The Lament of the KPO Middle Manager

& Why Clients Stop Pushing for Better

Insight article

This article assumes familiarity with offshore analyst models and examines a structural communication failure — the intermediation layer between client and analyst.

1. The pattern everyone recognises

A credit analyst in London spots an error in a quarterly update. The counterparty's leverage ratio looks wrong — the debt figure hasn't been adjusted for a recent refinancing. It's not a catastrophic mistake, but it's the kind of thing that erodes confidence in the offshore team's output over time.

The analyst raises it with the delivery manager — the person on the weekly call, the familiar face, the relay point between client and offshore team. The delivery manager didn't produce the report. They may have done a four-eyes check on it, but they didn't build the model, didn't pull the data, and don't carry the full context of this particular name. They are an intermediary, not an author.

What happens next is predictable. The delivery manager acknowledges the issue. They promise it will be corrected. They may offer an explanation that partially addresses the concern but doesn't fully resolve it — because they cannot fully resolve it. They didn't do the work.

At the other end, the analyst who actually produced the report receives a half-understood version of the complaint. The delivery manager has translated it — inevitably losing nuance, stripping context, softening the edges. The offshore analyst wants to discuss this directly with the client team. In most cases, they can't. They don't want to undermine the delivery manager, they may not be permitted to, and the structure doesn't allow it. So they make their best guess at what the client meant and revise accordingly.

The output comes back. Slightly improved. Still not quite right. The London analyst considers raising it again but weighs the effort against the likely outcome: another cycle through the same relay, another approximation, another partial fix. They decide to correct it themselves.

After two or three iterations of this, the client stops pushing for better. They accept adequate. They quietly absorb the rework. That is the real cost — not a communication failure in the conventional sense, but a progressive lowering of standards driven by a structure that makes excellence too expensive to pursue.

2. Why the delivery manager mollifies

It is important to be clear: this is not a failing of the individual. The delivery manager in a typical offshore research model is often the only person with any real experience in a team of relatively junior analysts. They may have been a decent analyst themselves before moving into a coordination role. But their function has shifted. They are no longer just producing work — they are managing the client relationship, doing quality assurance on output they didn't build, and supervising analysts who arrived with limited training and are still learning the client's requirements. They are stretched across all of these roles simultaneously, and none of them gets the attention it needs.

Their skill, by necessity of structure, becomes mollification. When a client raises an issue, the delivery manager's instinct is to manage upwards: smooth the concern, acknowledge it without escalating, translate it into something the team can act on without disruption. Clients can tell when they are being mollified rather than having their problem solved. Everyone can. But what else can the delivery manager do? They are the interface between a client who wants authentic, correct analysis and a team that doesn't have direct access to the person asking for it.

Managing downwards is no easier. The delivery manager relays a complaint that they only half understand, because they didn't do the work. The instruction arrives at the analyst's desk garbled — stripped of the context that would make the fix obvious. The analyst, who may be perfectly capable of resolving the issue if they could speak to the client directly, is instead working from a second-hand description of a problem they can see more clearly than the person relaying it.

The delivery manager persists in this role because it serves the provider's operating model. They are the continuity layer. In a high-turnover environment, the delivery manager is the stable relationship — the person the client knows, the person who joins the call, the person who keeps the engagement running. Their value to the provider is relationship management, not analysis. And for the provider, that is rational. But for the client, it means that the one person they speak to regularly is structurally the least equipped to solve their analytical problems.

3. The equilibrium that nobody designed

This pattern is not the result of bad intentions. No provider set out to build a mollification machine. It emerges from the interaction of reasonable decisions under structural constraints.

Providers scale by adding junior analysts supervised by a small number of experienced people. Direct client access for every analyst is operationally risky — not primarily because the analyst might say something wrong, but because they might not be there next week. In a high-turnover environment, building a direct relationship between client and analyst is a liability for the provider: every departure breaks a connection the client has come to rely on. The delivery manager layer exists partly to insulate the client from this churn. It is a sensible response to a real problem.

But the unintended consequence is that it suppresses exactly the behaviour that would make the offshore team genuinely valuable. Analysts who could communicate directly, take ownership of their output, push back when an instruction doesn't make sense, and engage authentically with the client's analytical needs are prevented from doing so by a structure designed to contain risk.

The equilibrium has lasted because the output is good enough. Reports arrive on time. Models are populated. Quarterly updates are produced. The surface-level deliverables meet the specification. For years, this was sufficient — the cost savings justified the overhead of the intermediation layer, and the friction was tolerable.

4. Why good enough is no longer enough

The economics of good enough have changed. In practice, output that is adequate but context-thin — reports that are technically populated but lack genuine analytical depth, models that run correctly but don't reflect real understanding of the name — is now what AI produces. Faster. Cheaper. Without a weekly call.

