RFP Policy

We'll Answer Your RFP on Two Conditions

The RFPs are pouring in

AI Search is new. Every executive can see the risk: customers asking ChatGPT instead of Google, AI Overviews answering the question before anyone clicks, competitors showing up in answers where your brand doesn't. Very few teams are ready to act on it. So they do what large organizations do when they're not sure what to buy. They issue an RFP. We're getting more of them in a few weeks than we used to see in a year. The deadlines are measured in days. And every one asks for the same things in a slightly different spreadsheet. We get why you're doing it. You've been told to pick a partner in a category that barely existed two years ago, and an RFP feels like the safe, defensible way to do that. But we're tried and true. We've been leading Search since 2014, we were building for AI Search before it had a name, and we're not going to spend our clients' time writing proposals for teams that aren't ready to act.

RFPs work when you know what you're buying

An RFP is a great tool when the thing you're buying has a spec. Servers. Paper. A media buy. You define the requirements, vendors respond to the same requirements, and you compare apples to apples. AI Search doesn't have a spec yet. The RFPs we receive prove it. SEO, AEO, GEO, AIO and LLMO get used interchangeably, sometimes in the same paragraph. Then the document asks the agency to define the KPIs, design the plan, pick the pilots and map out the first 90 days. In other words, it asks us to write the spec, so you can decide whether we're allowed to do the work the spec describes. That's a homework assignment, not a selection process. And when nobody has defined the problem, the RFP doesn't measure who does the best work. It measures who writes the best proposal.

What you're actually asking for is free consulting

Here's a sample of what recent RFPs asked us to hand over before anyone signed anything:
  • Three pilots, each with an objective, approach, required inputs, timing and success measures
  • Our recommendation on the company's core strategic question, on a 6-month and a 3-year horizon
  • 90, 180 and 365-day plans
  • A sample audit from a comparable client, plus a sample executive briefing
  • A forecasting methodology and KPIs for every workstream
  • An annual investment estimate, before any discovery
  • A gap analysis of their websites
That's not a questionnaire. That's the engagement. It's also expensive to produce. The average RFP response takes 33 hours and nine people, according to Loopio's 2026 benchmark. Our enterprise bids tie up a big chunk of our team for weeks. Those hours come out of the work we do for clients who already hired us. Then read the terms. What you submit usually becomes the buyer's property. It can be shared with their advisers and consultants. It has to stay valid for a year. Some buyers want 105 days to pay the invoice if you win. You wouldn't ask a law firm to argue half your case to earn the retainer. Don't ask your search partner to build half your strategy to earn the RFP.

The decision is usually made before the RFP goes out

Most RFPs aren't open questions. They're confirmations. You can see it in the requirements. Some are written so tightly around the incumbent or a preferred vendor that only one agency fits: a specific mix of advisor personalities, a particular tool stack, proof that a certain reporting cadence is already running for a similar client. Everyone else is there to fill out the comparison grid. Our time is collateral damage. Just as often, the RFP exists to educate the incumbent and drive their price down. Our ideas show the incumbent what good looks like. Our pricing becomes the number they're asked to beat. The terms make that easy, because everything we submit becomes the buyer's property. When it does come down to a real decision, procurement mechanics usually make it: a preference for staff-augmentation models, 90- or 105-day payment terms, then price. The RFPs we win start with a conversation long before the RFP exists. That's not a coincidence. It's why our first condition is time with your team.

