Getting Your Executive Director Published: Writing an Op-Ed With AI

Why Editors Reject Almost Everything

Rejection in opinion is usually silent, which deprives nonprofits of the feedback that would let them improve. It is worth knowing the reasons anyway, because they are consistent across outlets and almost all of them are fixable before you send. Industry commentary on the submissions pile returns repeatedly to the same short list: too broad, too promotional, or too obvious. Understanding which of those three you fell into is more useful than another draft.

No news peg. The piece could have run last year or next year, so it runs neither. Opinion sections are part of a news product, and the question an editor asks first is why this, this week. A peg does not have to be dramatic. A county report released Tuesday, a new state figure, a budget hearing on the calendar, a school board decision, a scheduled rate change, or the start of a season that makes a problem visible all work. What does not work is a general observation that a problem exists, however important the problem is. Writing guidance for prospective authors consistently puts timeliness and the strongest available peg at the center for exactly this reason.

Organizational self-promotion. This is the most common nonprofit-specific failure and the one editors find most irritating, because it reads as an attempt to get advertising space for free. Guidance from practitioners and from publications themselves converges on a firm rule: the organization should not appear in the body of the piece. The byline already carries whatever promotional value exists, and the text itself has to earn its place by being useful to a reader who has never heard of you and will not remember your name. If your organization is named three times in 650 words, the piece is about you, and opinion pages are not about the author.

No argument. Covered above, and worth repeating because it survives multiple drafts. The test is a single question: can you state, in one sentence, something a reasonable, informed, well-meaning person would deny? If you cannot, you have a theme and not a thesis, and no amount of polish converts one into the other.

Wrong length. Outlets publish word counts and mean them. CalMatters, for instance, states in its published submission guidelines that commentary runs to no more than 650 words and letters to no more than 150. Some papers are stricter still and will not consider a piece meaningfully longer or shorter than their stated target. A 1,100-word submission to a 650-word slot tells the editor you did not read the guidelines, which is a reasonable proxy for how much other care went into the piece.

Simultaneous submission. Most outlets require exclusivity, and many say so explicitly. Sending the same piece to four papers at once and hoping one bites is the single fastest way to burn a relationship with an opinion editor, because if two run it, both have been embarrassed. Exclusive means exclusive until declined, and the mechanics for handling that honestly are covered further down.

No local relevance. Regional and local papers want the local angle, and some now say outright that submissions should address local issues and come from people in the circulation area. A piece about a national trend with no county-level consequence belongs in a national outlet you probably cannot reach. The same piece rewritten around what the trend means for your three school districts belongs in your paper, and has a real chance.

The author has no standing. Opinion editors want the piece from someone whose position makes the claim credible, which generally means the executive director, a program leader with direct knowledge, a board chair, or someone with lived experience of the issue. A submission under a communications director's name on a substantive policy question is a harder sell, because the role does not supply the authority. This is a large part of why the byline matters strategically, and why these pieces run under the leader's name rather than the organization's.

Word Counts and Format Conventions by Outlet Tier

Length is not a matter of taste. It is a slot on a page or a template in a content system, and the number in the guidelines is the number. The right way to work is to find the target outlet's own published requirements before writing a word, because the limit shapes the structure. A 1,000-word piece can carry a secondary point and a short narrative opening. A 550-word piece can carry one argument and nothing else, which means the choice of argument has to be made before drafting rather than discovered during it.

Across outlets, three broad tiers appear. National papers and national policy outlets tend to run guest pieces in a band from roughly 400 to 1,200 words, with most slots clustering well below the top of that range. Regional and metropolitan dailies commonly target somewhere between 550 and 750 words, and some are explicit that they will not consider submissions much outside their stated figure. Nonprofit and philanthropy trade publications are generally the most generous: the Chronicle of Philanthropy's own guidance puts most published pieces at roughly 1,000 to 1,200 words and allows for somewhat longer, which is why the trade press is often the better home for an argument that genuinely needs room.

Letters to the editor are a separate and much tighter format. CalMatters caps letters at 150 words and asks that they reference the exact headline and publication date of the article being addressed, along with the writer's name, city, and organizational title where relevant. That is representative of the form generally. A letter is a response to something the outlet published, it is short enough to write in twenty minutes, and its acceptance rate is far better than an op-ed's, which makes it a seriously underused tool rather than a consolation prize.

