Skip to content
Ashish Kumar AgnihotriOperating Partner

Judging consultants

Best Operations Consultant in India: How to Judge Before You Engage

Most of what ranks for “best operations consultant in India” is a directory: self-submitted profiles, paid placement, badges nobody audits. This page takes the opposite approach. It teaches you to judge an operations consultant before you engage one — the consultant-versus-operator distinction, the deliverable test, the evidence standard, the red flags — and then puts my own record under the same light.

The “best operations consultant” search is owned by aggregators because aggregation is a business model: collect profiles, sell placement, rank by participation. Some listed firms are excellent; the list itself still tells you almost nothing, because no directory verifies whether a consultant’s recommendations survived contact with a live operation. Judgement does not aggregate. What you need instead is transferable criteria — what to demand as evidence, what the engagement should physically deliver, and which signals disqualify — so you can evaluate anyone: a global firm, a boutique, an independent, or the operator whose site you are reading. That is what this page provides, in the order you will use it.

Before comparing candidates, settle a prior question: do you need a consultant or an operator? A consultant studies your operation and hands you recommendations — valuable when you need analysis, a second opinion, or a decision de-risked. An operator takes a seat inside the operation and installs the fix — cadence, scorecard, process ownership, quality governance — then transfers it to your team. Both are legitimate. Confusing them is expensive in both directions: an operator is more than a bounded study needs, and a consultant’s deck will not fix a throughput problem that needs someone’s hands on the system for two quarters. “Best” starts with which product your problem actually requires.

My interest, declared: I work as an operator — a fractional COO and Business Excellence advisor — not as a report-writing consultant, and this page naturally reflects that view of where value sits. The criteria, though, are the ones I would use in your chair whichever way you decide: evidence with scope, deliverables you can hold, measurement agreed up front, engagements designed to end. At the close I state my own record — nineteen years, Publicis Groupe, four numbers — as one worked example of the standard, not as a ranking. If a bounded study is what you need, a good consultant will serve you better than I would, and this page will help you find that person too.

In depth

What you need to know.

Why directories cannot tell you who is best

Understand the mechanics and the badges lose their glow. Consultant directories and “top firms in India” aggregators are compiled from self-submitted profiles, sponsored placements and reviews the platform cannot verify; the ranking logic rewards participation and payment, not outcomes. No aggregator calls a listed consultant’s clients two years later to ask whether the recommendations were implemented, whether the numbers moved, or whether anything still runs. That is not cynicism; it is their unit economics — verification does not scale, listings do. Use directories for what they are: a discovery layer, a way to assemble a longlist quickly. Then do the judging yourself, with criteria that do not appear in any profile field: scope-attached numbers, inspectable artefacts, and references questioned about what actually held.

Consultant or operator: decide which you are buying

The distinction determines everything downstream — fee structure, timeline, and what you hold at the end. A consultant’s product is analysis: a diagnostic study, a process map, a recommendations deck, perhaps a pilot design. It suits bounded questions — should we consolidate these entities, why is margin leaking, is this operation ready for the growth we plan. An operator’s product is a working system: the weekly cadence running, the scorecard live, process owners named and trained, quality governance installed — built inside your business, then handed over. It suits continuing problems — throughput, quality, cycle time, decision latency — that a document cannot fix. Ask yourself one question: when this engagement ends, do I need to know something, or do I need something to be running? Buy accordingly.

The deliverable test: recommendation or installed system

Ask every candidate the same blunt question: “What do I physically hold when you leave?” Then listen for nouns. A consultant’s honest answer is a document set — findings, quantified opportunities, a sequenced roadmap, a business case. Judge it by asking to see a sanitised past example: real recommendation decks name systems, owners and numbers; weak ones re-describe your problem in frameworks. An operator’s honest answer is an artefact set — an operating cadence with an agenda that closes decisions, a scorecard on one source of truth, documented playbooks, trained process owners — plus the transfer plan that makes them yours. Vague answers to this question are the most reliable early filter in the market: anyone who has genuinely delivered either product can describe it in concrete parts, immediately.