Why this matters more in AI Search

Search has moved from ranking to recommendation. Winning now means being found, understood, trusted, cited and recommended across Google, AI Overviews, AI Mode, ChatGPT, Gemini, Perplexity and the agents that will soon shop on your customers' behalf. Nobody has a spec sheet for that yet. So the right partner isn't the one with the prettiest proposal. It's the one who can think through problems nobody has solved yet. You don't need an RFP to see that. You can see it in public work. Here's ours:
  • We were building for AI Search before it had a name. In 2023 we published how Search Generative Experience works and why retrieval-augmented generation is our future, two years before "AEO" and "GEO" entered the industry's vocabulary.
  • We wrote the manual. The AI Search Manual is our documentation for Relevance Engineering in AI Search. It's free to read.
  • We open-sourced the tooling. Qforia simulates the query fan-out AI search systems run behind the scenes, so you can see the sub-queries your content has to answer.
  • We published how to measure it. Our AI Search metrics framework splits measurement into input, channel and performance metrics.
  • AI recommends us first. In a July 2026 study of which agencies AI assistants recommend to buyers, iPullRank came out on top for all three buyer profiles tested. AI Search now drives more discovery for our own business than any other channel.
Our method pairs Relevance Engineering, which covers what the machines require, with Resonance Design, which covers what your audience needs. It spans shared, owned and earned media, so a narrow SEO spec sheet misses most of the job. Everything you need to judge how we think is already public. The only thing a proposal adds is your data, and that's the part worth paying for.

How we'll take part

We're not saying no to every RFP. We're saying no to RFPs that won't tell us who we'd work with or what problem we'd be solving. Here's how it works.
  1. We talk with your team first. Before we write anything tailored, we talk with the people who own the problem, not only procurement. If your process doesn't allow that, you get our standard response packet instead of a custom proposal.
  2. Strategy is paid. Pilots, 90-day plans, audits and recommendations are the work. When you want them, they go through our AI Search Strategy Program: Keyword Portfolio, Omnimedia Content Audit and Plan, Measurement Plan, Relevance Engineering Technical Assessment and AI Search Strategic Roadmap. Start where you are. There's an entry point whether you need clarity before you scale, want to challenge your market, or need to lead your category.

    You keep the work whoever you pick. It's the "prove it first" phase most RFPs ask for anyway, and unlike a proposal, it doesn't end as a document your team has to interpret alone. We stay involved through prioritization, implementation, QA and measurement.
  3. Everything else, we answer as usual. The packet covers capability questions. We'll sign a mutual NDA, complete security and vendor questionnaires, and share references once you're ready to move forward.

If you have to run a process, run a better one

Some organizations require a formal process. Fine. Here's how to pick an AI Search partner without wasting everyone's time, including yours.
  1. Define the problem before you write the document. What decision are you making, and what would change your mind? If you can't answer that, you're not ready to evaluate agencies.
  2. Give every shortlisted agency time with the team. Disclose your budget range, your incumbent, and how you'll decide. Agencies that can't ask questions can only guess.
  3. Cut the questionnaire to what changes the decision. Read the public work instead. How an agency thinks is already on the record.
  4. Pay your finalists for a short discovery sprint. You'll learn more from a few weeks of real work on your data than from a hundred pages of proposal.
  5. Check references at the end. Good agencies protect their clients' time, and so should you.
Do that, and you'll be picking a partner, not a proposal writer.

We'd rather work on your actual problem

The average RFP eats 33 hours. We'd rather spend them on your actual problem, not your marketing word problem. Request our standard RFP response. It answers the questions nearly every AI Search, SEO, AEO and GEO RFP asks: who we are, how we work, how we measure, what it costs and how we handle your data. An index maps the common RFP questions to our answers, so you can drop them straight into your template. Fill out the form and a member of our team will reach out shortly to send it over and answer any questions. Rather see how we'd think about your business? Start with the AI Search Strategy Program.

FAQ

What counts as strategy work?
Anything built on your business: pilots, 90-day plans, audits, forecasts, recommendations on your strategic questions, or sample deliverables using your data. Questions about who we are and how we work aren't strategy work. The packet answers those for free.
Our procurement rules don't allow vendor calls during an RFP. What now?
Then you get our standard response packet. If you can't talk to us, you can't see how we think, and how we think is what you're buying.
Do RFIs follow the same rules?
Yes. Most RFI questions are capability questions, and the packet answers them. Its index maps common questions to our answers so your team can fill in its template.
Will you sign our NDA?
Yes. We sign mutual NDAs.
Can we get references?
Yes, once you're ready to move forward. We get a lot of RFPs, and our clients' time matters to us as much as yours does.
Does this apply to RFPs already in progress?
We'll finish the bids where we've already spoken with the team. Everything new follows this policy.

Running an RFP? Start Here

Our standard RFP response covers the questions most teams ask before choosing an agency. Request a copy and we’ll reach out shortly.