How to Evaluate a B2B Web Design Agency (Before You Sign Anything)

TL;DR: Most B2B companies run agency evaluations built for vendor selection, not strategic partnership. They over-weight the pitch team, compare prices before understanding scope, and skip the questions that would predict delivery. This guide walks through a four-stage evaluation process: initial screen, pitch call, proposal review, reference check. Each stage includes the specific questions that separate agencies that can build a B2B website from agencies that say they can.

Your website project will be one of the more consequential decisions your team makes this year. The agency you choose will shape how your buyers experience your brand for the next three to five years, how your sales team uses the site in active deals, and whether your demand gen team has a foundation that supports their work.

The evaluation itself gets far less attention than it deserves. Most B2B companies run a process that was designed for procurement, not partnership. They issue an RFP, collect proposals, and compare them on price and aesthetics. Both factors are easy to rank and nearly useless as predictors of outcome. If you have already read through why B2B websites underperform, you have seen what happens when the wrong agency builds the wrong thing.

A Splunk website refresh produced a 113% increase in form conversion and a 59% increase in time on site. A Sift engagement generated a 192% increase in conversion rates. Those results did not come from the agency with the lowest bid or the prettiest slides. They came from a process where the right questions got asked before the contract was signed.

This guide gives you a framework for that process.

Stage 1: Screen before you schedule

The first cut should happen before anyone gets on a call. You are not looking for the best agency in the pool. You are removing agencies that do not belong in it.

Pull up their case studies. Read them carefully. Ask what you are not seeing.

Before you start screening agencies, you should already know what kind of project you need. If you are still working through whether a full redesign is the right call or whether a refresh would do it, the redesign vs. refresh framework is worth reading first. Once that decision is made, the evaluation can actually begin.

Many agency portfolios show the finished site without showing you the business problem it was built to solve, who the buyers were, or how the site performed after launch. The best B2B case studies include those details. They connect the work to a business context: what was breaking, who needed to use the site and how, and what changed after launch. If a case study reads like a design showcase with metrics bolted on at the end, that tells you something about how the agency thinks about the work.

Look for evidence of fluency, not just familiarity. B2B fluency means they can talk about a buying committee, a multi-step evaluation process, or a sales team that uses the site as a tool in active deals. If their portfolio is heavy on consumer brands, e-commerce, or lifestyle companies (and B2B appears to be a side practice), that tells you something about where their expertise sits.

What to check before the first call

Their portfolio: Does it include companies with complex products, long sales cycles, or multi-stakeholder buyers? Or does B2B mean “not retail”?

Their case studies: Do they report measurable outcomes (traffic, conversion, pipeline, engagement) or do they describe the process and leave the results vague?

Their point of view: Do they publish content that shows how they think, or is their site a capabilities brochure? An agency that has worked through real B2B problems usually has opinions about them.

Their RFP response, if you issued one: Did they respond to your brief or reflect it back at you? Agencies that pitch well have clearly read what you sent. Agencies that pitch generically are filling a template.

The goal at this stage is to narrow the field to agencies that have earned a conversation. Three to five is a workable number. More than that and the evaluation becomes its own project.

Stage 2: What to actually listen for on the pitch call

The pitch call is where most buyers get misled, and most agencies know it.

A well-run agency pitch is a polished performance. The team in the room is confident, prepared, and fluent in your brief. The work looks good. The process sounds thorough. It is easy to walk out feeling like you have found the right partner.

Two problems with that feeling: the people pitching you are rarely the people who will build your site, and agreement is not the same as alignment.

Ask to meet the delivery team before you make a decision. Ask specifically for the project or client services lead who would own your account. The relationship you will actually have is with them, not with the business development team. If the agency cannot make that introduction during the evaluation, that tells you something about how they run their engagements.

On the B2B fluency question: listen for how they talk about your buyers. Do they reference your buying committee, your sales cycle, your measurement environment? Or do they speak about “your audience” in terms that could apply to any company?

If they cannot describe your technical and economic buyers the way you do, they will design for a generic B2B buyer who does not really exist.

The pitch question that reveals the most

There is one question that consistently separates agencies that have genuinely thought about your project from those that have not.

