repeat while attention is still high. Ericsson's later writing (including the popular summary in Peak, with Robert Pool) stresses that the activity must target a specific aspect of performance with feedback, not merely accumulate domain hours. Mixing only whole performances is closer to playing the concert than to practicing the hard bar.
There is a second scheduling point. Shea and Morgan's 1979 study on contextual interference found that random (interleaved) practice of related tasks hurt performance during acquisition but helped retention and transfer relative to blocked practice. For case prep, that is a mild warning: do not grind thirty identical math prompts in a row and call yourself ready. Rotate rooms across the week. A structure day, a chart day, a synthesis day will feel harder than a math binge. Harder during practice is often the point.
Ned's rule. If a full-case debrief names one leak, run five drills on that leak before you sit another full case. The mock is the exam. The drills are the homework.
What drills cannot train
Be blunt with yourself. An isolated drill will not teach:
- Presence under a live interviewer. Eye contact, silence, and recovering when you blank are social skills. They need a spoken case with someone on the other side of the table or a voice interview that waits for you.
- Hint-taking and pushback. Real interviewers nudge. Drills give a score; they do not negotiate with you mid-answer the way a partner does.
- End-to-end case management. Knowing when to stop structuring and start quantifying is an integration skill. Only full cases force the handoff.
- Behavioral and PEI content. That is a different product surface (/behavioral), not a drill room.
- Firm-specific assessment formats. Casey, CCA, Solve-style games, and SOVA need their own simulators. Drills build the underlying skills those tests still require, but they are not the test.
If you only drill, you will get sharper at the skill and still sound thin in the room. If you only do full cases, you will stay vague on the skill that keeps costing you points. Use both.
How I would use the rooms in a recruiting week
Assume you have already sat one or two full cases and you know the leak. A practical week looks like this. Numbers are my estimate of useful load, not a published study result.
| Day | Assignment | Time (est.) | Why |
|---|---|---|---|
| Mon | 2–3 math or quick-math sprints | 15 min | Warm the formula + units habit |
| Tue | 2 structure drills | 15 min | Rebuild the tree without a full case |
| Wed | 1 charts + 1 synthesis | 20 min | Most candidates underweight both |
| Thu | 1 market sizing + 1 brainstorming | 20 min | Estimation logic and MECE ideation |
| Fri | Full spoken case at /interview | ~18 min + debrief | Integration test; note the next leak |
| Sat | Drill only the Friday leak | 15 min | Close the loop before Monday |
Quick math at /drills/quick-math stays free forever after the trial week. Use it as a warm-up, not as the whole plan. The six graded rooms are where the skill isolation happens.
After a full case, the debrief names a weak dimension and can hand you the matching drill room. That mapping is deliberate: drills and cases share the same skill vocabulary so the homework matches the scorecard.
Access, pricing, and what "open" means
Describe this honestly, because prep sites often do not.
Everything is open for seven days with no card: live interviews with Ned, seven-score debriefs, the drill bank, Casey, CCA extras, PEI, resume report. After the week, keep going with a Season Pass at $120 for 90 days, or Pro at $49 a month. Every paid plan is the whole stack; there is nothing to add on. There is a 14-day no-questions refund. Annual at $249 exists if you are recruiting across two cycles; pick Season Pass or Pro if you are in one season.
After the free week, Quick math, sample walkthroughs, the AI lesson, and the first two courses stay free forever. Your history stays either way.
Start without deciding on a plan: the five-minute first rep at /start, or open a room from /drills/structure and take one graded prompt.
Practice this today
Do not schedule another full case yet. Open the room that matches your last known leak. If you do not know the leak, start with structure: build a three-to-four-branch tree for a short prompt, submit, and read the model. Then run one math drill. That is thirty minutes of deliberate practice, not another foggy mock.
If you want the integration check afterward, sit CopperCup on voice at /interview once, take the seven-score debrief, and let it name the next drill. Seven days open; then Season Pass or Pro if the reps are working.
Frequently asked questions
How many drills does CoachNed have?
As of the public drills page in 2026, the bank is 550+ drills (listed as 554) across structure, brainstorming, math, market sizing, charts, and synthesis, plus separate quick-math sprints. Exact counts by room move as content is added; use the room pages for the current inventory.
Is the CoachNed drill engine free?
Everything, including unlimited access to the drill rooms and grading under the trial rules, is open for seven days with no card. After that, Quick math stays free forever. Full unlimited grading on the six skill rooms sits inside Season Pass ($120 / 90 days) or Pro ($49 / month), with a 14-day no-questions refund.
Can drills replace full case practice?
No. Drills isolate skills with feedback. Full spoken cases train integration, presence, and interviewer interaction. I want both: drills for the leak, cases for the exam.
Do I need an account to try a drill?
You can begin with the free starter flow at /start. The drill rooms themselves are part of the free week once you are in; Quick math remains available after the week without a paid plan.
What is the difference between math drills and Quick math?
Math drills are full business prompts with AI grading on equation, stages, units, and so-what. Quick math is a timed sprint track (guided warmup plus five-minute sprints) for speed. Use Quick math to warm up; use math drills to train the case habit.
Sources
- The role of deliberate practice in the acquisition of expert performance (APA PsycNet record) — Ericsson, Krampe, and Tesch-Römer (1993) framework for goal-directed practice with feedback. Checked 2026-09-24.
- Contextual interference effects on the acquisition, retention, and transfer of a motor skill (DOI) — Shea and Morgan (1979) on random vs. blocked practice and retention/transfer. Checked 2026-09-24.
- Peak (Penguin Random House / Bodley Head listing) — Ericsson and Pool's accessible summary of deliberate vs. naive practice. Checked 2026-09-24.
CoachNed is independent and not affiliated with McKinsey, BCG, Bain, Deloitte, or any firm named here.