A request for proposal is meant to get comparable answers from firms you would be happy to hire. Most fail at that: they ask fifty questions, withhold the budget, invite fifteen agencies and request speculative creative, with the result that the strongest firms decline and you choose from whoever was least busy.
State the budget, or accept incomparable proposals
Without a budget, each agency guesses at a different scale and you receive proposals that cannot be compared on anything. A range is enough. The fear that agencies will simply quote the maximum is largely unfounded and is in any case better than the alternative, which is a shortlist of proposals solving four different problems at four different prices.
Invite three or four, not fifteen
A serious response costs an agency real money, and the odds decide who bothers. Three or four invitations produce genuine effort; a dozen produce templates. Do the screening before the RFP through conversations and case studies, and tell the firms you invite how many others are in the process, which is the single question every agency asks and few clients answer.
Do not ask for free creative
Speculative creative made without research, access to your data or a proper brief is theatre: it tests presentation rather than thinking, and the better agencies increasingly decline to provide it. Ask instead for how they would approach the problem, what they would need, what they would do first, and a case study of comparable work. If you want real thinking on your brief, pay for a short paid discovery from two finalists.
Ask what you will actually use to decide
Who does the work day to day, what happens when they leave, how the fee is structured, who owns the accounts and the assets, what the notice period is, and what has gone wrong on a previous account and how it was handled. Those answers differentiate. Boilerplate about values and process does not, and every question you add reduces the care spent on the ones that matter.
Questions people ask about rfp advertising
How long should agencies get?
Two to three weeks for a considered response. A one-week deadline guarantees templated answers.
Should we publish the evaluation criteria?
Yes, with weights if you have them. It produces proposals aimed at what you care about instead of at what they guess you care about.
Is a pitch presentation necessary?
Meeting the actual team is. A polished presentation from people who will not work on the account tells you very little.
Should we pay for pitches?
For anything requiring substantial work, a modest fee to finalists is fair and noticeably improves what you receive.