Skip to content
Ashish Kumar AgnihotriOperating Partner

02Service

Business Excellence & Quality Audits

An independent assessment of operational quality, scored against a standard rather than an opinion.

Business Excellence is the practice of making quality a property of the system rather than of the people who happen to be careful that day. A quality audit is how you find out where you actually stand. Together they answer the question most operations cannot answer with evidence: is our work good, consistently, and will it stay good as we grow?

I have spent most of my career building and running this function at scale — most recently as Senior Director, Business Excellence at Publicis Groupe, where quality was scored across more than 2,000 campaigns and 450 clients, and where a programme lifted that score from roughly 95% to 99%. That is the lens I bring: quality measured against a standard, defects traced to the part of the operating model that allowed them, and remediation aimed where the cost actually is.

An audit here is not a clipboard exercise. It produces a board-ready read you can defend to clients, a baseline you can track quarter on quarter, and — if you want it — the controls that hold the gain rather than letting the score drift back the moment attention moves on.

01The problem

Most teams cannot answer a simple question with evidence: how good is our work, consistently, at scale? Quality is asserted, not measured. Defects are caught by clients before they are caught internally. When something goes wrong, the review is a search for a person to blame rather than the flaw in the system that allowed it. Without a standard, "quality" is just confidence — and confidence does not survive scale.

02Signs you need this

When this is the right call.

  • 01

    Clients catch defects before your own team does

  • 02

    Quality is described in adjectives, not measured as a number

  • 03

    Post-mortems end in blame rather than a change to the system

  • 04

    Quality scores swing with whoever is paying attention that month

  • 05

    You cannot show a client or board a defensible quality baseline

  • 06

    Remediation effort is spread thin instead of aimed at what costs the most

03The method

How the work goes.

  1. 01

    Define the standard

    Before anything is audited, we agree what good looks like — in terms specific to your delivery, your clients and your risk. A standard you can score against, not a generic checklist.

  2. 02

    Audit against evidence

    A structured review of live work and the processes behind it: sampling, scoring, and tracing defects back to their root in the operating model rather than to individuals.

  3. 03

    Quantify the gap

    A quality score that is honest and comparable over time, with the specific failure modes ranked by frequency and cost — so remediation goes where it actually matters.

  4. 04

    Close the loop

    The controls, gates and measurement that hold the gain: the difference between a one-off clean-up and a quality function that keeps the score where you moved it to.

04In depth

What this work really involves.

Why a quality score beats a quality opinion

Most leadership teams describe their quality in adjectives — strong, improving, a concern. Adjectives do not survive scale, and they cannot be managed. A score can. When quality is expressed as a number against an agreed standard, you can see it move, set a target, and tell whether last quarter’s effort actually worked. You can also compare across teams, sites and clients without it turning into an argument about whose work is harder. The first job of an audit is to replace confidence with a measurement honest enough to act on.

Defects are a property of the system, not the person

When a defect reaches a client, the instinct is to find who let it through. That instinct is expensive: it teaches people to hide errors and it leaves the real cause — the point in the process where the defect was both likely and invisible — completely untouched. Every audit I run traces failures back to the operating model: the missing gate, the handoff with no owner, the metric nobody watched. Fix the system and the same class of defect stops recurring. Blame a person and it simply waits for the next one.

Weighting remediation by financial exposure

Not all defects cost the same. A cosmetic slip and an error that puts client billings at risk should not get the same attention, yet most quality programmes treat the backlog as a flat list. I rank failure modes by frequency and by financial exposure, so the limited remediation budget goes to the handful of issues that actually move money and risk. This is the same logic behind the makegoods QA framework I built at a global advertising network, which protected more than USD 20 million in client billings by placing the heaviest checks exactly where the exposure was greatest.

Making the gain hold

The hard part of quality is not the clean-up; it is keeping the score where you moved it to once the project energy fades. That requires controls built into the running of the work — gates that cannot be skipped, a small set of quality metrics on the operating review, and an owner accountable for the number. Without that, every audit becomes an annual rediscovery of the same problems. The close-the-loop stage is what turns a one-off assessment into a quality function.

Where quality actually breaks under scale

Quality rarely fails because people stop caring. It fails because the system that kept work good at one size was never rebuilt for the next. The check that one experienced reviewer could do across ten jobs a day does not survive a hundred. The tacit standard everyone knew in a team of twelve evaporates in a team of two hundred. Handoffs that were a conversation become a ticket, and the context is lost in the gap. An audit that understands scale looks specifically at these seams — the points where a control that depended on small numbers or shared memory quietly stopped working — because that is where the defects your clients are about to find are already accumulating.

