AI tools for agency new business and pitching split into two groups: the ones that help you produce a pitch faster, and the ones that change which pitches you enter. The published cost data argues for the second. An agency spends an average of $204,461 on a new-business pitch, and two out of three reviews end with the incumbent kept.
What a pitch actually costs
The ANA and the 4As hired Advertiser Perceptions to price this, and their report, The Cost of the Pitch, surveyed both sides of the table in March 2023. A non-incumbent agency spends $204,461 responding to a pitch. An incumbent defending its own account spends $406,092, close to double, because it fields 18 people where a challenger fields 11. Add the clients internal costs and a three-agency review costs the industry around $1.02 million, or $1.19 million once an incumbent joins.
Europe looks the same. EACA surveyed 412 agencies for its Cost of Pitching 2025 and put the average agencys annual pitching bill at 650,937 euros, with the average new-business pitch at 43,804 euros. EACA also calculated a break-even: the process pays only if an agency wins business worth at least 7 million euros.
The number that should change your behaviour
The same researchers asked clients what happened after the review. Two in three kept the incumbent.
Then they asked what decided the selection. Cost and price came top at 62 percent. Creative execution scored 45 percent. The strategic big idea scored 36 percent.
So you are usually bidding against someone who already holds the account, in a contest where the deciding factor is your rate card rather than your thinking.
The AI is pointed at the wrong half
The AI tools sold into agency new business today mostly draft: the deck, the concepts, the credentials, the RFP response. They compress the hours between the brief landing and the room.
That works, and it is the cheaper half of the problem. A faster deck does not tell you whether to enter, and it does not move your price. You have lowered the cost of the bet and left the odds alone.
You could argue the opposite. If a pitch costs less to run, enter more of them, and EACAs break-even falls with it. But incumbency and price both sit on the clients side of the table, and neither responds to your production speed. Cheaper drafting buys more entries in a contest you lose twice out of three.
The expensive half is qualification, which means knowing which accounts are in play and holding a view of the client before the RFP lands rather than after. That is agency operations work rather than creative work, and most agencies run it out of somebodys inbox.
What always-on has to mean
Multiply settled its positioning in August 2026 as an Agency OS with always-on agents. For that to describe anything, something has to arrive before you ask for it.
In new business, that shipped on 27 August 2026 as Recurring Brief Proposals. Every Workspace now runs a weekly job that researches news for its client or market, drafts up to three brief proposals, and puts them on the Workspace Overview. A Workspace sits above projects and holds a clients strategy documents, brand guidelines and reference material, and that context flows down into the work beneath it.
What matters is the trigger. A proposal written on a schedule against a client profile you already keep feeds the decision about whether to pitch. A deck written on demand two weeks out comes after that decision.
Three questions before you enter
Is there an incumbent, and are they defending? Two in three reviews end where they started, and a defending incumbent spends about double what you will.
Where does price sit in the brief? Cost decided 62 percent of these selections, so if your rate card does not answer the brief, your creative will not rescue it.
Do you hold a view of this client already? If you will build one after the RFP lands, you are buying at deadline what the winner has been accumulating for months.
Ask those three before the team says yes. Most agencies would enter fewer pitches, and win a larger share of the ones they did.
