How Do Editors Decide What to Cover, and What Do Journalists Want From a Pitch?

What do journalists say decides whether a pitch lands?

Muck Rack's 2026 survey of journalists puts alignment with existing coverage at the top, cited by 70%. Nothing else comes close as a threshold condition: before a story is interesting, newsworthy or well-evidenced, it has to be the kind of thing this particular journalist writes about.

That ordering is rarely reflected in how pitching is done. A pitch optimised for the strength of the underlying story, sent widely, is optimising the second question while failing the first. A weaker story sent to someone who covers precisely that subject outperforms a stronger one sent to someone who does not, because the recipient of the second is not in a position to use it at all.

Below fit, the criteria are things a sender can actually provide. 58% want access to credible sources — a named person who will speak, on the record, with relevant standing. 40% value original data, meaning material that does not already exist in public.

The gap between those two numbers is instructive. Access is wanted half again as often as data, and access is the cheaper thing for most organisations to offer. A company with an executive willing to be interviewed on a subject they genuinely know has the more commonly requested asset.

One more asymmetry is worth drawing out of the ordering. Fit is a property of the match between sender and recipient, so it can be researched before anything is sent; sources and data are properties of the sender, so they can be built. Nothing on the list is a property of the pitch's prose, which is where most effort is spent.

How long should a pitch be?

Under 200 words, according to 69% of respondents. That is roughly the length of three paragraphs of the size you are reading now, and it has to contain the subject, why it fits this journalist, what is being offered, and how to get it.

The practical consequence is that most of what a pitch usually contains has to go. Company background, funding history, product description and adjectives all compete for the same two hundred words as the thing that would actually make the story usable.

It also implies that attachments and supporting material are not where a decision gets made. If a document is necessary to understand the offer, the offer has not been made yet.

The length preference also explains a common failure that looks like rejection but is closer to non-delivery. A pitch whose first sentence describes the sender rather than the story spends its opening on the one piece of information the recipient has least use for. Under two hundred words there is no recovery from that: the decision is usually made before the part explaining why this journalist, specifically, should care.

There is a structural reason the constraint is so widely shared. A reporter covering a beat may receive several hundred approaches a week against a publication schedule that can absorb a handful of stories. The filter is not a judgement about quality so much as a mechanism for processing volume, and brevity is what makes a pitch cheap enough to evaluate at all.

What does the AI finding mean for people sending material?

Two numbers from the same survey point in opposite directions. 82% of journalists report using at least one AI tool, up from 77% the year before — and 26% name unregulated or unchecked AI as a leading concern for the industry, up from 18%.

That is not a contradiction; it is a profession adopting a technology faster than it trusts it. Both figures rose together, which suggests use is being driven by workload rather than by confidence.

For anyone sending material, the risk is specific. A pitch that reads as machine-generated arrives in an inbox belonging to someone who is both more familiar with the output of these systems than the general population and more worried about it. Fluency is now weak evidence of effort, and the things that remain costly to fake — a named source who will talk, a document nobody has published, a specific factual claim that can be checked — carry proportionally more weight.

There is also a practical implication for how material should be formatted. If a substantial share of recipients are running pitches through summarisation tools, the details that survive that process are the concrete ones — names, figures, dates, a single clear offer. Anything carried by tone or emphasis is the first thing a summary discards.

What does this survey not establish?

It records what journalists say they prioritise, not what predicts coverage. Stated preference and observed behaviour diverge routinely in survey research, and no source reviewed here matches these responses against outcomes to establish which pitches actually became stories.

It is also a sample of journalists reachable through a media database, which is not a random sample of the profession. Reporters who maintain a public contact profile are more likely to be in it, and those reporters are, by definition, the ones most accustomed to receiving pitches.

And 897 responses across several countries cannot support fine-grained claims about particular beats or markets. The findings are robust as a description of broad professional preference and weak as a guide to any individual publication, which is the level at which pitching decisions are actually made.

The survey also cannot distinguish between what journalists want and what their publications will run. An individual reporter may genuinely prioritise original data while working somewhere whose commissioning decisions are driven by traffic patterns, seasonality or an editor's judgement about what the audience will read. Preferences expressed by the person receiving the pitch are not the whole of the process that determines whether a story appears.

The honest reading is therefore narrow: this is a good description of what journalists say they want, gathered from journalists who are easy to reach, and it should be used to shape an approach rather than to predict an outcome.

What follows for buying coverage?

The criteria journalists name are the criteria a paid placement bypasses rather than satisfies. Fit, credible sources and original data describe an editorial process, and a purchased placement substitutes a commercial transaction for that process.

That is not an argument that one is illegitimate. It is an argument that they are different products with different properties, and the survey describes what the earned route is actually gated on. For an organisation with a genuine expert willing to be interviewed, or with data nobody else holds, the gate is more passable than it is usually assumed to be — 58% and 40% of journalists are explicitly asking for exactly those two things.

Media Morpheus sells media and podcast placement. The finding above is that two of the three things journalists most want from a pitch are assets a company either has or does not, and cannot be bought from anyone.

There is a further consequence for how placement should be evaluated against the earned route. The two are usually compared on price and certainty: a placement has a fee and a delivery date, while a pitch has neither. What this survey adds is that the earned route's gate is narrower and more concrete than its reputation suggests — three named criteria, two of which are assets rather than skills — which makes the comparison less a matter of certainty versus chance than of whether a given organisation actually holds what journalists are asking for.

Put plainly: the survey describes a gate with three named locks, and money opens none of them.

Methodology and limitations

Figures are from Muck Rack's The State of Journalism 2026, published 19 March 2026. The survey contacted 1,044 journalists between 30 January and 2 March 2026, with 897 responses in the final analysis, predominantly in the United States with additional respondents in the United Kingdom, Canada and India. Four limits apply. First, Muck Rack is a media database and PR software company, and its respondents are reached through that database, so the sample is drawn from journalists who maintain a public contact profile rather than from the profession at large. Second, the figures record stated preferences rather than observed behaviour, and no analysis reviewed here links these responses to which pitches resulted in coverage. Third, respondents could select multiple priorities, so the percentages overlap and do not sum. Fourth, the sample size does not support breakdowns by beat, market or publication size, and none are reported here.

Conclusion

The thing journalists name most often is the thing least amenable to being purchased or improved: whether the subject is what they already cover. Below that, the two most-cited wants are access to credible sources and original data, and both are assets an organisation either possesses or does not. The length constraint is severe and nearly universal, which means the pitch itself is a filtering exercise rather than an argument. None of this makes paid placement illegitimate; it clarifies that the earned route is gated on a small number of concrete things, two of which cannot be bought, and that a company holding them is closer to the gate than it probably assumes.

Frequently Asked Questions

Sources

1. vendor research Muck Rack, 2026. The State of Journalism 2026, published 19 March 2026. Survey of 1,044 journalists, 897 in final analysis, fielded 30 January to 2 March 2026. Published by a media database and PR software company.
2. trade press GlobeNewswire, 2026. “Muck Rack's 2026 State of Journalism Report Finds 82% of Journalists Use AI”, 19 March 2026.