Ask it directly: “What would you NOT recommend we do, and why?”

A strong answer names something specific from your brief and explains why it is not the right use of your budget. Maybe the homepage redesign you have budgeted for is not your real problem. Maybe the AI feature you have been asked to include costs four times what you have allocated and will not move the needle for your buyers. Whatever it is, they give you something real to discuss and a clear reason to reconsider.

A weak answer sounds like this: “Everything in your brief looks good. We would just refine it as we get into discovery.” Or they call the question interesting without answering it.

“If they can’t say no during the pitch, when they’re trying to win your business, they won’t say no in week eight when it’s even harder.”

— Steve Ohanians, CEO, Clear Digital

An agency that has skin in the outcome pushes back when something is wrong. One that wants the contract will agree with you through the pitch and raise concerns later, when you are already committed and the change is expensive.

Other questions worth asking on the call

“Who specifically would work on our account, and what does their current capacity look like?”
Strong answer: names, roles, and an honest account of current load.
Weak answer: “We assign a dedicated team based on project needs.”

“How do you handle scope changes mid-project?”
Strong answer: a specific process: who identifies it, how it gets priced, how it gets communicated.
Weak answer: “We work collaboratively to address changes as they come up.”

“What happens when something goes off track?”
Strong answer: they name a specific scenario from a past project and explain what they did.
Weak answer: “We maintain open communication throughout.”

Red flags that do not show up until the proposal

A proposal is not just a price. It is a document that reveals how an agency thinks about your project: what they scoped, what they left out, and what assumptions they made to hold the number together.

Read it with that in mind.

Vague discovery scope. If the proposal describes discovery as “a series of stakeholder interviews and workshops” without naming who participates, what gets produced, or how long it runs, there is no shared definition of what discovery will accomplish. Scope disputes start here.

Good discovery is not a phase that happens before the real work. It is the foundation every subsequent decision is built on. One way to evaluate how seriously an agency takes this: ask whether their delivery team, not just the account team, is involved before scope is locked. At Clear Digital, we run a Velocity Workshop™ at the start of every engagement. The full project team on both sides gets in a room together to align on goals, requirements, and realities before a single wireframe is drawn. If an agency’s discovery process is something that happens to the client rather than with them, that distinction will show up in execution.

No mention of your MarTech environment. For a B2B company, the website does not exist in isolation from your CRM, your marketing automation platform, and your attribution model. If the proposal does not address how the site connects with your marketing automation system (whether that is HubSpot, Marketo, Pardot, or something else), how the CMS supports your demand gen team, or what the measurement plan looks like post-launch, you will discover in week ten that no one planned for any of it.

The MarTech fluency test

Ask these questions before you sign:

• How does your team approach marketing automation integration during the build?
• Who on your team works directly with the demand gen side of the house?
• How do you structure attribution for organic and paid channels in the CMS?
• What does your post-launch measurement plan typically include?

An agency with real B2B experience will have clear answers. An agency that treats MarTech as a handoff to the client will hedge.

Price-first comparison. Price is the easiest variable to rank across proposals, and it hides the most. Two proposals at similar price points can include radically different amounts of discovery, different levels of post-launch support, and different assumptions about who does what. The proposal with the lowest number has usually made the most assumptions about what you will figure out later.

AI claims without evidence.

If they cannot name one or describe the methodology, they are describing a future capability, not a current one.

Reference checks are not a formality

Most buyers treat references as a box to check. They call two names the agency provides, ask if they would recommend them, get two yeses, and move on.

That process tells you almost nothing.

Ask the agency for references from clients with similar complexity: similar product lines, similar stakeholder maps, similar scope. If they push back or cannot provide them, that is useful information.

When you get on the call, ask these instead:

“Describe a moment the agency pushed back on something you wanted to do.” An agency that pushed back at least once during the engagement was doing its job. If the reference cannot name a moment, the agency agreed with everything the client wanted. That is not a reference; it is a testimonial.

“How often did you hear from the project or client lead, and what did that communication actually include?” You are looking for a specific cadence and a real description of what got communicated. “They were very responsive” is not an answer.

