Consulting Skills for Interviews: Contracting, Diagnosis and Recommendation
One consultant walks into a meeting and hears, “Our productivity is low,” then spends three weeks building dashboards nobody asked for. Another pauses, contracts the problem, diagnoses the real constraint, and recommends a decision the client can actually make on Monday.
- Contracting defines the work before the work: objective, scope, roles, access, timeline, decision rights and success measures.
- Diagnosis turns symptoms into causes using hypotheses, issue trees, data, stakeholder interviews and triangulation.
- Recommendation is not a slide deck summary; it is a clear decision backed by logic, evidence, trade-offs and implementation steps.
- The golden sequence is: contract the question - diagnose the cause - recommend the choice - secure commitment.
- Use MECE thinking in diagnosis: no overlaps, no gaps in how you break the problem.
- The best recommendation answers four questions: What should we do? Why? What will it take? What could go wrong?
- The common trap is solving the stated problem without testing whether it is the real problem.
Big Picture: Consulting Is a Decision-Making System
Contracting, diagnosis and recommendation are not three separate “soft skills.” They are a single consulting operating system. Contracting creates the right question, diagnosis finds the real cause, and recommendation converts insight into action.
The Core Skill: Do Not Accept the First Version of the Problem
Clients usually present symptoms, not problems. “Sales are down,” “attrition is high,” “costs are rising,” and “teams are unproductive” are starting points. A consultant’s job is to convert that noise into a decision-ready problem statement.
The shift is from being a task taker to being a trusted advisor. A task taker asks, “What analysis do you want?” A trusted advisor asks, “What decision are you trying to make, and what evidence would change your mind?”
Contracting: The Conversation Before the Project Starts
Contracting is the explicit agreement on the work: goal, scope, roles, access, timeline, decision rights, success measures and ways of working.
Think of contracting as a small but high-leverage negotiation. You are not negotiating fees in a campus case. You are negotiating clarity. A weak contract creates rework, politics and vague recommendations. A strong contract tells you exactly what problem you are solving and what “good” looks like.
If a CHRO says, “Improve workforce productivity,” a weak consultant starts benchmarking headcount. A strong consultant contracts the question: productivity for which workforce group, measured by what output, over what time horizon, and with which constraints such as hiring freeze, automation budget or labour rules? The strategic so what: good contracting prevents you from solving a politically attractive but economically irrelevant problem.
Diagnosis: Turn Symptoms into Causes
Diagnosis is structured root-cause discovery: forming hypotheses, breaking the problem into MECE drivers, testing evidence and identifying the few causes that matter most.
A good diagnosis has two qualities. First, it is structured: the issue tree covers the problem without overlap. Second, it is evidence-seeking: every hypothesis must be tested through data, interviews, observation or benchmarking.
Use this five-step diagnostic loop when the problem is ambiguous:
Recommendation: Convert Insight into Commitment
Recommendation is a specific course of action supported by logic, evidence, trade-offs, risk mitigation and implementation ownership.
The most useful recommendation is not “Here are three options.” It is “Choose Option B because it best meets the decision criteria; here is the evidence, risk, cost, sequencing and owner.”
How to Pressure-Test a Recommendation
Before presenting, test whether your recommendation is decision-ready. These measures are practical consulting checks, not academic metrics.
Definitions You Can Say in One Breath
- Contracting: Explicit agreement on goals, scope, roles, access, success measures, timelines and decision rights.
- Diagnosis: Structured root-cause discovery using hypotheses, evidence and prioritisation.
- Recommendation: A specific action choice backed by logic, evidence, trade-offs, risks and implementation plan.
- MECE: Mutually exclusive, collectively exhaustive - no overlap, no gaps in a problem breakdown.
- Hypothesis: A testable explanation of what may be causing the observed business problem.
Case Study: Meesho and the Shift from Social Commerce to Marketplace Discipline
Meesho shows how a company can re-contract the problem, diagnose a changing market and recommend a sharper operating model for value-conscious Indian shoppers.
Situation: Meesho was widely known for social commerce, where resellers helped sell products through personal networks. As Indian online shopping matured, many users began buying more directly, especially in value fashion, home and unbranded categories. The visible symptom could have been framed narrowly as “increase reseller productivity.”
The better contract: The sharper question was not just how to improve resellers. It was: how should Meesho serve price-sensitive Indian customers and small sellers at scale when direct discovery, low prices and logistics reliability matter more?
Diagnosis: The primary driver was a shift in user behaviour toward direct, app-based value shopping. Supporting drivers included a large base of small Indian sellers seeking digital demand, high price sensitivity beyond metro markets, and the need for a cost-conscious operating model.
The move: Meesho leaned into a marketplace model with a strong value proposition for small sellers and budget-conscious customers. Publicly, the company has emphasised low-cost selling, broad seller participation and a focus on affordability. The recommendation logic was clear: compete not by copying premium e-commerce playbooks, but by designing for Bharat-style value commerce.
Outcome and lesson: Meesho became one of India’s most visible value e-commerce platforms. The consulting lesson is powerful: if the team had solved only the stated symptom, it might have optimised the old model. By re-contracting and diagnosing the real market shift, the strategic move became clearer.
How AI Changes Contracting, Diagnosis and Recommendation
AI does not replace consulting judgment, but it changes the speed and quality of the first draft. In 2026, the best students will use AI to think better, not to outsource thinking.
- Contracting becomes faster but riskier: AI can convert meeting notes into a draft project charter, scope, assumptions and open questions. The risk is hallucinated agreement, so the consultant must confirm every decision right and success measure with the client.
- Diagnosis becomes evidence-rich: Tools can summarise interview transcripts, cluster pain points, compare themes across stakeholders and scan annual reports or policy documents. The human skill is still hypothesis design and deciding what evidence matters.
- Recommendation becomes more scenario-based: AI can draft option trees, risk registers, implementation plans and sensitivity questions. The consultant must choose, simplify and defend the recommendation.
Use NotebookLM for practice: upload a company annual report, two news articles and your case notes; ask it to generate a project charter, top five diagnostic hypotheses, evidence needed for each hypothesis, and three likely interviewer challenges. Then rewrite the final recommendation in your own words.
Interview Relevance
“A client says employee productivity has fallen across business units. How would you structure the consulting engagement from contracting to recommendation?”
In your answer, keep saying “decision.” Interviewers trust candidates who show they know consulting is not analysis for its own sake; it is analysis that helps a client choose.
Common Mistake
The biggest mistake is jumping from symptom to solution: “productivity is low, so automate” or “attrition is high, so increase pay.” It costs candidates because it shows weak diagnosis and poor client management. One-line fix: contract the decision, diagnose the drivers, then recommend the action.