Most candidates who search "Ecosystem Building" in 2026 are preparing for a game that is no longer the default McKinsey Solve sitting. The thesis is simple: treat McKinsey's own assessment pages as the roster authority, treat forum module names and timings as candidate reports, and spend your hours on the games your invitation actually implies. After this piece you should be able to separate confirmed facts from folklore, run a constraint-chain check the way the reef module once rewarded, and know which CoachNed sims match the live Solve family.
McKinsey's public careers material describes Solve as a gamified assessment of problem-solving, says there is no need to prepare in the classic study sense, and is clear that integrity violations (AI tools, screenshots, help from others) can end an application. It does not publish a fixed 2026 line-up of game titles on that page. That gap is where prep blogs invent certainty.
What McKinsey has actually said in public
Start with primary text, not with someone else's "complete guide."
In its early recruiting write-up, McKinsey framed the assessment around wildlife and habitats and used the line build a reef, save a species. That is the public origin story for what candidates later nicknamed Ecosystem Building: place organisms so a habitat stays viable under diet and environment rules. Separately, a careers piece on the game's genesis said applicants are offered one to three tasks from a library set under water, on mountain tops, and in prairies, and that the online version generating many permutations went live around Fall 2019 (Innovating recruiting through online gaming).
More recently, McKinsey's game-based innovation lab interview names Sea Wolf explicitly: candidates select and evaluate microbes to clean contaminated sites, as part of Solve, which the firm says reaches on the order of 300,000 early-career consultant candidates per year (From apprenticeship in space to selecting microbes). That is firm language for a current module. It is not a guarantee that every invite includes Sea Wolf and nothing else.
The Solve FAQ PDF (linked from the careers assessment page) adds operational facts that still matter: your invitation email states the assessment length; each task starts with an untimed tutorial; scenarios vary so two people rarely see identical parameters; results sit alongside the rest of the application. McKinsey still does not list "Ecosystem Building," "Redrock," or "Sustainable Futures Lab" as named chapters in that FAQ.
Confirmed roster versus candidate reports
Use this table as a filter, not as a prophecy.
| Claim | Status as of 2026-09-24 | Source type |
|---|---|---|
| Solve is a gamified recruiting assessment; no business knowledge required | Confirmed | McKinsey careers / FAQ |
| Assessment length is in the invitation; tutorials are untimed | Confirmed | McKinsey FAQ |
| Task library historically included wildlife / habitat scenarios ("reef") | Confirmed (historical design) | McKinsey recruiting blogs |
| Sea Wolf (microbe selection for site cleanup) is used in Solve | Confirmed (named by McKinsey) | McKinsey GBIL interview |
| Ecosystem Building = one closed food chain in a single habitat | Candidate-reported label for the reef-style task | Forums / prep chatter |
| Default ~65-minute sitting is Redrock plus Sea Wolf | Candidate-reported, widely repeated | Not on McKinsey's roster page |
| Longer ~85-minute sittings add Sustainable Futures Lab (or similar) | Candidate-reported | Not on McKinsey's roster page |
| Plant Defense and Ecosystem Building are retired from the common path | Candidate-reported | Consistent with McKinsey naming Sea Wolf instead of reef in newer public copy |
| Exact pass rate, scoring weights, or fixed species count | Not published | Folklore if stated as fact |
If your email says roughly 65 minutes and two tasks, I would not spend the week drawing food webs. If the in-test screen or a recruiter note names a reef, ecosystem, or food-chain task, the constraint logic below still applies. Read the invite and the tutorial. Those beat any blog, including this one.
The skill the reef was measuring
From the interviewer's side, Ecosystem Building was never about memorizing coral biology. It was a constraint-satisfaction drill dressed as ecology.
You receive a habitat with abiotic rules (depth, temperature, substrate) and a deck of species cards with roles (producer, grazer, predator) and diets. A legal build is one where every non-producer has a present, legal food source and no species sits outside its terrain. An illegal build is anything that looks "complete" but leaves a mid-level consumer with nothing it can actually eat, or parks a deep-only animal on a shallow reef.
That maps to case work more cleanly than people admit. When a PEI or case interviewee stacks recommendations without checking feasibility, I score the same failure: decorative producers, hungry middle, apex species with no base. The reef just made the failure visible on a board.
What I do not treat as confirmed: a magic target of eight species, a fixed map bank, or "diversity always wins." Those are folklore numbers. The tutorial in front of you defines the objective for that permutation.
Ned's rule. Before you place a single predator, name the producer that feeds the bottom of the chain and verify every arrow down to it. If you cannot draw that path on a napkin, the board is not ready, and neither is the recommendation.
Worked check: seven cards, two illegal diets
Invented habitat for practice only. Not a McKinsey item.
Habitat. Temperate kelp shelf: cool water, mid depth, rocky. Illegal for tropical-only and deep-only cards.
Cards available (you will not use all).
| Species | Role | Diet on card | Terrain |
|---|---|---|---|
| Kelp | Producer | Light / nutrients | Mid rock |
| Drift algae | Producer | Light | Mid |
| Sea urchin | Grazer | Algae | Mid |
| Rock crab | Mid | Invertebrates | Mid |
| Kelp bass | Mid | Small fish / crabs | Mid or shallow |
| Harbor seal | Apex | Fish | Mid or shallow |
| Lantern squid | Mid | Plankton | Deep only |
Scoring heuristic I use when coaching this style of puzzle (Ned's estimate, not a firm rubric): start at 100. Deduct 25 for any consumer with no legal food present. Deduct 25 for any terrain violation. Deduct 10 for an apex with only one fragile mid-link and no second path. Deduct nothing for leaving cards unused.
