Skip to content
Ashish Kumar AgnihotriOperating Partner

01Industry

Media & newsrooms

Newsrooms and content operations live or die on throughput and quality at speed. The constraint is almost never the size of the team — it is how work moves through it. Fix the operating model and output multiplies; leave it unmeasured and adding people just adds cost and confusion.

A newsroom is a production line that refuses to admit it is one. Copy moves from assignment to research to draft to edit to publish, and at every handoff it can stall, get re-keyed, or sit unworked while someone is busy elsewhere. The newsroom feels frantic, yet most of a story’s life is spent waiting, not being written. That is the paradox of high-velocity content: the people are flat out and the output is still capped, because the cap is set by the flow between desks rather than by the effort inside them. You cannot fix a flow problem by asking tired people to try harder.

The instinct, when output plateaus, is to hire. More writers, another desk, a bigger night shift. But adding people to a pipeline you cannot see usually lowers output per head — the new hires inherit the same invisible queues, and now there are more handoffs to lose work in. Throughput is a property of how the desks are arranged and how work passes between them, not of headcount. Until the pipeline is mapped and the waiting is measured, every expansion is a bet placed blind, and the run-rate barely moves while the wage bill climbs.

This is the work behind a newsroom whose throughput roughly quadrupled to around 400 stories a day — and where the account then expanded by two news desks because the model could finally absorb them. Nothing about it relied on working journalists faster. It came from redrawing the production pipeline so stories stopped queueing, building quality into the flow rather than bolting it on at the end, and giving editors a live view of where work actually was. Speed and standards stopped being a trade-off. The newsroom got both because the operating model finally supported both.

What tends to break

  • Stories wait in invisible queues between desks rather than being worked.
  • Quality is checked too late to do anything but reject finished work.
  • Leadership has no real-time view of throughput or where it’s stalling.
  • Scaling means hiring — which adds cost without lifting the run-rate.

How I help

  • Map the true production pipeline and expose where work waits.
  • Redesign desks and handoffs so ownership is clear and quality is built into the flow.
  • Install a daily cadence and the few throughput/quality metrics editors steer by.
  • Scale deliberately — adding capacity only once the model can absorb it.

Sound familiar?

01

You’re producing at a fraction of your potential and can’t see why.

02

More people hasn’t meant more output.

03

Errors are caught late, or by the audience.

Proof in this sector

4×

Quadrupling newsroom throughput to 400 stories a day

Read the case

The fit

Process Transformation & Lean Six Sigma

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

Book a consultationExplore the service

In depth

The operating detail for this sector.

Throughput is a flow problem, not an effort problem

The first thing I do in a newsroom is time the work, not watch the people. Where does a story actually spend its hours — being written, or waiting in someone’s queue for attention? In almost every content operation the answer is waiting, and by a wide margin. A piece is drafted in an hour and then sits for six before an editor frees up; it clears the edit and waits again for a slot on the desk that handles media or publishing. The newsroom looks busy because everyone is occupied, but occupancy is not output. Once you map the flow and put numbers on the dead time, the lever becomes obvious: you lift throughput by attacking the queues between desks, not by demanding more speed inside them.

Build quality into the flow, not at the end

Most newsrooms inspect quality at the worst possible moment — after the story is finished, when the only options are publish it or send it all the way back. A late rejection is pure waste: the work is done, the time is spent, and now it is done again. The fix is to move the check upstream, into the flow, so a story is shaped correctly while it is cheap to shape. Clear standards at the point of assignment, a quick structural check before the deep edit, defined ownership of what each desk guarantees before it passes work on. Built in this way, quality stops fighting speed. The errors that used to surface at the end — or worse, after publication, in front of the audience — are caught when fixing them costs minutes rather than a full rewrite.

The daily cadence editors actually steer by

A newsroom needs a rhythm, but not more meetings. The cadence I install is a short, sharp daily look at the few numbers that govern the day: how many stories are in each stage, where the backlog is building, what is at risk of missing its window. It replaces the gut-feel question "are we behind?" with a visible answer. Editors stop discovering a bottleneck at 6pm when the evening desk is drowning, because they saw it forming at 11am and moved a person. The point of the cadence is not control for its own sake — it is to make the state of the pipeline legible early enough to do something about it, every single day, without a status meeting eating the time the newsroom needs to publish.

Why hiring rarely lifts the run-rate

When output stalls, "we need more writers" is the easy diagnosis and usually the wrong one. If the constraint is the edit desk or the publishing step, adding writers just lengthens the queue in front of the real bottleneck — more drafts piling up where the jam already is. Throughput is set by the tightest stage in the pipeline, and capacity added anywhere else is capacity wasted. I find that constraint first, relieve it, and only then talk about headcount. Often the run-rate climbs sharply before a single hire is made, simply by rebalancing the desks around where the work actually backs up. When expansion does come, it lands on a model that can absorb it — which is exactly how a newsroom can take on additional desks and have them add output rather than confusion.

