McKinsey Solve Ecosystem Building: Species Selection, Scoring, and Full Strategy (2026)

McKinsey Ecosystem Building is the hardest module in the McKinsey Solve game (also called McKinsey PSG — Problem Solving Game). Most candidates fail it not because they pick wrong species — but because they optimize for the wrong criteria. They treat it as a biology quiz, trying to identify which animals in real life eat other animals. McKinsey built this game to test systems thinking, structured decision-making under uncertainty, and iteration behavior. A candidate who builds an imperfect ecosystem through clear, systematic logic consistently outscores a candidate who builds a better ecosystem through instinct and guessing.

This guide covers what the game actually tests, how to work through the 39 species in under 3 minutes, how to build a sustainable food chain, and what McKinsey tracks in your process that most candidates never think about.

Important context: McKinsey Solve has evolved significantly since 2020. Ecosystem Building was part of the original Solve suite but has been phased out from July 2025 for most candidates; candidates with a 90-100 minute Solve invitation may still encounter it. This guide is written for candidates who receive a 90-100 minute Solve invitation or who want to understand Ecosystem Building for historical context. If you're taking Solve in 2026, confirm which modules you'll face before investing prep time here.

What Ecosystem Building Tests (System Thinking, Not Biology Knowledge)

McKinsey partnered with Imbellus to build Solve on a specific premise: the skills that make a great consultant — structured problem decomposition, hypothesis formation, iteration under uncertainty — can be measured through game-based simulations without requiring business knowledge. Ecosystem Building is the clearest expression of this principle.

The game is set in an ecological context so that prior knowledge (consulting frameworks, financial modeling) gives zero advantage. Everyone starts equal. The advantage goes to candidates who:

1. Decompose the problem systematically — understanding that "build a food chain" actually means "select an apex predator, then identify sufficient prey at each trophic level, ensuring the location supports all selected species"
2. Form and test hypotheses — starting with a hypothesis about which species to include rather than randomly exploring the full pool
3. Iterate based on feedback — when the game shows a species can't survive (because it has no prey or the location conditions are wrong), updating your chain rather than starting over or ignoring the signal
4. Allocate time efficiently — recognizing which decisions are reversible (species swaps) vs. which constrain everything downstream (apex predator selection, calorie surplus validation)

Biology knowledge helps at the margin. If you already know that wolves eat elk and elk eat grass, you'll confirm those connections faster than someone who doesn't. But CaseBasix's Ecosystem Building research shows that candidates with biology backgrounds do not outperform non-biology candidates at a statistically significant rate — because the game's species relationships are displayed in the UI. You don't need to know them from memory.

The 39 Species: How to Cluster and Filter in Under 3 Minutes

The game presents approximately 39 species, each with a profile card showing: habitat type (terrestrial, aquatic, mixed), temperature range, elevation or depth tolerance, diet type, and predator/prey relationships. Processing 39 cards before making any decision is the wrong approach. Filter in sequence.

Step 1: Confirm your assigned scenario (first 60 seconds)

Your scenario is assigned randomly — you do not choose it. When the game loads, you will be placed in one of three scenario types: coral reef, mountain ridge, or jungle. Each scenario type has its own group of approximately 13 species. Spend the first 60 seconds confirming which scenario you've been assigned, then immediately begin filtering species from your assigned group. Do not waste time wishing for a different scenario — all three are equally solvable with the right approach.

Step 2: Identify apex predators in your scenario group (2-3 minutes)

An apex predator is a species that has no natural predators in your ecosystem. Start here — the apex predator defines the top of your chain, and everything else must support it. Look for species cards showing "No natural predators in this region" or equivalent language. In a jungle scenario this might be a jaguar or harpy eagle; in a mountain scenario, a mountain lion or golden eagle; in a reef scenario, a shark or large marine predator.

Select 1-2 viable apex predators from your scenario group. If one doesn't work later (e.g., fails the calorie surplus check with available prey), having a backup saves you from restarting.

