Key Takeaways

  • A good RFP should show you how an agency thinks, not just what they charge. If every agency can give you basically the same answer, the RFP isn't doing its job.
  • You really only need nine sections. Company context, the problem, scope, success metrics, your stack, budget, timeline, agency questions, and evaluation criteria.
  • Share your budget range. Withholding it is the single fastest way to get proposals you cannot compare and to waste four weeks of your own calendar.
  • Ask agencies to respond to your problem rather than recite their capabilities. The best signal in any proposal is whether they reframed what you asked.
  • Send the RFP to three to five agencies, not fifteen. More than five and you are running a beauty contest instead of a hiring process.

Most of the B2B tech RFPs that land in our inbox were written to protect the buyer from making a bad decision, not to help them make a good one. You can feel it in the first page. Thirty questions about ISO certifications and DEI policy, two sentences about the actual business, no budget, and a deadline that assumes we have nothing else going on. We answer them anyway, because sometimes there is a real company underneath the template, but the process the buyer just kicked off is going to hurt them more than it hurts us.

This is the RFP I wish more B2B tech marketing leaders would send. The goal is pretty simple: get a few proposals you can actually compare, from agencies who understood the problem well enough to push back on it. Copy it, change the words, send it. If you want the longer version we use with clients who ask us to help them run the search itself, it is inside our agency selection blueprint, goes beyond the RFP and walks through the larger selection process. For now, start with the nine sections below.

What a B2B tech agency RFP is actually for

The mistake I see most often is confusing an RFP with a spec sheet. A spec sheet lists what you want built. An RFP is asking someone to tell you how they would build it, and, more importantly, whether they think you are building the right thing at all. In B2B tech, where the buying committee is big, sales cycles are long, and the ICP shifts every time a new competitor raises a round, the strategy question matters more than the deliverables question. Your RFP has to make room for that.

Which means two things. First, the problem section has to be longer than the deliverables section. Second, you have to leave the door open for the agency to disagree with your framing. If every question is closed-ended, you will get closed-ended answers, and you will pick the cheapest one, and six months later you will be writing another RFP.

The nine sections, in order

Everything below goes in the document, in this order. Keep the whole thing under ten pages. If it is longer than ten pages, you have not decided what the project is yet, and no agency response is going to fix that for you.

1. Company context (one page, no more)

  • What we do, in one sentence a non-customer would understand.
  • Who we sell to. Named ICP, named personas, and the industries where we win most often.
  • Stage and size. Revenue band, headcount, funding stage if relevant, growth rate.
  • The three competitors we lose to most, and why we think we lose.
  • What has changed in the last twelve months that made this project necessary.

That last bullet is the one people skip and it is the most important. Agencies read it to figure out what is actually going on. A new CRO, a repositioning, a flat quarter, a product launch that underperformed. Tell them.

2. The problem (this is the section that matters)

  • The business outcome we are trying to move, in plain business language rather than a marketing metric.
  • What we have already tried, and what happened when we tried it.
  • Our current hypothesis about why the problem exists.
  • What we think we need help with, and what we are explicitly not asking for help with.
  • An invitation: if you think we have framed this wrong, tell us in your response.
That invitation is the single highest-signal line in the whole RFP. Any agency worth hiring will take you up on it. Any agency that ignores it and just answers your questions as posed is telling you exactly how the engagement will go.

A recent RFP that came across our desk had thirty questions in it and never once named the business outcome the team was trying to move. It asked about our SOC 2 status and our PTO policy and the tenure of our creative director. The winning agency was going to be whoever priced lowest, because there was nothing else in the document to evaluate against. That is exactly what happened.

3. Scope of work

  • The specific deliverables or workstreams in scope. Website, demand gen, content, ABM, brand, product marketing support, whatever it is, name it.
  • What is out of scope, so nobody wastes a page pitching it.
  • Whether this is a project, a retainer, or a project that could become a retainer.
  • What internal resources will be available to the agency. Named people, hours per week, decision authority.

4. Success metrics

  • The one or two metrics we will use to decide whether this worked.
  • The baseline for each metric today, and the target.
  • The time horizon for hitting the target.
  • What data sources we will use to measure, and who owns them internally.
If you cannot fill this section in, you are not ready to send the RFP. Go back and figure it out. An agency cannot design a program against a goal you have not defined, and any agency that pretends they can is going to invoice you against a goal they invented.

5. Current marketing stack and team

  • The systems in play: CRM, marketing automation, CMS, analytics, attribution, ABM platform, content tools. Name them.
  • Who owns each system today, internally or externally.
  • The gaps you already know about.
  • Any tools you are actively evaluating or planning to replace.

