Most of what a PR agency does day to day is research and pitching that produces nothing, and that is not a failure of the agency, it is the structure of the work. In 2026 journalists receive over 200 pitches a day, up from 50 to 80 in 2020, with many reporters seeing 500 or more emails a week, and typical cold media pitch reply rates sit in the low single digits. A mid market retainer at $8,000 to $15,000 a month buys roughly two to three placements, which means the majority of billed hours go into media list building, monitoring, drafting, and outreach that never converts. Understanding that arithmetic is the difference between a productive agency relationship and a frustrated one.
Here is the actual composition. A typical month at a mid market retainer breaks into five recurring workstreams: monitoring and research, message and asset development, media list building, pitching and follow up, and reporting. Only the fourth of those produces anything a client sees as output, and it depends entirely on the first three being done well. The tools that make it possible are the same across most agencies: Cision, Muck Rack, Prowly, Meltwater, Qwoted, Featured, and Google Alerts, plus a wire account with PR Newswire, Business Wire, GlobeNewswire, or EIN Presswire.
What are the five things a PR agency does every month?
Every retainer, at every price tier, resolves into the same five workstreams. The tier changes the depth, not the categories.
1. Monitoring and research
Someone reads the news for you every morning. That means scanning trade publications, newsletters, journalist social feeds, competitor coverage, and alerts through tools like Meltwater, Muck Rack, Talkwalker, or Google Alerts, looking for both risk and reactive opportunity. This is where newsjacking opportunities come from, and it is the most consistently undervalued part of the work because it produces nothing visible on the days nothing happens.
2. Message and asset development
The messaging document is the most important deliverable a PR agency creates, and most clients never think about it. It becomes the source material for pitches, releases, website copy, executive talking points, investor narratives, and social content. Alongside it sit the press kit, executive bios, boilerplate, fact sheet, and a story bank of angles that can be pitched over the coming quarter.
Curious whether the coverage you already have is showing up when buyers ask AI engines about your category? Get a free AI visibility audit and see what those answers currently say.
3. Media list building
Not a purchased list. A real list is built reporter by reporter, checking who currently covers your beat, what they wrote in the last ninety days, what they explicitly say they do not want, and how they prefer to be contacted. Muck Rack and Cision provide the database layer, but the filtering is manual and it is where junior hours go. Lists decay fast, since media layoffs and beat changes mean a list built a year ago is substantially wrong today.
4. Pitching and follow up
The visible work. Personalized pitches sent in waves, each tailored to the specific reporter and their recent coverage, followed by one or two follow ups. Against a 200 pitches per day inbox, personalization is the only thing that moves reply rates above noise. A list of 150 reporters producing a handful of conversations, of which a few become coverage, is a normal successful month.
5. Reporting and account management
Coverage reports, a monthly call, and increasingly some measure of message pull through and share of voice. This is also where scope gets renegotiated, which is why the monthly call matters more than clients typically treat it.
Where does the money actually go?
Roughly speaking, senior strategy time is the smallest share of hours and the largest share of cost, while list building and outreach are the largest share of hours at the lowest hourly rate. That blend is what a retainer price reflects.
A mid market retainer at $10,000 a month typically covers something in the range of 40 to 60 hours of blended team time depending on the agency’s rate structure. Those hours split across a senior strategist who sets direction and handles the highest value relationships, an account manager who runs the day to day, and a coordinator who builds lists and handles logistics. When an agency sells you a senior name and staffs the work with a coordinator, that is the specific failure mode to watch for, and asking who will actually be sending pitches is the question that surfaces it.
The other cost most clients do not see is tooling. Cision and Meltwater subscriptions run into five figures annually for an agency, Muck Rack and Prowly are cheaper but not free, and wire distribution is billed either as a pass through or built into the retainer. Ask which model applies, because a wire release billed separately at national rates can quietly add a meaningful amount to a monthly invoice.
What has changed about the work in 2026?
Two things changed materially: inbox saturation from AI generated pitching, and the fact that coverage now feeds AI answers as much as it feeds referral traffic.
The saturation problem is measurable. Reporter inboxes went from 50 to 80 pitches a day in 2020 to over 200 in 2026, and a large share of the increase is AI generated outreach that reads as templated. The effect is counterintuitive: as automated pitching increased, the value of genuinely researched, human pitches went up rather than down, because the baseline got so much worse. Agencies that responded by sending more pitches faster are producing worse results than agencies that cut volume and increased specificity.
The second change reshaped what a placement is worth. A trade publication piece that sends 40 referral clicks used to be considered a weak placement. In 2026 that same piece may be the reason ChatGPT, Perplexity, or Google AI Overviews names your company when a buyer asks for recommendations in your category, because AI engines ground their answers in recognized third party domains rather than in your own marketing site. That makes outlet authority and topical relevance more important than traffic, and it makes evaluating an agency on click volume actively misleading. Our digital PR for AI visibility guide covers how the two disciplines have merged.
A third smaller change: contributed content and expert sourcing platforms have partly replaced the old HARO workflow. Qwoted, Featured, and similar services now handle a meaningful share of expert quote placement, and a competent agency should be working those queues daily rather than treating them as an afterthought.
What does a PR agency not do?