Step 3: Filter for mid-level predators and herbivores (3-4 minutes)

Work down from the apex. What does your apex predator eat? Filter for species in your scenario group that match those prey items. Then filter for what those prey species eat — typically herbivores. Build the chain downward: apex predator → mid-level predator or large herbivore → small herbivore → producer (plant species). At each link, verify the calorie surplus: the prey must provide more calories than the predator requires.

A viable chain structure in a jungle scenario might look like:

• Apex (L4): Jaguar
• Mid-predator (L3): Ocelot
• Large herbivore (L3): Tapir
• Small herbivore/omnivore (L2): Agouti
• Producer (L1): Fig tree, broadleaf shrub

Step 4: Verify calorie surplus at every link and check ecological balance

For each predator-prey link in your chain, confirm the prey species' calorie value exceeds the predator's calorie requirement. A link that fails this check will cause the chain to collapse regardless of ecological plausibility. Beyond calorie balance, aim for more prey species than predators at each level — the chain needs a stable base to support its upper levels.

Food Chain Logic: Building a Sustainable Predator-Prey Loop

"Sustainable" in Ecosystem Building has a specific meaning: each species must have sufficient food sources to survive, and predators cannot drive their prey to extinction. The game models population dynamics — if a predator cannot meet its calorie requirement from available prey, it cannot survive in your ecosystem. An unsustainable chain collapses and scores poorly even if the species are individually plausible.

The calorie surplus mechanic — the actual filter to apply:

Each species card shows a calorie value: how many calories it provides if eaten. Predators also have a calorie requirement — how many calories they need from their prey to survive. A valid predator-prey link requires the prey to provide MORE calories than the predator requires. When building your chain, check that each predator-prey link produces a calorie surplus. A chain where a predator is calorie-deficient will fail even if the species types are logically compatible.

Apply this filter at every link as you build downward:

1. Does your apex predator's prey provide enough calories to meet the apex's requirement?
2. Does the mid-level predator's prey provide enough calories to meet its requirement?
3. Does the herbivore's food source (producers) provide enough calories to sustain it?

A species that is ecologically plausible prey but calorie-insufficient is not a valid link. Replace it with a species that clears the calorie surplus threshold.

Connected vs. disconnected chains:

Every species in your ecosystem must be connected to the food chain — either as prey for something above it, or as a predator consuming something below it, or as a producer at the base. Orphaned species (present in your ecosystem but not connected to anything) count against your product score. Before submitting, trace every species: is it eaten by something? Does it eat something? If neither, remove it.

Producer layer:

Always include 2-3 producer species (plants, grasses, shrubs). These form the base of the food chain and are often overlooked by candidates who focus on the more interesting predator-prey relationships. An ecosystem with no producers has no chain base and will score very low regardless of how good the upper trophic levels are.

Your Assigned Scenario: What Changes Between Reef, Forest, and Mountain

Your scenario is assigned randomly at the start of each Ecosystem Building session — you do not choose it. There are three scenario types: coral reef, mountain ridge, and jungle (forest). Each scenario type maps to its own group of approximately 13 species. What changes between scenarios is the species pool available to you; the underlying chain-building logic is identical across all three.

What changes per scenario:

1. Coral reef: Species include reef fish, filter feeders, large predatory fish, and marine producers such as algae and coral organisms. The chain runs from producers through small grazers up to apex marine predators.

2. Mountain ridge: Species include alpine mammals, birds of prey, small rodents, mountain ungulates, and hardy plant producers. Temperature and elevation tolerances are the key compatibility filters.

3. Jungle (forest): Species include rainforest insects, reptiles, mid-level mammals, large apex predators, and diverse plant producers. Diet breadth and canopy level tend to be the key compatibility filters.

How to orient in the first 60 seconds:

When the game loads, identify your scenario type immediately. Look at the environment displayed and the species names visible. Once you know whether you're in a reef, mountain, or jungle, you know which third of the species pool is yours. Candidates who spend time examining species outside their scenario group waste critical minutes on irrelevant information.