Measuring a newsroom without killing the craft

Journalists are rightly wary of metrics — quotas tend to reward volume over judgement and hollow out the thing that makes the newsroom worth reading. So the measurement I build is about the flow, not the individual. I track where stories wait, how long stages take, where rework originates — operational signals the editors own, not a leaderboard that pits writers against each other. The standard for the work itself stays editorial, held by editors, expressed as what good looks like rather than how many. Done this way, measurement protects the craft instead of corroding it: it removes the friction and the waiting that frustrate good journalists, and it leaves the editorial judgement exactly where it belongs.

When a newsroom redesign is the wrong call

Not every content operation needs this. If the work is genuinely low-volume or bespoke — long-form investigations, a handful of considered pieces a week — there is little flow to optimise and the gains are small; the constraint really is craft and time, and a pipeline redesign would be solving a problem you do not have. The same is true if the bottleneck is upstream of production entirely: a commissioning strategy that has not been decided, or a demand for output the audience does not actually want. I would rather say so than re-engineer a pipeline that is not the thing holding you back. This work earns its keep where volume is high, the cadence is relentless, and throughput is visibly capped by how the work moves.

Questions

Common questions.

By fixing the flow rather than adding headcount. In most newsrooms a story spends far more time waiting between desks than being written, so output is capped by handoffs and queues, not by effort. Map where work actually waits, relieve the tightest stage in the pipeline, and rebalance the desks around it. The run-rate usually climbs before a single hire is made. Adding people to an unseen pipeline tends to lower output per head, because the new hands inherit the same queues and create more handoffs to lose work in.

Only if quality is left as a final inspection. The reason speed and standards feel like a trade-off is that most newsrooms check quality at the end, when the only choices are publish or reject. Move the check upstream — clear standards at assignment, a structural check before the deep edit — and a story is shaped correctly while it is still cheap to shape. Built into the flow this way, quality stops fighting speed. The newsroom I worked with quadrupled throughput while errors fell, because fewer mistakes reached the end of the line to begin with.

It means redrawing how a story travels from assignment to publication so it stops queueing. In practice: timing each stage to expose where work waits, clarifying what each desk owns and guarantees before it passes work on, removing handoffs that add delay without adding value, and building quality checks into the flow rather than bolting them on at the end. The aim is a pipeline where work moves continuously instead of sitting in invisible queues. It is the same operating discipline used on any production line, applied to editorial work without compromising the craft.

I measure the flow, not the individual. The signals that matter operationally are where stories wait, how long each stage takes, and where rework originates — and those belong to the editors, not to a leaderboard ranking writers by volume. Quotas reward output over judgement and tend to hollow out the work. The standard for the journalism itself stays editorial, held by editors and expressed as what good looks like. Done this way, measurement removes the friction and waiting that frustrate good journalists, and leaves editorial judgement exactly where it should sit.

The diagnosis is quick — a couple of weeks of timing the real flow usually makes the main bottleneck obvious, and relieving it can lift the run-rate noticeably soon after. The deeper gains, where quality is built into the flow and the daily cadence is holding on its own, build over the following weeks as the new model beds in. The sequence is deliberate: relieve the tightest constraint first for an early, visible win, then install the structure that makes the gain permanent rather than a temporary push that fades.

It depends on volume and cadence, not headcount. A small newsroom publishing at high velocity has the same flow problem as a large one and benefits just as much. But if the operation is genuinely low-volume or bespoke — a few considered, long-form pieces a week — there is little flow to optimise, and the real constraint is craft and time rather than how work moves. I would tell you so. This work earns its keep wherever the cadence is relentless and throughput is visibly capped by the pipeline, regardless of the size of the team.

No — they gain control. Today many editors discover a bottleneck only when the evening desk is drowning. A live view of the pipeline lets them see the backlog forming hours earlier and move a person before it becomes a crisis. The cadence makes the state of the work legible; it does not take decisions away from the people who run the desk. Editorial judgement stays with editors. What changes is that they are steering with the day visible in front of them rather than reacting to it after the fact.

A tool automates whatever process you already have — including the queues and the late checks. If you install a workflow system on top of a pipeline that was never re-engineered, you tend to cement the waiting in place at higher cost. The order matters: fix how the work flows first, prove the simpler version works, then let tooling support the process that genuinely needs to exist. This is a design problem before it is a software problem, and most of the throughput gain comes from the redesign, not the platform.

Other sectors

Advertising & agenciesBPO, GBS & shared servicesScaling & founder-led companiesMulti-entity & enterprise operations