How Many Agencies Should You Invite to Pitch? Fewer Than You Think
You have a dozen or more social media agencies on a spreadsheet, and someone suggests sending all of them the brief. Do not. Three finalists is the working number, and the real work is cutting the long list before any agency writes a word.
Why one agency and ten agencies both fail
The buyer who talks to one agency, usually because of a warm referral, loses three things at once. There is no price anchor, so a quote of any size looks plausible. There is no way to tell a strong proposal from an average one, because you have nothing to hold it against. And there is no leverage in negotiation, since the agency knows it is the only option on the table.
The buyer who invites ten loses something less obvious. By the seventh proposal you are skimming. They blur together, and the decision quietly falls back on chemistry or the lowest number. Neither of those is what you set out to evaluate.
To be plain about it, this is a process judgement from vetting agencies, not a data finding. I am not citing a study. I am describing what tends to happen to attention and incentives.
The cost of a pitch that nobody puts on the invoice
A serious proposal takes senior time, and an unpaid pitch has to be recovered somewhere. My reasoning, and it is reasoning rather than a measured fact, is that the likeliest places are the fee quoted to whoever wins, or the seniority of the people who end up doing the real work after the pitch team moves on.
Picture two agencies with the same brief. The first believes it has roughly a one in three chance of winning, so it sends its best strategist and a tailored plan. The second has been told, or has guessed, that ten agencies are competing. It estimates a one in ten chance, so it recycles a template and hands it to a junior. Same agency, very different pitch.
You do not control how much effort an agency chooses to spend. You do control the odds each one sees, because the number of invitations sets them. Invite three, say so, and the agencies that matter to you have a reason to bring their best people.
Why three, and when it is two or four
Three gives you a real comparison without a tiebreaker meeting. With two, you cannot tell whether either is typical. With three, one agency tends to land in the middle, and that proposal anchors how you read the other two.
There are honest exceptions. Two is reasonable for a narrow specialist need, where few agencies qualify and inviting more means inviting bad fits. Four can be justified when the scope spans several markets or several platforms and the finalists differ by type.
The limit is how many proposals you can honestly read closely. Three is my judgement from vetting agencies, not a study result and not an industry standard.
How to cut a long list before anyone writes a word
Write your criteria down before you look at the list. If you set them afterwards, you will find reasons to keep the agencies you already like. Then apply these cuts in order, using only what you can learn without a proposal.
- Fit by agency type and size. Passing: the agency’s usual client is roughly your size and it offers the service you need, whether that is management, paid social, content or influencer work. Failing: a boutique that has never handled your volume, or a network whose smallest client dwarfs your budget.
- Category and platform experience. Passing: it works in your category or a close neighbour and runs the platforms you need as a core service. Failing: platforms listed as an afterthought, or a category history unlike yours.
- Named staffing. Passing: the agency can say who would run your account, their role and roughly how much of their time you would get. Failing: “a dedicated team” with no names or roles.
- Pricing posture against your budget range. Passing: how the agency prices, whether by retainer, project or package, plausibly sits near your range. Failing: it only sells far above your range, or so low that real senior staffing is implausible. For whether to share your number at all, see whether to tell agencies your budget.
A worked cut from fifteen to three
This example is hypothetical. Picture an unnamed regional retail brand that built a long list of fifteen agencies from a directory plus a few referrals, then ran the four cuts above in three passes.
| Pass | Agencies left | Who fell out and why |
|---|---|---|
| Pass 1: type and size | 10 | Two large networks whose typical client was far bigger. Two influencer-only shops. One freelancer collective that could not cover the volume. |
| Pass 2: category, platform and staffing | 6 | Three agencies with no retail or consumer experience. One that could not name anyone who would staff the account. |
| Pass 3: pricing posture | 3 | Two agencies whose package structure started well above the brand’s range. One whose pricing was so low the brand doubted any senior person was included. |
The cut the brand might regret skipping was a well regarded agency, strong on creative, that fell out in pass 3 on pricing. The brand had a fixed budget range, and a pitch it could not afford to accept would have wasted both sides’ time. Cutting it was still right.
What to tell the agencies you cut, and the ones you keep
A short, specific decline now is better for an agency than a silent wait of three weeks, and it costs you two sentences: say you are going a different direction and name the category, such as budget fit or scale. You owe a timely no, not long feedback.
Finalists need three facts: how many other agencies are pitching, the timeline, and whether the pitch is paid or unpaid. Honest odds let an agency size its effort. If the scope is complex, consider a paid option, as covered in the paid pilot post.
Sharing brand assets with finalists raises an NDA decision. Make it before the materials go out, and read what to put under NDA before a pitch first.
Where this sits in the larger process
To run the RFP itself, start with the RFP process post. For where to find agencies worth evaluating, see the sourcing post.
Build your long list
Browse vetted agencies, currently 82 published, to build the long list you will then cut. If you run an agency yourself, you can list your agency free.
FAQ
Is three agencies always the right number?
No. Three is a working number and an editorial judgement, not a finding. Two can be right for a narrow specialist need and four for a multi-market scope, and the test is how many proposals you can read closely.
Can I ask just one agency for a proposal?
You can, but you lose the price anchor and any way to judge whether the proposal is strong. If you have a referral you trust, still bring in at least one comparison.
Sources
- Partner in Social agencies directory, fetched 2026-10-07 (HTTP 200)