Most PR plans fail in the same way: they are a list of publications somebody wants to appear in. That is a wish, not a plan. A plan states what the business needs, who has to believe something different, what that something is, and which activities in which weeks will move it.
Here is the structure that survives contact with a real quarter, followed by a filled-in example.
What a PR plan is actually for
Two jobs. First, it forces a decision about what PR is supposed to change: pipeline, hiring, investor perception, buyer confidence late in a deal cycle. Second, it makes PR schedulable. Most PR that never happens dies not from lack of ambition but because on a random Tuesday nobody could tell what to do.
A working plan fits on two pages. If yours runs twenty slides, you have built a deck for approval, not a document for work.
Part 1: Objectives tied to business goals
Write the business goal first, then the PR objective under it, then the signal you will watch. Never write the PR objective alone, because an objective with no business parent is how teams end up chasing logos.
- Business goal: close enterprise deals faster. PR objective: give buyers third-party proof at the evaluation stage. Signal: coverage and analyst mentions cited in deal reviews.
- Business goal: hire twelve engineers. PR objective: make the engineering culture legible outside the company. Signal: inbound applicants who name a specific article or talk.
Two or three objectives. Not seven. A quarter has room for two or three.
Part 2: Audiences
Publications are not audiences. People are. For each objective, name the human you need to reach and where they actually read.
Be specific enough that a tactic falls out of the description. "Developers" produces nothing. "Platform engineers at 500–5,000 person companies who evaluate tools through their peers, read a couple of practitioner newsletters, and distrust vendor blogs" tells you to pursue engineering podcasts and practitioner-authored technical posts, not a trade-press campaign.
Part 3: Messages
Three to five sentences you want repeated back to you. Each one should be defensible with evidence you can produce today: a number, a customer, a working product.
The test: strike out any message a competitor could also claim. What is left is your actual message. If nothing is left, you have a positioning problem that PR cannot fix, and you should solve that before you spend a dollar on distribution.
Keep a short list of proof points under each message, and a shorter list of things you will not say. The second list is what keeps a spokesperson out of trouble.
Part 4: Tactics
Now, and only now, pick activities. Map each to an objective and an audience, and be honest about what each one produces.
- Announcements. Funding, launches, partnerships, milestones. High effort, brief lift, and only worth it when the news clears the bar. Run it through the newsworthiness test before committing a week to it.
- Distribution. Puts the news on the record and makes it findable, but it does not make a journalist write. See what distribution does and does not do.
- Direct pitching. The thing that produces actual articles. Requires a real list (see building a media list) and a pitch that reads like a tip rather than a brochure.
- Bylines and contributed articles. Slow, controllable, durable. Good for messages that need explanation rather than announcement.
- Podcasts and speaking. Best for founder-led categories and for messages that benefit from a human voice.
- Reactive commentary. Responding to journalist source requests and breaking news. Cheap, unpredictable, excellent filler between announcements.
A healthy mix is roughly one announcement a quarter and a steady hum of everything else. Companies that do announcements only are silent ten months a year.
Part 5: Calendar
Put the tactics on a twelve-month grid, working backward from fixed dates: funding close, product GA, conference CFP deadlines, fiscal-year buying seasons. Fill the gaps with the always-on work.
Attach lead times, because they are the thing people get wrong. Award submissions and conference CFPs close three to six months ahead. Monthly-magazine features work two to four months out. A byline takes two to six weeks from idea to publication. An announcement wants a two- to three-week runway for pre-briefs.
Part 6: Measurement
Pick metrics before the work starts, and pick ones you can actually collect.
- Output metrics: placements, bylines published, talks delivered. Honest but shallow.
- Quality metrics: did the article carry your message, did it name your product, is the outlet one your buyer reads.
- Position metrics: share of voice against three named competitors, tracked quarterly.
- Business metrics: branded search volume, inbound that cites coverage, coverage referenced in sales conversations.
Set the review cadence now: monthly for output, quarterly for everything else. And set expectations about time: PR compounds slowly, and the realistic timeline is quarters, not weeks.
A worked example: Series A SaaS
Hypothetical: a 40-person company selling incident-response tooling to platform engineering teams, six months after a Series A.
Objectives. (1) Support a pipeline goal by making the company a credible name during vendor evaluation, with the signal being mentions in deal reviews and branded search. (2) Support hiring ten engineers, with the signal being inbound applicants citing published work.
Audiences. Platform and SRE leads at mid-market companies, who learn from practitioners rather than vendors. Secondarily, senior engineers considering a move, who read the same sources.
Messages. (1) Most incident tooling optimizes for the alert, not the recovery. (2) The company has data from a large volume of real incidents showing where recovery time is actually lost. (3) The product is built by people who carried pagers.
Tactics. One announcement: a research report built from anonymized incident data, which is genuinely new information rather than a product update. Direct pitching to a list of roughly twenty-five reporters and newsletter writers who cover reliability engineering. Two bylines from the CTO. Four podcast appearances. Two conference CFPs. Reactive commentary whenever a major public outage is in the news.
Calendar. Report research in Q1, launch in Q2 with pre-briefs two weeks ahead. CFPs submitted in Q1 for autumn events. Bylines in the off-months. Podcasts spread one per month.
Measurement. Quarterly share of voice versus three named competitors. Monthly count of qualified placements. Branded search volume. A one-line question added to the sales call notes template: did the buyer mention seeing us anywhere.
Common failure modes
- Naming target publications before naming the audience. Always backwards.
- Treating the plan as permanent. Revisit quarterly, and kill tactics that produced nothing twice in a row.
- No owner. Every line needs a name next to it, including the founder's own hours.
- Measuring only volume, which rewards publishing noise.
Start with one page covering objectives, audiences, and messages. The rest is scheduling. If you want the scaffolding, our templates cover the announcement pieces the plan will call for.
Want help turning the plan into published news? [See how it works](/how-it-works).