Evidence that survives scrutiny

Whichever product you are buying, the evidence standard is identical. Every claimed result should carry three attachments: a baseline and endpoint defined the same way — “improved efficiency by 30%” means nothing until you know what was measured and how; a scope — how many people, entities, clients or transactions the number spans, because scale is what makes a result hard; and a fate — what happened after the engagement ended, which separates installed change from performed change. Then triangulate: artefacts you can inspect, references you can question, roles you can verify publicly. Consultants and operators with real records produce all of this readily, often with visible relief that someone finally asked. The ones without it will offer you client logos and adjectives. Logos are not numbers.

The scale question most consultants cannot answer

One question sorts the Indian operations field faster than any other: “What have you personally run — not studied, not advised, run?” Operating judgement is forged by owning consequences: the quality score with your name on it, the SLA breach at two in the morning, the month-end close across dozens of entities, the client escalation that ends in a makegood negotiation. A consultant can be brilliant without that history — but for operating mandates the difference shows within weeks, because recommendations written by someone who has absorbed consequences are buildable, and the other kind are merely elegant. When I answer that question, the specifics are nineteen years inside operations, most recently running Business Excellence at Publicis Groupe across 500+ clients and teams of 2,000+. Demand equivalents from everyone you shortlist.

Red flags before you sign

Some patterns should stop the process regardless of pedigree. Frameworks as the headline: Lean, Six Sigma and operating-model vocabulary offered where numbers should be — methodology is a toolset, not a track record. Results without scope, which are unverifiable by design. No measurement plan: a serious proposal names the metrics it expects to move and how they will be counted; “we will define KPIs in phase two” means the accountability is being scheduled for later. Team bait-and-switch: the partner sells, juniors deliver — ask who exactly does the work, in writing. Open-ended time-and-materials billing on an operating problem, which prices delay attractively. And no designed ending: an engagement without an exit and a transfer plan is structured to renew, not to finish.

How the best structure operations engagements in India

Good structure shows up before the contract does. For advisory work: a defined study with a fixed fee, a stated evidence base — time on your floor, not only in your data room — named metrics, and a deliverable specified in advance. For operating work: a short paid diagnostic first, so both sides commit on findings rather than chemistry; then a retainer scaled to cadence — days per week, decision rights, what is owned versus advised — with measures agreed before it starts and reviewed weekly. In India, expect proper GST invoicing and clarity about who is actually staffed on the work. In both models the best practitioners price the assessment separately, welcome verification, and design the ending — because their next client comes from this one’s numbers, not from this one’s renewal.

Where I sit against these criteria

Stated as evidence for the test, not as a ranking. I work as an operator: nineteen years in operations, most recently Senior Director, Business Excellence at Publicis Groupe — quality and delivery across 500+ clients, teams of 2,000+, and more than USD 750 million in annual media spend. The numbers, with scope: quality raised from 95% to 99% across 2,000+ campaigns for 450 clients; a billing cycle compressed from roughly two months to fifteen days across 75 entities and 2,000+ employees; Republic World’s newsroom scaled to about 400 stories a day — four times prior output — while adding two desks; USD 20M+ in makegoods protected at a global media network. Engagements run diagnostic-first, retainer-based, never hourly, and end with your team holding the system. Apply the deliverable test to me.

Questions

Common questions.

The one whose evidence survives scrutiny — no directory, badge or list can answer this for you, because none of them audits outcomes. Run the test this page describes: decide first whether you need analysis or an installed system; then demand numbers with baseline and scope, inspectable artefacts, and references who can say what still runs. As one worked example of the standard: my record is nineteen years in operations, Business Excellence leadership at Publicis Groupe across 500+ clients, and results like 95% to 99% quality across 2,000+ campaigns. Judge me with exactly the same test.

