17 Questions to Ask a Digital PR Agency Before You Sign

Most agency selection processes fail in the last meeting, not the first. By the time a buyer reaches the final call, the deck has done its work: the case studies look impressive, the logos are recognizable, and the account lead is polished. What has not happened, in most cases, is a real interrogation of how the work actually gets done — the media relationships, the reporting standards, the people who will handle the account after the ink dries.

The gap between how agencies present themselves and how they deliver is not academic. According to Swydo's 2026 research, clients rank poor delivery as the number one reason they end agency relationships, while agencies rank it only seventh. The disconnect is a diagnostic: buyers routinely fail to test for what will later break the engagement.

What follows is a working script for the pre-signature call. Seventeen questions, grouped by what they surface, with the answers that should end the conversation.

Questions That Test Strategy, Not Just Enthusiasm

Strategy questions reveal whether the agency has thought about your business or is running a template. A credible digital PR shop should be able to explain a plan in terms of your audience, your competitive set, and the specific journalists who cover your category — not in terms of "brand awareness" and "thought leadership" as free-floating goals.

1. What is the single story you would pitch about us in the first 30 days, and to whom? A concrete answer names a reporter, an outlet, and an angle grounded in something you have said or done. A weak answer describes a "narrative pillar."

2. Which trade publications matter most in our category, and why? Trade coverage often outperforms tier-one for pipeline, particularly in B2B. If the agency defaults to Forbes and TechCrunch for a niche software buyer, they are optimizing for the pitch, not the outcome. Our own view on this sits in the piece on trade publication coverage.

3. What data do we already have that a journalist would want? The best digital PR programs mine internal data before commissioning surveys. Muck Rack's 2026 State of Journalism report found that journalists rank clear relevance to their beat (70%), interview access to relevant sources (58%), and original data or research (40%) as the most valuable pitch elements. If the agency has not asked to see your dashboards, they are guessing.

4. What would you refuse to pitch on our behalf? Agencies that say "everything is worth trying" are describing volume, not judgment. Cision's 2025 State of the Media Report found that 86% of journalists say they will immediately reject a pitch that isn't aligned with their beat or audience. Every misfire costs future access.

Questions That Expose the Delivery Model

The pitch team is almost never the delivery team. This is the single biggest source of post-signature disappointment, and it is trivially easy to test for.

5. Who, by name, will be pitching journalists on our account? Ask for LinkedIn profiles. Ask how long they have been at the agency. Ask what other accounts they staff.

6. What is that person's caseload? A senior outreach specialist juggling twelve accounts is not going to know your category. Six is aggressive. Ten is a warning.

7. What work is outsourced, and to whom? Some agencies contract writing, data analysis, or outreach to freelancers or offshore teams. That can be fine, if disclosed. It is not fine if you discover it after signing.

8. Show me your outreach cadence and QA process. A serious operator can walk you through their sending schedule, follow-up logic, and pre-send checks. The mechanics we advocate live in the outreach QA checklist. If the answer is "we use Muck Rack," they are describing a tool, not a process.

Questions That Force Proof, Not Case Studies

Case studies are marketing artifacts. They select for wins, omit the timeframe, and rarely name the campaign lead. Proof is different: it is verifiable, recent, and specific to the person on the call.

9. Send me live placement URLs from the last 90 days. Not lifetime highlights. Not "representative" examples. Actual links, earned inside the current quarter, ideally on accounts the pitching lead ran personally. Refusal to share them should end the meeting.

10. What was the pitch that earned this coverage? Pick one of the placements and ask for the email that landed it. A confident agency will share it, redacted where necessary. This tests whether the placement was earned through outreach or acquired through a marketplace.

11. Are any of these placements paid, sponsored, or from link exchange networks? Paid placements are a legitimate product when sold as paid; the problem is paid dressed as earned. BuzzStream's 2026 analysis of 257,267 sites on link marketplaces found that only 1.37% qualify as high quality or top tier. Most of what circulates as "digital PR inventory" is not what a serious buyer wants attached to their domain.

12. Can I speak to a current client on a similar retainer, unsupervised? Written testimonials are polished. Live conversations are not. Reluctance here is diagnostic.

Questions That Reveal How They Measure

Measurement questions expose whether the agency is optimizing for what you will pay for or what makes their monthly report look full. Impressions, AVE, and clip counts remain the standard filler of underperforming programs.

13. What are the three metrics on our monthly report, and why those three? The right answers vary by objective, but they should ladder to search visibility, brand mentions in target outlets, or qualified inbound. If E-E-A-T signals and LLM citations matter to your search strategy, they should be measured; the mechanics of the latter sit in getting cited by ChatGPT, Perplexity, and Claude.

14. How do you distinguish a placement that moved the needle from one that filled a slot? The link-count era is over. Backlinko's analysis of 11.8 million Google search results found that the #1 result has an average of 3.8x more backlinks than positions #2–#10 — but the composition of that link profile matters more than the raw count. An agency that cannot articulate placement quality is running a volume play.

15. What is a realistic timeline before we see ranking or pipeline movement? A credible answer sounds like: first placements in weeks two to three, search movement in months three to six, compounding after that. Anything faster is either lucky or bought.

Questions That Clarify the Contract

The final two questions are the ones buyers routinely skip because they feel adversarial. They are not. They are the questions that determine what happens when something goes wrong.

16. What are the exit terms, and what do we own on the way out? Media lists, pitch templates, custom research assets, and any earned content built during the engagement should be yours. Notice periods should be reasonable. Retainer minimums that outlast the notice period are a trap.

17. What would you tell us to walk away from this engagement? A confident agency will describe a scenario where they are the wrong fit — a category they cannot cover, a data profile too thin to be newsworthy, a leadership team unwilling to be quoted. An agency that cannot name a disqualifying condition is selling, not partnering. The strategist voice we look for in press quotes is described in the anatomy of a great press quote.

How to Read the Room During the Call

Answers matter, but so does texture. Watch for three patterns.

• Deflection under specificity. When you push for a name, a number, or a URL, does the agency get sharper or vaguer? Sharpness is the tell.
• Pushback on your brief. An agency that agrees with everything you say has no point of view. The right partner will challenge your positioning, your data hooks, or your target outlet list before signing.
• Named disqualifying conditions. When the agency says "this is when we would tell you to hire someone else," take it seriously. That is what a strategist sounds like.

The pre-signature call is the last moment of leverage a buyer has. After signing, the questions get harder to ask and much easier to defer. Seventeen questions is a lot, but the entire point of the meeting is to make the vendor earn a contract rather than close one. If any of the disqualifying answers surface, the honest response is to end the process. If none do, you have the beginning of an engagement worth building. Further diligence on engagement structure sits in the comparison of retainer vs per campaign models, and the broader philosophy of the work is described on how our campaigns actually work.