Format conventions beyond length are mostly about removing friction. Submit in the body of the email or as a plain attachment, matching whatever the guidelines specify. Put the word count at the top. Supply a two-line author identification exactly as you want it to appear, since that tag is where your organization legitimately lives. Include a daytime phone number, because fact-checkers call and an unreachable author can lose a slot that was otherwise going to run. Many outlets ask for an author photograph and a short biography, so have both ready rather than scrambling when an editor replies on a deadline. Avoid formatting that will be stripped anyway, meaning no headers, no pull quotes, no logos, no attached letterhead. Provide sources for every factual claim, either as a short list at the end or as inline links, because verification is the step that most often stalls an otherwise accepted piece.

Op-Ed, Letter, or Trade Byline: Picking a Target You Can Actually Hit

Most nonprofits aim too high on their first attempt, get silence, and conclude that opinion placement is not available to organizations their size. The better approach is to treat the three available formats as a portfolio rather than a hierarchy, because they do different jobs and have wildly different acceptance rates.

The local or regional op-ed is the realistic flagship for most organizations. Your community paper, your metro daily's regional section, a statewide nonprofit news site, or a well-read local outlet gives you the audience that actually matters: the people who fund you, serve on the boards you need, sit on the commissions that decide your issue, and live in the neighborhoods you work in. A national placement is a nicer sentence in a newsletter. A local placement moves things. It also has materially better odds, because the pool of credible local voices on your issue is small and you are in it.

The letter to the editor is the most underrated and the most neglected. It is 150 to 250 words, it responds to something the outlet just published, it takes a fraction of the effort, and it gets accepted at a far higher rate. It also builds exactly the thing you need for a future op-ed, which is a record of being a reliable, publishable voice on a subject. A communications function that produces one short, sharp letter a month whenever the paper covers your issue is doing more for its leader's public standing than one unsuccessful op-ed attempt a year. Note that letters often require the writer to live in the circulation area and to cite the headline and date of the piece they are answering.

The trade or sector byline is where a longer, more technical argument belongs, and where the audience is funders, peer organizations, and sector leadership rather than the general public. Philanthropy and nonprofit publications accept pieces from inside and outside the sector, allow more room, and are explicitly interested in arguments that complicate conventional thinking rather than confirming it. Guidance from both the Chronicle of Philanthropy and Nonprofit Quarterly's submission page points the same direction: accessible writing, a real argument, examples beyond your own organization, and minimal field jargon. If your argument is about how the sector works rather than about a local decision, this is the right home for it and the general-interest desk was never going to take it.

A sensible annual plan for a small communications function might be a dozen letters, three or four local op-ed attempts tied to predictable news moments such as budget season or the release of an annual report, and one trade piece developed properly over a few months. That is achievable by one person with AI assistance, and it compounds. The leader who has been published eleven times locally is a different proposition to an editor than the one submitting cold, and the broader positioning work this supports is covered in our article on AI-supported advocacy for nonprofits.

Where AI Actually Helps in This Pipeline

Everything above is the craft, and none of it is new. What has changed is that a one-person communications function can now run this pipeline without a month of evenings, because most of the labor is research, structuring, compression, and correspondence. Work through the stages in order, and note that the model's role gets smaller as the piece gets closer to the leader's own words.

Finding the news peg. The hardest practical problem is not having an argument, it is noticing the week when the argument becomes timely. A model with web search can scan recent local coverage of your issue area, list what has been published in the last fortnight, and identify which items have an angle nobody has taken yet. Ask specifically for scheduled events too: hearings on the calendar, reports with known release dates, budget cycles, and seasonal patterns. That turns opinion writing from reactive scrambling into a planned activity, since the peg for your February piece is usually visible in December. The standing version of this capability is a monitoring habit, which we treat in our article on AI media monitoring for nonprofits.

Stress-testing whether the thesis is arguable. This is the single highest-value AI task in the whole process and it takes two minutes. Give the model your one-sentence thesis and ask it to make the strongest honest case against it. If it struggles, or produces only objections about implementation, your thesis is not arguable and the piece will not place. If it produces something that makes you uncomfortable, you have a real argument and you also have the counterargument paragraph you were going to need anyway. Ask it as well for the three most obvious reader objections and whether the claim would be read as self-serving coming from an organization in your position, which is a question nonprofits are usually too close to answer.

Starting from the leader's spoken argument, not a blank page. This is the technique that makes the whole thing work, and it inverts the usual order. Do not ask a model to write an op-ed about your issue. Instead, record the executive director talking for fifteen minutes: what they think, why they think it, what they have seen, what they would say to someone who disagreed. Transcribe it, then ask the model to find the actual argument inside the transcript, list the evidence the leader cited, flag the strongest phrases in their own words, and propose an outline built around them. What comes back is not generic commentary. It is your leader's thinking, organized. The fifteen minutes of talking is the part that cannot be delegated, and it is also the part leaders are happy to do, because talking is not what they have no time for. Writing is.