Bad board (score 50). Kelp + rock crab + harbor seal. Kelp is a producer, so the board "has biomass." The crab's diet is invertebrates, not kelp, so the crab is starved (−25). The seal needs fish; you placed no fish (−25). Two failure modes, one decorative producer.
Still bad (score 75). Kelp + urchin + bass + seal + lantern squid. Urchin eats algae (legal). Bass does not eat urchin on this card set and no crab is present (−25). Squid is deep-only on a mid shelf (−25).
Legal board (score 100 under this heuristic). Algae → urchin; add bass only if legal prey (crab or small fish) is present; add seal only if bass (or another fish) is present and the seal card allows that prey. Leave lantern squid in the deck. Unused cards are not a penalty here; illegal cards are.
Reproduce the check: write three columns — Species, Food present (Y/N), Terrain legal (Y/N). Any N fails the build. That napkin test is the whole game.
Prep allocation when your invite is ambiguous
Candidates waste weeks on the wrong module mix. Use time the way a staffing partner would allocate analysts.
| If your invite / tutorial suggests… | Prep weight (Ned's split) | What to skip |
|---|---|---|
| ~65 minutes, two tasks, data + microbe language | 50% Redrock-style chart math, 40% Sea Wolf constraints, 10% reef skim | Memorizing food-web maps |
| Longer sitting, third task named or hinted | Keep the 50/40 split, add 15–20% multi-objective tradeoff practice for SFL-style decisions | Treating Ecosystem Building as the main event |
| Explicit reef / ecosystem / food-chain language | 40% closed-chain constraint drills, 30% Sea Wolf transfer, 30% Redrock | Leaked "real" ecosystems |
| Vague email, only a total duration | Follow duration heuristics carefully, confirm names on the first tutorial screen | Pre-committing to one blog's roster |
McKinsey's FAQ is explicit that length lives in the email. Candidate reports that map 65 minutes to Redrock plus Sea Wolf, and longer invites to an added lab module, are useful planning assumptions. They are still reports. If the first tutorial names something else, pivot in the room; do not argue with the screen.
Integrity is not optional color. The Solve assessment page bars AI tools, screenshots, and help from others except approved disability accommodations. Chasing a leaked reef map is both a values problem and a practical one: permutations differ by design.
Practice the live modules, not the folklore map
If Ecosystem Building appears, the transferable habit is filter-infeasible-then-place. That habit also shows up in Sea Wolf-style site matching. If it does not appear, the same habit still helps cases: kill illegal branches before you optimize.
CoachNed's Solve-style practice for the modules candidates most often see lives at /mckinsey-sea-wolf, /mckinsey-red-rock, and /mckinsey-sfl. Those are independent sims for format familiarity, not McKinsey's exam and not leaked item banks. For the constraint habit itself, a short structure filter drill also helps:
Assignment for today (40 minutes). Open Sea Wolf-style practice and force yourself to list hard constraints before you maximize any score. Then run the seven-card table above on paper with a timer set to eight minutes. Grade yourself only on the napkin columns (food present, terrain legal). Everything is open on CoachNed for seven days with no card; then $120 for a recruiting season or $49 a month. After Solve, the offer still turns on cases and PEI, not on whether you once built a pretty reef.
CoachNed is not affiliated with McKinsey & Company.
Frequently asked questions
Is Ecosystem Building still part of McKinsey Solve?
McKinsey does not publish a named 2026 roster that lists Ecosystem Building as a current default module. It has publicly described reef-style wildlife tasks historically and now names Sea Wolf in firm materials. Whether a food-chain task still appears for some roles or regions is candidate-reported; confirm from your invitation and tutorials.
How is Ecosystem Building different from Sea Wolf?
As candidates describe them, Ecosystem Building is one habitat and a closed food chain; Sea Wolf is multi-site microbe (or species) matching under attribute ranges. McKinsey has confirmed Sea Wolf by name. Treat detailed Sea Wolf scoring rules from forums as reports until your tutorial says otherwise. See the Sea Wolf guide.
What does McKinsey say about preparing for Solve?
On the official assessment page, McKinsey says there is no need to prepare and to approach the games with curiosity. The FAQ adds that each task has an untimed tutorial. Format familiarity is your choice; leaked maps and prohibited tools are not.
Should I practice Plant Defense or Sustainable Futures Lab instead?
Plant Defense is widely described as an older library task. Sustainable Futures Lab (or similarly named lab modules) appears in candidate reports of longer sittings. Neither is specified as a fixed chapter on McKinsey's public Solve overview. Weight prep to what your invite length and on-screen names support, and use /mckinsey-sfl only as independent tradeoff practice.
Does a strong Solve score replace the case interview?
No. McKinsey states Solve results are considered with the rest of the application and other assessments. Live cases and PEI remain the hiring conversation. Use Solve as one data point, then practice interviewer-led cases.
Sources
- Solve, McKinsey's assessment game — official description, integrity rules, "no need to prepare," results considered with the rest of the application. Checked 2026-09-24.
- Solve FAQs (PDF) — invitation length, untimed tutorials, scenario variation, multi-factor decision. Checked 2026-09-24.
- Build a reef? Save a species? — McKinsey's public framing of wildlife / reef-style digital recruiting tasks. Checked 2026-09-24.
- Innovating recruiting through online gaming — one-to-three tasks from a habitat library; online deployment timing. Checked 2026-09-24.
- From apprenticeship in space to selecting microbes — Sea Wolf named; ~300,000 early-career Solve candidates per year. Checked 2026-09-24.