Large language models can generate first-pass credit summaries, populate templates, pull data into standardised formats, and produce surveillance reports that meet a basic specification. They do not understand the counterparty. They do not carry institutional memory. They cannot exercise judgement about whether a covenant breach is technical or substantive. But they produce adequate output at marginal cost and near-instant speed.

An offshore team whose output sits at the same level as AI — technically correct, contextually thin, produced without genuine engagement with the analytical question — is now competing with AI on AI's terms. And it is losing. It is slower, more expensive, requires management overhead, and introduces the communication friction described above. The cost case, which originally justified offshoring, now undermines it: why pay multiples of the AI cost for output that is no better?

The only justification for human analysts — offshore or onshore — is the work that AI cannot do. Judgement. Direct communication with the people who use the analysis. Ownership of the output, meaning genuine accountability for whether it is right, not just whether it is delivered. The ability to say "I think this assumption is wrong" or "this name has deteriorated and here's why" without waiting for permission to have that conversation.

The intermediation model structurally suppresses all three of these capabilities. It prevents direct communication. It diffuses ownership through the relay chain. And it removes the conditions under which judgement develops — because judgement requires engagement with the question, not just execution of the task.

5. What breaks the pattern

The solution is not a process improvement. It is not better templates, more frequent calls, or a revised escalation protocol. These treat symptoms of a structural problem.

What breaks the pattern is a different kind of analyst — and a model that lets them operate.

Analysts who communicate directly with the onshore team. Not through a relay. Not through a delivery manager who translates and softens. Direct access means the analyst who produced the work is the person who discusses it, defends it, and revises it. Feedback is unmediated. Context is preserved. Issues resolve in one cycle, not three.

Analysts who stand up for themselves. This means people who are trained to communicate clearly, who have the personality to take ownership, and who accept that ownership involves risk. They can be in trouble if it goes wrong. They reap the reward if it goes right. That is not an offshore-specific requirement — it is the baseline for being an analyst of any standing, anywhere in the world. The difference is that most offshore structures actively select against this trait, because the delivery manager layer exists to insulate the client from the analyst's rough edges. A model that values direct communication selects for it.

Training that develops communication as a core skill, not an afterthought. Analytical training in most offshore models focuses on technical competence — how to build a model, how to spread financials, how to populate a template. Communication is assumed to follow. It does not. The ability to explain your analysis to a sceptical senior banker, to flag a concern without causing alarm, to ask the right clarifying question — these are skills that require deliberate development, typically under the supervision of someone who has done it themselves at a senior level.

The structural requirement is clear: remove the intermediation layer, train analysts to communicate and take ownership, and select for the personality types who thrive under that model. What emerges is not just a communication improvement — it is a fundamentally different kind of offshore engagement, one where human value is preserved because human capabilities are actually deployed.

6. How to tell whether this is happening to your team

You are being mollified, not served

When you raise an issue, the response is acknowledgement and reassurance rather than a substantive analytical discussion. You hear "we'll look into it" more often than "here's what happened and here's why."

You never speak to the person who did the work

Your primary contact is the delivery manager. The analysts who produced the report, built the model, or wrote the summary are not on the call, not in the email chain, and not available for direct questions.

Issues take multiple cycles to resolve

A problem that should be resolved in one direct conversation takes two or three relay cycles — complaint to delivery manager, delivery manager to analyst, revised output back to you, still not quite right, repeat.

You have stopped asking for improvements you know are needed

You have mentally categorised certain quality gaps as "not worth raising" because the correction cycle is too slow and too imprecise. You absorb the rework instead.

Output is adequate but never surprising

The work meets specification but never exceeds it. You never receive an unprompted insight, a flagged concern, or an observation that makes you think differently about a name. The team executes but does not engage.

You find yourself comparing the output to what AI could produce

If the thought has crossed your mind that a language model could produce similar quality faster and cheaper, the team's value proposition has already been structurally undermined — not by AI, but by the intermediation layer that constrains the team to AI-level work.

The question of who owns the analytical process — who reviews, who escalates, and who is accountable for whether the output is right — determines whether an offshore engagement compounds in value or slowly degrades. Where ownership is clear and communication is direct, the model improves with time. Where ownership is diffused through an intermediation layer, even well-intentioned structures produce the pattern described above.

This article is part of a series on structural dynamics in offshore research models.

For a broader view of how India-based analyst teams are structured and why integration determines outcomes, see: India-Based Analyst Teams.

For related analysis of how turnover compounds the communication problem, see: The Real Cost of 2.2-Year Analyst Turnover in KPO.

If your current offshore model exhibits these patterns and you're evaluating alternatives, see: Upgrade to Frontline.