Building a quality function, not a clean-up

A one-off audit can lift a score for a quarter. A quality function keeps it lifted for years. The difference is ownership and rhythm: someone accountable for the number, a standard that is maintained rather than rediscovered, gates embedded in the workflow, and a quality review that sits inside the operating cadence rather than off to one side. The most valuable outcome of an audit is therefore not the report; it is the small, durable function it justifies — the difference between a business that cleans up periodically and one that does not let the mess accumulate in the first place.

What the board sees

Boards and clients do not want adjectives about quality; they want a number they can trust and a trend they can read. A defensible quality baseline — scored against an agreed standard, comparable quarter on quarter — changes the conversation. It lets you show a prospective client evidence rather than assurance, demonstrate to the board that a remediation programme actually worked, and price risk honestly in commercial terms. In regulated or client-audited environments, it is also the difference between passing a scrutiny event calmly and scrambling. The audit produces exactly this artefact: a board-ready read that turns quality from a claim into evidence.

05What it looks like

What an engagement looks like

  • A defined audit window with a written, board-ready assessment
  • Scored against an agreed standard — repeatable quarter on quarter
  • Optional remediation support to install the controls that hold the gain
  • Suited to delivery operations, BPO/GBS, agencies and regulated functions

Outcomes

  • A measured quality baseline you can defend to clients and the board
  • Defects caught internally, before they reach the customer
  • Remediation effort aimed at the failure modes that cost the most
  • Reviews that fix systems instead of assigning blame
Discuss this engagementSee the results

Questions

Common questions.

Business Excellence is the discipline of making good outcomes a property of how the work is designed, rather than of who happens to be careful. It combines a defined standard, measurement against that standard, root-cause analysis of failures, and the controls that hold improvements in place. The goal is quality that survives growth and staff turnover.

A certification confirms you follow a documented process; it does not tell you whether the output is actually good or where it will break under scale. My audit scores the real work against a standard specific to your delivery and risk, ranks failure modes by cost, and recommends targeted remediation. The two are complementary — certification proves conformance, an audit proves quality.

Most audits run over a defined window of a few weeks, depending on the breadth of the operation and how much live work we sample. You receive a written, board-ready assessment at the end, with a quality score and a ranked list of failure modes. Remediation, if you want it, is scoped separately.

No. The audit works from live work and existing process, with sampling designed to be light on the people doing the job. The tone matters: this is a review of the system, not of individuals, and I run it that way so the team gives you honest evidence rather than a tidied-up version.

Yes. The close-the-loop stage installs the gates, metrics and ownership that hold the gain, and remediation support is available either as a standalone engagement or as part of a broader fractional COO arrangement. The audit is most valuable when it leads to a quality function, not a filed report.

For most operations, a thorough baseline audit followed by a lighter quarterly re-score works well — frequent enough to catch drift while it is small, not so frequent that it becomes a burden. The cadence should match how fast the operation changes: a business adding clients or sites rapidly benefits from tighter intervals, while a stable operation can stretch them. The point of repetition is the trend line. A single audit tells you where you stand today; a series tells you whether the system is improving, holding, or quietly slipping.

It covers both the output and the system that produces it. On the output side, we sample live work and score it against the agreed standard. On the system side, we examine the process, the controls, the handoffs and the points where quality depends on individual diligence rather than design. The result ties the two together: not just how many defects, but where in the operating model they originate and what they cost — so remediation is aimed at causes rather than symptoms.

We define it together, before anything is scored. A useful standard is specific to your delivery, your clients and your risk — what a defect actually is in your context, which failures are merely cosmetic and which put money or compliance at stake. This is deliberately not a generic checklist; a standard borrowed from another business measures the wrong things. Agreeing the standard first is what makes the resulting score meaningful and defensible rather than an outsider’s opinion.

Particularly well. High-volume, multi-client delivery environments are exactly where quality has to be a property of the system rather than of individuals, and where an honest, comparable score is most valuable commercially — clients increasingly audit their providers, and a defensible baseline is a competitive asset. Much of my experience is in precisely these environments, where the same work is performed at scale across many clients and consistency is the product.

Other ways I work

01

Fractional COO

An operating partner inside the business — not a deck about it.

03

Process Transformation & Lean Six Sigma

Rebuilding how the work flows — measured, not just reorganised.

04

Operations Governance & BI Systems

The measurement and decision rights that let leadership steer, not guess.