“Did the agency define what success looked like at the start of the project?” If success was never defined at kickoff, there is no shared standard for what “done” means at delivery. Disputes about whether a project is complete almost always trace back to this moment.

“What would you do differently if you ran the evaluation again?” Clients who have been through the project have hindsight. Some of what they would do differently will point directly at the agency.

Why a B2B evaluation is different from any other agency search

Generic advice about choosing a web design agency (look at their portfolio, check their process, trust your gut) was written for a broad market. It does not account for what makes B2B website decisions different.

Your buyers are not individual consumers making a single decision. They are committees. Your sales cycle runs months, sometimes years. Your site needs to support a salesperson in an active deal, not just attract a visitor. Your measurement environment includes pipeline attribution, not just session counts. The agency you hire needs to understand all of that before they start designing a single page.

Most agencies that claim B2B expertise treat it as a segment of a broader practice. It is one row in a list of industries served. For Clear Digital, B2B is the only practice. Twenty-five years in Silicon Valley B2B means the frameworks we bring to a web design engagement were built from B2B data, not adapted from consumer work.

When Veritone came to us, their site was failing on speed and organic performance. The rebuild produced a 154% improvement in PageSpeed scores and an 84% increase in organic traffic. That did not come from running a good process. It came from understanding what their buyers needed the site to do and building for that.

The Splunk engagement addressed a more complex problem: a full website refresh aligned to new product architecture and brand messaging. The site needed to reorganize their product categories, create clearer user pathways, and do it across a multi-stakeholder organization where no two teams agreed on priorities. The result was a 59% increase in time on site, a 113% increase in form conversion, and an 84% reduction in drop-off rates. Cisco acquired Splunk for $28 billion shortly after. We are not claiming causality, but the site was not holding them back.

Sift needed a revitalized web experience that could support a repositioned brand. The UX work produced a 192% increase in conversion rates.

Those results came from agencies that understood what B2B buyers actually do when they land on a page.

If you are running an evaluation now and want a conversation with someone who has been on both sides of it, who has pitched and been pitched, we are straightforward to reach. Talk to us about your project →

Frequently asked questions

How many agencies should I include in an initial evaluation?

Three to five is the right range for most B2B companies. Fewer than three limits your frame of reference. More than five and the evaluation becomes time-consuming enough to distract your team from other work. The screen stage should bring you from a longer list down to three to five before the first call happens.

Should I issue a formal RFP or skip it?

A brief RFP can be useful for complex projects where you need consistent information across proposals. A full RFP with dozens of questions tends to favor agencies that are good at writing RFP responses, which is not the same as being good at building websites. If you go the RFP route, keep it focused on scope, business objectives, and the questions you need answered, not a comprehensive questionnaire.

How do I evaluate an agency’s AI capabilities without being misled?

Ask for specifics. Every agency will claim AI optimization capabilities right now; very few can point to a client where they improved a measurable AI search visibility metric. Ask them to show you what that looked like: which client, which metric, what changed, and over what time period. Vague answers about “preparing for the future of search” are not the same as a track record.

How do I know if an agency is genuinely B2B-focused?

Look at what they have actually built. B2B-focused agencies show portfolios with complex products, multi-stakeholder buyers, and long sales cycles. They write and speak about the B2B buying process with specificity. They ask questions during the pitch about your sales cycle and your measurement environment. If B2B appears to be one row in a list of industries they serve, it is probably a side practice.

What should be defined before a project kicks off?

At minimum: what success looks like at delivery (not just “a new website”), who owns decisions on the client side at each stage, how scope changes get handled, what the communication cadence looks like, and how the site will be measured post-launch. If any of these are not defined before the contract is signed, you are starting the project with assumptions on both sides. Assumptions become disputes.

What does a realistic B2B website project timeline look like?

Most B2B website redesign projects take three to six months, depending on scope, features, and content needs. That range assumes a structured process: discovery, UX design, development, and testing. Projects with more complex integrations, larger site maps, or more stakeholders involved in approvals will run toward the longer end. Any agency quoting a timeline significantly shorter than that is compressing a phase, usually discovery or QA, that tends to cost more to fix after launch than it would have cost to do properly.