The seat. An operations consultant works from outside: studies the operation, quantifies the problems, recommends the fix, and leaves you to implement. A fractional COO takes a part-time seat inside the leadership team: owns processes, runs the operating cadence, installs the systems, and is accountable for the numbers moving — then transfers the model and leaves. Consultants suit bounded questions; operators suit continuing problems. Fees differ accordingly: project fees versus monthly retainers. If your issue will still exist next quarter unless someone owns it weekly, you are shopping for an operator, whatever title the market uses.

Attach three things to every claim: definition, scope, fate. Definition — what exactly was measured, and identically at both ends? Scope — across how many people, entities, clients or transactions? Fate — what remained twelve months after the engagement? Then triangulate: ask for sanitised artefacts (a real recommendations deck or an installed scorecard), question references about what specifically changed and what still runs, and check stated roles against the public record. This takes an afternoon and filters most of the market. Practitioners with genuine records tend to enjoy the exercise; that reaction is itself a data point.

Treat them as discovery, never as judgement. Most are compiled from self-submitted profiles and paid or sponsored placement, and none that I am aware of verifies outcomes with clients after the fact — the economics of aggregation do not fund verification. That does not make every listed firm weak; strong practices appear in weak lists. It means the badge carries no evidentiary weight. Use directories to assemble a longlist quickly, then apply your own test: deliverable clarity, scope-attached numbers, artefacts, references. Ten minutes of that outweighs any award logo in a website footer.

It depends on which product you bought, and you should know before signing. An advisory engagement should deliver findings with quantified impact, recommendations specific enough to act on without hiring the author again, a sequenced roadmap, and the evidence base behind all of it. An operating engagement should deliver a running system: weekly cadence, live scorecard on one source of truth, documented playbooks, named and trained process owners — plus a transfer that survives the consultant’s departure. In both cases, insist the deliverable is specified in the proposal. “Phase one: discovery” is not a deliverable; it is a meter starting.

India publishes few benchmarks, so judge structure before size. Bounded advisory work should carry a fixed project fee tied to a defined deliverable. Embedded operating work should run on a monthly retainer scaled to days per week and scope. Be cautious with open-ended hourly or time-and-materials billing on operating problems — it prices delay attractively. Anchor the fee against the cost of the problem: a stuck billing cycle, a quality score bleeding clients, margin leaking through rework. Senior judgement is never the cheapest line item; it is usually the cheapest fix. My fractional COO cost guide covers the retainer arithmetic in detail.

When the problem needs an owner, not an opinion. If throughput, quality, cycle time or decision latency will still be broken next quarter unless someone runs the fix week by week, a recommendations deck — however good — lands on the desk of the same team that could not fix it while doing their day jobs. That is an operator’s problem: someone who takes the seat, installs the cadence and scorecard, and transfers a working system. The reverse also holds: bounded questions deserve bounded studies, and hiring an operator for one means paying for a seat you do not need.

Six earn their place in every conversation. What do I physically hold when you leave — nouns, not phases? What have you personally run, at what scale, and what broke while you ran it? Show me a sanitised past deliverable. Which of my numbers will this engagement move, and how will we count them? Who exactly does the work — names, not the firm? How does this end, and what transfers to my team? The pattern in the answers — specific versus fluent — will decide most engagements before references are ever called. My annotated question guide goes deeper.

The next step

A short conversation settles most of this — and a fixed-fee diagnostic settles the rest.

Book a consultationThe diagnostic

Keep exploring

Business excellence & quality auditsThe installed-system version of operations consulting.Case studies with measured outcomesThe evidence standard, applied to my own work.Best business excellence consultantThe same test, applied to the BE discipline.Questions to ask a fractional COOThe annotated interview list.What is a fractional COO?The operator model explained.Hire a fractional COOIf your problem needs an owner, not a report.