How to choose a B2B marketing agency: the questions that actually matter

Why agency selection processes produce bad outcomes

The typical agency selection process works like this. A company identifies three to five agencies, sends them an RFP or a brief, receives responses, reviews the decks, shortlists two or three candidates, invites them to present, and then makes a decision based on some combination of chemistry, price, and the quality of the presentation. The agency that wins is almost always the one that ran the best sales process, not necessarily the one that would produce the best marketing outcomes.

This selection mechanism has a built-in flaw. Agencies that are good at winning new business have invested in their pitching capability. The best performers at a new business presentation are often senior people who move on to other pitches once the contract is signed. The work is then done by more junior team members who were not in the room when the contract was won. The company selected an agency based on the performance of people who will not be doing the work.

There is also a fundamental misalignment in what the RFP process reveals. A well-constructed RFP evaluates how an agency responds to a brief. It does not evaluate whether the agency can write a good brief, because the brief was provided by the client. The actual capability being evaluated in an agency selection is the agency's ability to diagnose a marketing problem and design an approach that addresses it. The RFP process, by providing the brief, removes exactly the diagnostic step where strategic capability is revealed or absent.

The result: companies select agencies based on execution capability demonstrated in a controlled presentation context, then discover three to six months into the engagement that the agency does not have the strategic depth to make the fundamental positioning and channel decisions the work actually requires.

The difference between a marketing agency and a marketing partner

Before asking which agency to hire, the more useful question is what type of relationship the company needs. Most companies in the $5M to $100M revenue range do not need a marketing agency in the traditional sense. They need a marketing partner: a firm that can diagnose the strategic problems above the execution layer, help the company understand what it should be doing before it does anything, and hold accountability for business outcomes rather than just deliverables.

The full distinction between these two models, and the specific questions to ask to determine which type of relationship is appropriate for the company's current situation, is addressed in the post on the difference between a marketing agency and a marketing partner. The short version: if the company already has a clear strategy and just needs execution capacity, an agency relationship may be appropriate. If the company needs help figuring out what the strategy should be, it needs a partner relationship, and evaluating agencies through an RFP process will almost certainly produce the wrong result.

The six questions that reveal actual capability

These six questions are specifically designed to probe strategic depth rather than execution competence. They should be asked in a live conversation, not submitted as a written request. The quality of the answer in real time reveals things that a prepared written response cannot.

Question 1: Show me the brief you wrote before you produced this work. Every agency has a portfolio. Most portfolios show the finished campaign, the creative execution, the performance numbers. Very few agencies can produce the strategic brief that preceded the campaign. If the agency cannot show the strategic document that defined the problem, the target audience, the message architecture, and the success criteria before the work was built, there is a reasonable chance the work was built without one. The brief is where strategy lives. Campaigns without a brief are execution without direction.

Question 2: What was the specific business problem you were solving, not the campaign objective? There is a meaningful difference between "we were asked to generate 500 MQLs in Q3" and "the company was entering a new vertical and needed to establish credibility with a buyer audience that had no existing awareness of the product category." The first is an execution objective. The second is a business problem. An agency that cannot articulate the business problem behind any given engagement is almost certainly being used as an execution vendor rather than a strategic partner. That may be appropriate, but it needs to be understood going in.

Question 3: What did you recommend the client NOT do? Strategic capability is often more visible in restraint than in ambition. An agency that recommended a client not invest in a particular channel, not pursue a particular campaign, or not enter a particular market, and can explain the reasoning, is demonstrating independent judgment. An agency that says "we execute what the client asks for" is describing a vendor relationship, not a strategic one.

Question 4: What would you change about this engagement if you ran it again? This question is designed to probe intellectual honesty. An agency that has a thoughtful answer demonstrates both the self-awareness to recognize where the work fell short and the transparency to share that assessment. An agency that answers "we are very proud of the results we achieved" is performing rather than reflecting. Perfect satisfaction with completed work is a signal that the agency is not learning, which is a problem in a field that changes as rapidly as marketing does.

Question 5: Who specifically will be doing the work, not presenting it, but doing it? Get names. Ask for samples of work from the specific people who will be on the account. Understand whether the senior strategist who impressed in the pitch will be involved in weekly execution reviews or only in quarterly strategy sessions. The gap between the people who pitch and the people who execute is one of the most consistently cited disappointments in agency relationships.

Question 6: What metrics will you report on at month three that are leading indicators of the business outcome, not vanity metrics? An agency that answers this question with website traffic, social engagement, and email open rates is describing activity reporting. An agency that answers with pipeline contribution by source, MQL-to-opportunity conversion rate by channel, and CAC trend by audience segment is describing outcome reporting. The difference in these answers reflects a fundamental difference in how the agency understands its accountability.

Agencies that win RFPs are good at pitching. The capability that matters is whether they can diagnose the actual problem and design a motion that addresses it, before any execution begins.

Red flags in the sales process

Several patterns in the agency sales process are reliable indicators of problems that will appear in the engagement itself.

Agencies that lead with a creative portfolio for a strategy role. If the agency's primary response to "we need help with marketing strategy" is a deck of creative work, brand campaigns, and visual identity systems, there is a mismatch between what is being sold and what is needed. Creative work is downstream of strategy. An agency that leads with creative in a strategy conversation either cannot do the strategy work or does not understand that it is being evaluated on strategy.

Agencies that cannot explain their process without referencing tools. "We use HubSpot for automation, SEMrush for search visibility, and Hotjar for conversion analysis" is a description of a tool stack, not a process. A strategic marketing process has a diagnostic phase, a strategy development phase, a prioritization phase, and an execution phase, each with specific activities and specific outputs. An agency that can describe its process in tool terms but not in strategic terms has not developed a methodology. It has assembled a technology stack.