An agency does not guarantee coverage, does not control what a journalist writes, and cannot manufacture news you do not have. Those three limits explain most client disappointment.
Guarantees are the clearest signal to evaluate. Any agency guaranteeing placement in a specific named tier one outlet for a fee is selling paid placement or advertorial rather than earned media, and under FTC disclosure rules that content should be labeled. Paid contributor programs like Forbes Councils, the Fast Company Executive Board, and the Entrepreneur Leadership Network are legitimate and useful, charging four figure annual membership fees for vetted contributor bylines, but they are content distribution rather than press coverage and should be priced and reported as such.
Agencies also do not fix a fundamental story problem. If your company has nothing new to say, no data, no customer results, no executive perspective, and no product change, then no amount of pitching skill converts. The most common reason a retainer underperforms is not agency quality. It is that the client could not supply a genuine story roughly monthly, and the program drifted into releases about nothing.
Finally, agencies do not move fast without you. Slow client approvals kill more placements than bad pitching does, because reporter windows are hours, not days. If your legal review takes a week, tell the agency up front and structure around it.
How should you evaluate whether an agency is doing the work?
Ask for the inputs, not just the outputs, because the inputs are where you can actually tell. Four questions surface almost everything.
Ask to see the media list with the rationale column. A real list explains why each reporter is on it and what they covered recently. A purchased list has neither, and it will show.
Ask how many pitches went out, to whom, and what the reply rate was. A competent agency tracks this. If the answer is only “we pitched widely,” the work is not being measured internally either.
Ask what got killed and why. Pitches that reached a conversation and did not convert are the most informative data in the whole engagement, and an agency that only reports wins is hiding the diagnostic information you need.
Ask where the coverage landed on the authority spectrum, not just how much there was. Ten pickups on syndication aggregators is a different thing from one substantive piece in a recognized trade publication, especially now that AI engines weigh source authority when deciding whom to name. Our publication tiers guide explains how those tiers sort in practice, and our press placement services page shows how outlet tiers map to real programs.
Frequently asked questions
What does a PR agency do on a daily basis?
Five recurring workstreams: monitoring news and journalist activity through tools like Meltwater, Muck Rack, and Google Alerts; developing messaging documents, press kits, and story banks; building and refreshing media lists reporter by reporter; writing and sending personalized pitches with follow ups; and reporting on coverage. Only pitching produces visible output, and it depends entirely on the other four. Most billed hours go to work the client never sees.
Why does a PR agency cost so much if I only get two or three placements?
Because most of the work does not convert by design. Journalists receive over 200 pitches a day and typical cold pitch reply rates sit in the low single digits, so producing two to three placements a month requires building lists of a hundred or more relevant reporters, personalizing outreach to each, and following up. A $10,000 monthly retainer typically covers 40 to 60 hours of blended team time plus tooling costs for Cision, Meltwater, or Muck Rack subscriptions.
What tools do PR agencies use in 2026?
The standard stack includes Cision and Muck Rack for media databases and journalist contact data, Meltwater or Talkwalker for monitoring, Prowly for pitching and newsroom hosting, Qwoted and Featured for expert quote sourcing, Google Alerts for basic monitoring, and a wire account with PR Newswire, Business Wire, GlobeNewswire, or EIN Presswire for distribution. Ask whether wire costs are passed through or included, since national releases can add meaningfully to an invoice.
Can a PR agency guarantee coverage in a specific publication?
Not for earned editorial. Any agency guaranteeing placement in a named tier one outlet is selling paid placement or advertorial, which under FTC disclosure rules should be labeled as sponsored. Paid contributor programs such as Forbes Councils, the Fast Company Executive Board, and the Entrepreneur Leadership Network are legitimate and charge four figure annual membership fees for vetted bylines, but they are content distribution rather than earned press and should be reported separately.
Has AI changed what PR agencies do?
In two ways. Inbox saturation from AI generated pitching pushed reporter volume from 50 to 80 pitches a day in 2020 to over 200 in 2026, which raised the value of genuinely researched human outreach rather than lowering it. And coverage now feeds AI answers directly, since ChatGPT, Perplexity, and Google AI Overviews ground responses in recognized third party domains. A low traffic trade placement can be the reason an engine names your company months later.
How do I tell if my PR agency is actually working?
Ask for inputs rather than outputs. Request the media list with a rationale for each reporter, the count of pitches sent and the reply rate, an account of which conversations did not convert and why, and where coverage landed on the authority spectrum rather than just how much there was. Agencies that report only wins are withholding the diagnostic information you need, and agencies that cannot produce pitch metrics are not measuring internally either.
Press coverage now feeds the answers your buyers get from AI engines, whether or not it sends clicks. Run a free AI visibility audit and find out what your existing coverage is actually doing for you.
The uncomfortable truth about PR is that you are buying a probability, not a product. An agency’s job is to raise the odds that a journalist with 200 pitches in their inbox opens yours, and every hour of monitoring, list building, and message development exists to move that probability a few points. Judged as a purchase of guaranteed coverage, every retainer looks overpriced. Judged as sustained access to reporter attention, with assets that compound and coverage that now trains the AI engines your buyers ask for recommendations, the good ones are worth what they charge. Ask for the inputs, supply a real story every month, and answer emails fast. Those three things separate the engagements that work from the ones that quietly do not.
Tagged