What does NOT change:

• The chain-building logic (top-down from apex predator, with sufficient prey at each level)
• The calorie surplus mechanic (each predator-prey link must produce a surplus — see Food Chain Logic below)
• The process score criteria (systematic navigation, purposeful iteration, connectivity checking)
• The time limit: 35 minutes regardless of scenario

What McKinsey Actually Tracks Beyond Your Final Chain

This is the piece that changes how you should play the game from start to finish. McKinsey's platform tracks your behavior throughout the session, not just your final ecosystem. The process score captures:

Reading behavior: Do you open and read species cards systematically, or do you jump between cards randomly? Systematic readers — who open cards in a logical order (e.g., all apex candidates first, then mid-level, then herbivores) — signal structured problem-solving. Random card-opening generates a lower process score.

Placement iteration: When you add a species to your chain and then remove it, does your behavior suggest you've learned something (you immediately try a different species) or that you're confused (you cycle through the same 3 species repeatedly)? Purposeful iteration is rewarded; circular iteration is penalized.

Time allocation: Are you spending proportional time on high-leverage decisions (scenario orientation, apex predator selection) vs. low-leverage ones (which of two similar herbivore species to include)? Candidates who spend 15 minutes on herbivore selection after spending only 2 minutes on apex predator identification signal inverted priorities.

Decision speed after information: How quickly do you make a decision after opening a species card? Pausing for 5-10 seconds after reading a card (to process and integrate information) is normal and good. Clicking away in under 2 seconds signals you didn't read it; taking 45+ seconds suggests indecision.

Chain connectivity checking: Do you verify that your chain is connected before finalizing? Candidates who visually trace their chain and remove orphaned species signal quality control — a core consulting behavior.

From Wall Street Oasis community data on McKinsey Solve, candidates who rank in the top 30% on process score advance to case interviews at 2.3x the rate of candidates with equivalent product scores but lower process scores. Process is not secondary — it may matter more.

Full 35-Minute Walkthrough: Building a Winning Chain Step by Step

This walkthrough uses a jungle (forest) scenario as the example — the same logic applies to reef and mountain scenarios with different species.

Minutes 0-2: Scenario orientation

The game loads. Identify your assigned scenario immediately — in this example, jungle (forest). This tells you which group of ~13 species is yours. Spend 60 seconds confirming the scenario type, then move directly to species cards. Do not explore the full 39-species pool — filter to your scenario group only.

Minutes 2-8: Apex predator identification

Open species cards from your assigned group systematically — start with the largest carnivores. You're looking for: (1) no natural predators listed, (2) habitat listed as compatible with your scenario, (3) a high calorie requirement (apex predators tend to have the highest requirements). Identify 2 candidate apex predators. In this example: jaguar (viable) and harpy eagle (viable). Select jaguar as primary; note harpy eagle as backup.

Minutes 8-14: Chain construction (top-down) with calorie surplus checks

Jaguar eats: tapir, peccary, large rodents. Find tapir card — check that tapir's calorie value exceeds jaguar's calorie requirement. If yes, add it. Find peccary — same calorie check. Add it if it clears the threshold. Tapir eats: fruit, leaves, aquatic plants. Find producer species that match: fig tree (viable), broadleaf shrub (viable). Add both.

Now add mid-level: find ocelot (eats agoutis and lizards, compatible with jungle). Check calorie link: do agoutis provide enough calories for the ocelot? Yes — add ocelot. Find agouti — eats seeds and fruit, viable. Add agouti.

Chain so far: Jaguar → Tapir, Peccary → Fig tree, Shrub. Jaguar → Ocelot → Agouti → Seeds/Fruit producers.

Count: 1 apex + 1 mid-predator + 2 large herbivores + 1 small omnivore + 2 producers = 7 species across 4 trophic levels.

Minutes 14-28: Expand and verify calorie links

Look for additional species in your group that extend the chain or strengthen existing links. For each new species added, confirm the calorie surplus on both the link above (what eats it?) and the link below (what does it eat?). Aim for 8-12 species. Add 1-2 more producers to strengthen the base.