This one saves everyone time. If you are on HubSpot and the agency is a Marketo shop, both of you want to know that before the pitch, not after.

6. Budget range

Share a range. I know the traditional procurement advice is to withhold it and let the market bid, and I know why that feels safe, but it produces worse outcomes for you. When budget is hidden, agencies either lowball to win and cut corners later, or shoot high because they do not know what you can absorb. Either way, you end up comparing proposals that are not actually comparable, and the evaluation turns into a spreadsheet exercise instead of a strategy conversation. Say the number.

  • Annual or project budget range, with a floor and a ceiling.
  • How the budget breaks down between fees and media or production, if relevant.
  • Whether there is flexibility if the agency proposes a scope you had not considered.

7. Timeline

  • RFP sent date.
  • Questions due back from agencies, and how you will answer them (usually a shared FAQ so nobody has an information advantage).
  • Proposal due date. Give people at least three weeks. Two weeks is a red flag; one week is an insult.
  • Shortlist notification date.
  • Pitch or working session dates.
  • Decision date.
  • Kickoff date.

8. What we want the agency to tell us

Keep this section tight. Six to eight questions, not thirty. Every question should be one you would actually use to decide.

  • How would you reframe the problem we described, if at all, and why?
  • What is your recommended approach in the first ninety days, and what would you expect to learn from it?
  • Who specifically would work on this account, with names, roles, and the percentage of their time we would get?
  • Show us one piece of work you have done for a B2B tech company at a similar stage, and walk us through what you would do differently now.
  • How do you price this kind of engagement, and what would a rough fee structure look like against our budget range?
  • What would have to be true for this engagement to fail, and how would we know early?
  • Two client references we can talk to, ideally one current and one former.

That last question, the failure one, is the one I would fight to keep in. It flushes out the agencies who think in scenarios versus the ones who only think in wins.

9. Evaluation criteria

Tell them how you will score. Not the exact weights, but the shape.

  • Strategic fit and quality of thinking on our specific problem.
  • Relevant B2B tech experience at our stage.
  • The team we would actually work with, day to day.
  • Commercial fit within our budget range.
  • Cultural fit and how the conversations felt.

Notice what is not on that list: agency size, awards, client logo count. Those shouldn't decide the search. If you weight them, you will end up with a bigger agency than you need and a more junior team than you were promised.

When an RFP is the wrong call

Not every agency search needs one. If you already know two or three agencies through your network, and you trust the people who referred them, you will get a better result from three ninety-minute working sessions than from a formal RFP. The working sessions cost you less time, they show you how the agency thinks in real time, and they surface chemistry problems that a written proposal will hide. We have won and lost plenty of both, and the working-session process is a better predictor of the engagement going well.

The RFP is the right tool when you have procurement involvement you cannot skip, when the spend is large enough to require a paper trail, or when you genuinely do not have a network to draw from and need to run a wider search. Otherwise, consider whether you are adding process because you need it or because it feels safer. Sometimes the honest answer is the second one, and that is fine, but at least name it.

How to send it, and to whom

Three to five agencies. Not fifteen. Every agency you add past five creates more work without giving you much more useful information, because agencies at your stage cluster into a small number of archetypes and you are mostly getting variations of the same three answers. Pick two agencies you found through referral, one you found through content that impressed you, and one wildcard, then stop. If none of them work out, you learned something and you can run a second round with a sharper brief. That is a better outcome than shortlisting from a field of fifteen mediocre responses.

Send the document as a PDF, with a short cover email that names the person on your side who will answer questions, and a link to a shared FAQ doc. When an agency asks a good question, add it to the FAQ so every other agency sees the same answer. This is how you keep the process fair and, more selfishly, how you avoid answering the same question five times.

Frequently Asked Questions

How long should a B2B tech marketing RFP be?

Under ten pages. If yours is longer, you have either not decided what the project is, or you are asking questions you will not use in the decision. Cut anything that will not change your shortlist.

Should we share our budget in the RFP?

Yes, as a range. Withholding budget produces proposals you cannot compare and turns the evaluation into a pricing exercise instead of a strategy conversation. It matches what we see on the receiving end every week: hidden budgets produce worse hires.

How many agencies should we send it to?

Three to five. Past five, you're doing a lot more work without learning much more. If you cannot narrow to five, your brief is not tight enough yet.

How much time should we give agencies to respond?

Three weeks minimum, four if you can. Anything under two weeks tells serious agencies you do not respect their time, and the ones who respond fastest are usually the ones with the least going on, which is not the signal you want.

What is the single best question to include?

Ask each agency how they would reframe the problem you described. The answer tells you whether they are actually thinking, and it is the closest a written document can get to showing you what working with them will feel like.

Sources