Preserving the voice rather than flattening it. The default failure mode of AI drafting is institutional mush: smooth, balanced, unobjectionable, and completely anonymous. That is fatal in opinion writing, where the voice is a large part of the value. Two practices help. First, feed the model real samples of the leader speaking and writing, and instruct it explicitly to preserve the characteristic sentence rhythms and vocabulary rather than improving them. Second, mark the sentences that came verbatim from the transcript and protect them through every subsequent edit, because those are the lines an editor will quote and a reader will remember. When a draft starts producing phrases your leader would never say, the fix is not another prompt, it is returning to the transcript. The general discipline here is the same one we describe in our article on brand voice consistency with AI, applied to one person instead of an organization.

Cutting 1,200 words to 650 without losing the spine. Compression is tedious, mechanical, and the task AI does best. The approach that works is not asking for a shorter version, which produces a blurred summary. Instead, ask the model to identify which paragraphs serve the central thesis and which serve a secondary point, then cut the secondary material entirely rather than trimming everywhere. Then ask it to find the sentences that restate something already said, the qualifiers that weaken claims without adding accuracy, and the sector jargon that could be plain English. A 1,200-word nonprofit draft usually contains about 400 words of hedging and repetition, and removing those gets most of the way to a 650-word target before any real content is sacrificed. Review the result against your protected verbatim lines to confirm the voice survived the surgery.

Building the outlet target list. This is pure research and it is worth doing once properly. Have a model assemble a list of plausible outlets with, for each one, the opinion desk contact, the published word limit, the exclusivity requirement, the stated response window, any local-residency or subscriber rules, what the author tag should contain, and whether they state an AI policy. Then verify every line against the outlet's own guidelines page, because submission requirements are exactly the kind of specific, changeable detail models get confidently wrong. Used as a research brief rather than an answer, it turns a week of hunting into an afternoon of confirming, and the resulting table serves every future submission.

Drafting the pitch note and the follow-up. Short correspondence in a consistent tone is straightforward delegation. Have the model produce the three-sentence cover note, the single follow-up after your stated window, the thank-you to an editor who ran the piece, and the note that offers it to the next outlet with a revised opening. Keep these as templates so the tenth submission costs ten minutes rather than an hour.

Repurposing after publication. A published piece is reusable in a dozen directions, and this is the stage small teams skip because the win feels complete. The method and the formats are covered in our article on repurposing content with AI, and the specifics for an op-ed appear in the next section.

What to Do With the Piece After It Runs

Publication is the midpoint, not the end, and this is where the return on the whole effort is either collected or left on the table. Start with the mechanics of permission, because they are easy to get wrong. The outlet typically holds rights to the published version, so check the terms before reposting the full text anywhere. The reliable approach is to link to the original from your own channels, quote a short excerpt, and add context rather than duplicating the article, which also avoids competing with the outlet for search traffic on the piece you want them to keep publishing.

Then work the audiences the piece gives you access to. Send it to your board with a short note on why the argument matters now, because this is one of the few artifacts that shows governance what public leadership from their organization looks like. Send it to the funders who care about the issue, as evidence of positioning rather than as an ask. Send it to the officials or staff at agencies whose decisions the piece addresses, with a single line offering to talk further, since the whole point of a specific ask is that someone might act on it. Send it to peer organizations, because a shared argument advanced by a sector is more durable than one organization's op-ed. Each of those is a two-sentence email, and all of them can be drafted at once.

The piece then becomes source material, and this is the straightforward part to delegate. One op-ed yields a newsletter item framing the argument for supporters, a short social thread pulling out the three strongest lines, talking points for the next time the leader is interviewed on the topic, a paragraph for the next grant application that demonstrates thought leadership, a section for the annual report, and the opening of a longer piece for your own site that can carry the material the word limit forced out. Our articles on scaling storytelling with AI and AI-assisted blog writing both cover the production side of that in more detail.

Finally, invest in the relationship, because that is the compounding asset. Thank the editor briefly and without asking for anything. Note what they changed in editing, since those edits are free instruction in what the outlet wants. Record the outlet, contact, date, and outcome in a simple log. If the piece draws responses, including critical letters, answer them with grace in public and then leave it, because engaging an argument well in the days after publication is often what makes an editor think of you next time. Six months of that is how an organization goes from submitting cold to being asked.