Minutes 28-33: Connectivity check and refinement

Trace every species: is it connected? Remove any orphaned species. Verify that no predator in your chain has a calorie deficit from its prey. If any link fails the calorie surplus check, replace the deficient species now.

Minutes 33-35: Final review and submit

Quick scan of all placed species. Verify no duplicate niche species. Submit with at least 2 minutes to spare — do not let the clock run out on a disconnected chain.

Common Ecosystem Building Mistakes (and What the Data Says)

Mistake 1: Exploring the wrong scenario group. Your scenario is assigned — not chosen. Candidates who spend time reading species cards outside their assigned scenario group (reef, mountain, or jungle) waste 5-10 minutes on irrelevant information. Confirm your scenario type in the first 60 seconds, then filter immediately to your group of ~13 species. Hacking the Case Interview's Solve guide notes that misallocated early time is the most common structural error.

Mistake 2: Ignoring the producer layer. Approximately 25% of candidates fail to include sufficient producer (plant) species at the base of their chain. Without producers, herbivores have nothing to eat and the chain collapses. Always include 2-3 producer species.

Mistake 3: Overloading predators. Some candidates add 4-5 apex or near-apex predators to make their ecosystem feel "complete." This produces an ecologically unstable chain — too many predators, not enough prey. The predator-to-prey ratio should be roughly 1:3 or better at each level.

Mistake 4: Circular iteration. Swapping the same species in and out repeatedly without changing strategy generates a negative process score signal. If a species isn't working, diagnose why (wrong habitat? no prey?) before deciding what to replace it with.

Mistake 5: Running out of time with a disconnected chain. Candidates who run out of time submit whatever is on screen — including orphaned species and broken chain connections. A 6-species connected chain scores better than a 12-species disconnected one. Set a 5-minute buffer for connectivity checking.

Comparison: Ecosystem Building vs Redrock Study vs Sea Wolf

Understanding how Ecosystem Building differs from the current Solve modules clarifies what skills transfer and what doesn't.

The skills that transfer from Ecosystem Building to current Solve modules: systematic navigation, iteration behavior, and time allocation under pressure. The specific strategies (top-down species selection, food chain logic) are unique to Ecosystem Building. See the McKinsey Solve guide for current module strategies.

30-Day McKinsey Solve Prep Plan

If you're facing the current Solve format (Redrock Study + Sea Wolf), see the McKinsey Solve guide. If you're preparing for Ecosystem Building specifically:

Test Your Knowledge

Practice Drills: Ecosystem Building Scenarios

Related Guides

• McKinsey Solve guide — the current Solve format with Redrock Study and Sea Wolf strategies
• McKinsey Redrock Study guide — specific strategy for the data interpretation module
• McKinsey Sea Wolf guide — specific strategy for the ecosystem management simulation
• McKinsey case interview guide — how to prepare for live McKinsey case interviews after passing Solve
• McKinsey PEI guide — the Personal Experience Interview that accompanies every McKinsey case round
• What is a case interview — foundational overview for candidates new to consulting interviews
• AI case interview practice guide — how to use AI tools to build case skills efficiently

Sources and Further Reading (checked March 7, 2026)

• McKinsey Solve official page: https://www.mckinsey.com/careers/interviewing/the-mckinsey-problem-solving-game
• IGotAnOffer McKinsey Solve guide: https://igotanoffer.com/blogs/mckinsey-case-interview-blog/mckinsey-solve-guide
• Hacking the Case Interview, McKinsey Problem Solving Game: https://www.hackingthecaseinterview.com/pages/mckinsey-problem-solving-game
• CaseBasix McKinsey Solve Ecosystem Building: https://www.casebasix.com/mckinsey-solve-ecosystem-building/
• Wall Street Oasis McKinsey Solve tips and discussion: https://www.wallstreetoasis.com/forum/consulting/mckinsey-solve-tips