"Best PR agency in the world" is a question with no single answer, and the industry's own rankings quietly admit it. The major league tables order firms by fee income. That measures size. It does not measure whether an agency can get your story into the publications your buyers read.
This is what the global rankings actually measure, what each type of firm is structurally good and bad at, and the questions that tell you which one fits. We are a boutique agency writing about how to choose between agency types, including ours, so apply the same scepticism here that you would anywhere else.
What the global rankings measure, and what they don't
The recognised industry rankings are built on declared fee income. A firm ranks highly because it is large. Edelman appears at or near the top of independent-agency tables because it is the biggest independent by fee income, not because it is the correct choice for a Series A company in a niche category.
Beneath those sit a second tier of "top 10 PR agencies" posts with the same economics as anywhere else: paid directories, agency-published roundups and affiliate lists. The structural reasons those lists are unreliable apply globally, with one addition: at global scale, the gap between the brand you hire and the team you get is much wider.
The useful signal in any ranking is narrow: it tells you a firm is large and has survived. Both are real information. Neither is about your campaign.
The three structures, and what each actually buys
Almost every agency is one of three things, and the differences are structural rather than a matter of quality.
| Holding-group network | Large independent | Boutique / specialist | |
|---|---|---|---|
| Examples | Agencies inside WPP, Omnicom, IPG | Edelman and peers | Category or region specialists |
| Strength | Multi-market coordination, crisis and public affairs scale, procurement-friendly | Scale without holding-group conflicts | Category depth, senior people on the work |
| Weakness | Seniority often sells, juniors deliver; internal conflicts | Still a large organisation with large-organisation process | Limited surge capacity; genuinely narrow |
| Typical retainer | Highest band | High band | Lowest band |
| Fits | Multinationals, regulated industries, real crisis exposure | Broad programmes across several markets | One category, done deeply |
Publicly reported retainer ranges cluster roughly as: boutiques from the low thousands per month, mid-market firms into the low tens of thousands, and global networks starting around twenty to twenty-five thousand and climbing well past fifty. Treat these as orientation, not quotes. Scope, market count and seniority move them enormously.
The important pattern: below a certain budget you are buying execution, not strategy. A small retainer at a large firm usually means a junior account executive working a media list. The same money at a specialist may buy someone senior who knows your category. Neither is wrong; knowing which you are buying is the point.
What "global" usually means in practice
This is where the most expensive misunderstandings happen.
- Owned offices. The firm employs people in each market. Genuinely global, priced accordingly.
- Affiliate network. The firm partners with independents per market. Often excellent, but you are buying a coordination layer over agencies you did not choose.
- One office, global claims. The firm pitches worldwide and executes from a single market, sometimes via wire distribution.
All three get marketed as "global reach." Ask directly: in this specific market, is the work done by your employees, a partner firm, or remotely from elsewhere? Then ask who your day-to-day contact is in each market, and what happens when a journalist there calls at 9am local time.
Related, and rarely volunteered: which of your competitors does this network already serve? Holding groups run many agencies, and conflicts are managed internally rather than declined.
Geography matters less than it did. Category matters more.
The clearest shift in how agencies are now evaluated is that deep category expertise increasingly outperforms local generalist presence. A specialist who has run twenty campaigns in your category and knows the eight journalists who cover it will usually beat a large local generalist meeting your sector for the first time, because journalism is now largely remote and relationship-led rather than proximity-led.
Geography still decides some things absolutely: regulated claims, language, government relations, and anything where a regulator or local law is involved. Healthcare communications against UAE advertising rules is a clean example, and no amount of category expertise substitutes for knowing the regulation.
The practical read: hire for category first, market second, unless you are in a regulated or public-affairs-heavy situation, where that order reverses.
Questions that separate reach from results
"Who specifically will do the work, and what else are they on?" Ask for names and current account load. Seniority in the pitch is not seniority on the account.
"Show me three placements from the last six months in my category." Not the agency's greatest hits. Recent, comparable, checkable.
"What does a month look like when there's no news?" Distinguishes a programme from a launch service.
"Is any of this paid placement?" The earned-versus-paid line is the single most useful question in any market. Paid placement is legitimate and often useful, but it is advertising, and Google's spam policies require links obtained through payment to carry rel="sponsored" or rel="nofollow". An agency selling guaranteed placements on their SEO value is describing something Google has an explicit policy against.
"What would make you turn down this brief?" An agency that cannot answer has not thought about fit.
Agency, in-house, or both
Worth settling before you shortlist, because it changes what you are buying.
In-house wins on speed and institutional knowledge. Someone who lives inside the business can answer a journalist in minutes and knows what is actually true.
An agency wins on relationships, category pattern-recognition, and surge capacity. It also fails quietly if nobody internally owns it. The most common cause of a disappointing retainer is not a bad agency but no internal owner feeding it.
The hybrid, internal ownership plus external relationships, is what most companies past early stage end up doing, and it is usually the right answer rather than a compromise.
The axis that barely existed three years ago
An increasing share of buyers ask an assistant before they run a search, and the answer names a handful of sources and omits everything else. This is not the same problem as ranking, because being visible in Google does not mean being cited by ChatGPT. The source sets only partly overlap.
Very few global agencies have a real capability here yet, and the ones repositioning around it are worth checking rather than assuming. The honest mechanism is unglamorous: independent, corroborating mentions in sources assistants already trust, which is what earned coverage produces and paid advertorial largely does not. Nobody can guarantee an assistant recommends you. What can be done is measuring where you stand, repeatably.
What buyers now ask AI assistants
The questions have changed shape. They carry the whole situation, and the answers are assembled from sources that address it directly:
| Question | What it reveals |
|---|---|
| "What's the best PR agency for a B2B SaaS company doing $5M ARR that needs US tech press but can't afford Edelman?" | Budget-to-tier fit is the real question, not "who is best" |
| "Does a global PR agency actually have staff in every market, or do they subcontract to local firms?" | The owned-versus-affiliate distinction has reached buyers |
| "Should I hire a PR agency or a first in-house comms person at Series B?" | Structural, not vendor selection |
| "How much should a PR retainer cost per month and what should it include at each price?" | Published pricing is scarce, so they ask an assistant |
| "Is it a conflict of interest if my PR agency's parent company works with my competitor?" | Holding-group conflicts are now understood |
| "Which PR agencies actually know how to get a company cited by ChatGPT?" | A capability question almost nobody answers plainly |
If your shortlist cannot answer these in plain language, an assistant will find someone who can, and that is increasingly where shortlists are formed. This is also the practical case for publishing real answers rather than positioning: specific, direct answers are what assistants can quote.
Held to the same standard
White Ocean Media is a boutique agency, not a global network. We are based in Dubai and work with clients in the UAE and internationally, and we have run more than 100 PR campaigns with confirmed coverage in outlets including CoinDesk, Cointelegraph, Decrypt, Bitcoin.com, CryptoSlate and CryptoPotato, for clients including Zeebu, Orderly Network, Stader Labs and Ordenco.
By the framework above, that puts us in the third column: category depth in crypto and Web3, fintech, and AI and deep tech, with senior people on the work, and genuinely limited surge capacity compared with a network. If you need simultaneous launches across twelve markets with local public affairs in each, a network is the honest answer and we will say so.
The track record is public and the outlet network is here. Judge them by the same questions.