Agencies that propose a retainer before diagnosing the actual problem. If the first commercial conversation results in a proposal for a monthly retainer covering "content, inbound, social, and paid" without a diagnostic phase, the agency is proposing execution before understanding what the company actually needs. A retainer is appropriate after the strategic architecture is defined. Proposing one without that foundation is selling time, not outcomes.

How to evaluate case studies and references

The standard case study format, in which an agency presents a client name, a campaign description, and a performance number, is nearly useless as an evaluation tool. The performance number is rarely contextualized against a baseline, the campaign description rarely specifies what strategic decisions were made, and the client name carries social proof without substantive evidence.

Third-party review and research platforms can fill some of that gap, but only if the buyer understands what each platform is actually measuring. G2's research methodology scores providers on review quality, recency, source, and service-specific attributes like professionalism, expertise, responsiveness, and execution ability, with additional detail in its scoring methodologies documentation and review validation process. That makes G2 useful as a filter for verified-buyer feedback, not as a final authority. Clutch plays a similar role for service providers, structured around verified client reviews and market research. For enterprise-grade validation, the Forrester Wave on Marketing Creative and Content Services, Q1 2025 evaluates and scores the most significant providers in the category against a defined criteria set, which is the closest thing to an objective third-party benchmark. The practical takeaway: G2 and Clutch tell you what buyers say about an agency; Forrester tells you how analysts evaluate the category. Neither is a substitute for the diagnostic questions above, but they are reasonable proof signals to cross-reference once a shortlist exists.

The information that is actually useful in a case study: the state of the marketing program before the agency engagement began, specifically the specific problems that existed and why they had not been solved. The strategic decisions made early in the engagement: what channels were prioritized, what were deprioritized, and why. The specific constraints that shaped the approach: budget, timeline, internal team capability, competitive situation. The outcomes, with a baseline that makes the performance number meaningful rather than just large.

Ask for the strategy document that preceded the campaign. Not the creative brief and not the media plan. The strategic document that defined the problem, the ICP, the message architecture, and the success criteria. If the agency produced one, they should be willing to share it in a reference review. If they cannot produce one, the campaign was built without a strategic foundation.

References should be asked one specific question that goes beyond satisfaction: what did the agency recommend that you were initially skeptical of, and how did it turn out? This question reveals whether the agency was exercising independent judgment or confirming the client's existing assumptions throughout the engagement.

What the engagement model should look like before you sign anything

Before a retainer or project contract is signed, three things should be explicitly defined in the engagement documents or a pre-engagement conversation.

First, a definition of success that is tied to business outcomes rather than deliverable volume. "Twelve blog posts per month" is a deliverable. "Increase marketing-sourced pipeline by 35% over six months" is a business outcome. The contract should reference the latter, and there should be a shared understanding of how it will be measured. Understanding what budget is appropriate to set against a given agency relationship is one of the exercises covered in the B2B marketing budget framework.

Second, a definition of the diagnostic phase. The first deliverable in any strategy-level engagement should be a diagnostic document that describes the current state of the marketing program, identifies the primary constraints on growth, and proposes a strategic architecture. The diagnostic should be completed before any ongoing execution begins. An agency that wants to skip the diagnostic and move directly to execution is either confident it already has enough information, which is rarely warranted, or is more interested in beginning the billing cycle than in building the right foundation.

Third, clarity on who is accountable for what. The client is accountable for providing access, approvals, and input on the company's strategy and market context. The agency is accountable for the strategic architecture and the quality of execution. When something underperforms, the accountability framework determines whether the problem is in the client inputs or the agency outputs. Without this framework, underperformance conversations tend to become political rather than diagnostic.

Specific considerations for B2B technology companies

B2B technology companies evaluating agency relationships face a set of considerations that do not apply equally to all B2B sectors. The agency needs to understand B2B buying cycles, not just digital channels. The distinction matters because B2B technology purchases typically involve multiple stakeholders, long evaluation periods, and content requirements that map to the evaluation stages rather than a single conversion event.

An agency that specializes in direct response or consumer digital marketing can often translate some of those skills to B2B contexts, but they will have a learning curve around B2B buyer psychology, multi-stakeholder deal dynamics, and the content architecture that supports a 90-day enterprise evaluation. That learning curve has a cost that is paid from the engagement budget.

Fintech companies specifically face an additional constraint. The considerations covered in the post on fintech marketing strategy are relevant to agency selection: an agency that has not worked with regulated financial products will not instinctively understand the trust-building requirements, the compliance constraints on marketing claims, or the conference and publication ecosystem where fintech buyers form vendor assessments. These are learnable, but they take time and the company pays for that learning during the engagement.

The most reliable proxy for an agency's ability to serve a specific B2B technology vertical is not a case study from a company in that vertical. It is evidence that the agency's team members have spent time understanding the buyer's professional context: the problems they face, the constraints they operate under, the way decisions get made, and what trust signals are meaningful to them. An agency that demonstrates that depth of buyer understanding in a capability conversation is more likely to produce relevant strategy than one that demonstrates familiarity with the same marketing tools without that buyer-level context.

The difference between an agency that delivers work and a firm that builds marketing capability is addressed at length in what a marketing strategy diagnostic actually produces. The diagnostic is often the right starting point before any agency relationship begins: it clarifies what the strategic architecture should be, which makes it significantly easier to evaluate whether a given agency is equipped to execute against it or whether what is actually needed is a different type of firm entirely.

Frequently asked questions