The 12-Question Audit Most Web Agencies Will Fail

Most business owners hiring a web agency in 2026 are choosing between three or four proposals that look almost identical on the surface. Same deliverables. Similar prices. Same vague language about strategy, design, and development. The owner picks based on gut feeling, prior relationship, or which proposal had nicer formatting. Six months later, the website is live and lead volume has not moved.

The proposals look identical because the agencies actually are similar. Most of them sell the same package with different branding. The way to separate the ones who will produce real outcomes from the ones who will produce invoices is to ask twelve specific questions during the sales process. Most agencies cannot answer them clearly. The ones that can are the ones worth hiring.

Print this list. Bring it to your next agency call. The agency that gets uncomfortable with these questions is telling you exactly what their work is going to look like.

How to use this list.

You do not need to ask every question in one call. Pick the five or six most relevant to your situation and work them into the discovery conversation. Listen for specificity. Vague answers are the signal. An agency that has actually shipped real outcomes can answer these in concrete numbers, named clients, and clear processes. An agency that has not will pivot to generalities and impressive-sounding adjectives.

Each question below has two sections. The good answer is what you want to hear, with examples of how a credible agency would actually phrase it. The red flag is the language to watch for. Both versions are pulled from real agency proposals we have reviewed for clients evaluating vendors.

Question 1: What is the closing rate of leads from your last five client websites?

This is the single most important question and the one almost no agency can answer. It separates agencies that measure outcomes from agencies that measure activity.

Good answer.

A specific number, attached to a specific client, with the caveat about how it was measured. Something like: "EM Landscape closes about 25 percent of qualified leads from the website into projects above $50K. Four Corners closes around 30 percent of LSA leads. Salt Creek Dental hits about 18 percent on new-patient inquiries from organic traffic. The numbers vary by category and price point."

Red flag.

A pivot to leads generated, traffic increases, or conversion rates without ever mentioning closed deals. If they cannot tell you how much of the website-generated lead volume their last clients actually closed into revenue, they were never tracking it. Which means the work was being optimized for something other than your revenue.

We covered the specific tracking layer that produces this data in our writeup on how a landscape company closed $500K from less than $30K in ad spend. Without a closed-deal feedback loop, every other optimization is guesswork.

Question 2: Can I see the technical performance of websites you launched twelve months ago?

Anyone can ship a fast website on day one. The harder test is whether it stays fast after a year of content updates, plugin additions, and the slow accretion of marketing tags. Ask to see Core Web Vitals on real production sites that have been live for a year.

Good answer.

They send you live URLs of three or four real client sites that launched 12 to 18 months ago. You run PageSpeed Insights yourself and see LCP under 2.5 seconds, INP under 200 milliseconds, and CLS under 0.1 on mobile. The numbers are still green.

Red flag.

They show you Lighthouse scores from staging environments, brand-new launches, or PDF screenshots from a sales deck. They cannot or will not name live client URLs you can audit yourself. They blame any performance issues on "client edits after launch" rather than building a system that holds up to normal use.

Question 3: What schema markup do you implement, and on which page types?

Structured data is the single most leveraged technical SEO investment in 2026. It is also the one most agencies skip because their template-driven workflow does not include it. Ask specifically about JSON-LD coverage.

Good answer.

They name specific schema types and the page types they apply to. Organization or LocalBusiness on every page. Service on every service page. FAQPage on every page with FAQs. Article or BlogPosting on every post. BreadcrumbList everywhere. Review or AggregateRating where applicable. They can show you a live page with proper schema by viewing the source.

Red flag.

They say "we handle SEO" without specifics. Or they list Yoast or RankMath as their SEO solution, which suggests they are running on WordPress with default plugin output rather than building schema intentionally. Or they cannot define what JSON-LD is. None of those agencies will get your site cited in AI search results.

The depth of schema coverage that drove Salt Creek Dental from zero to 52,950 percent organic traffic growth in one year is the bar to compare against. Most dental marketing agencies cannot match it.

Question 4: How do you measure and report on what is working?

The reporting an agency provides tells you exactly what they are optimizing for. Agencies report on the metrics they can move. If they report on impressions and clicks, they are running campaigns to maximize impressions and clicks. If they report on closed revenue, they are running campaigns to maximize closed revenue.

Good answer.

A monthly report that connects channel activity to qualified leads to closed deals. The report names specific campaigns or content pieces that produced revenue. It includes a section on what changed last month and what will change next month. The agency walks through it with you on a real call, not just an emailed PDF.

Red flag.

Reports full of impressions, click-through rates, "engagement," and rankings without any connection to revenue. PDFs that look impressive but require you to trust the agency on whether anything is actually working. "We are seeing positive momentum" without numbers attached.

Question 5: Who actually writes the content?

Most agencies outsource content production to freelance writers or, increasingly, to AI tools. The output is generic content that reads like it was written by someone who has never met your customers. It does not rank well, it does not convert, and it makes your business sound like every competitor in your category.

Good answer.

They name the people doing the work. The senior strategist who interviews subject matter experts. The writers who specialize in your category. The editorial process that catches generic language before it ships. They show you content samples from real client engagements where the voice is clearly differentiated.

Red flag.

Vague language about "our content team" without names. Samples that read like every other agency blog. Reluctance to specify whether they use AI to write drafts. Any agency selling 8-10 blog posts per month for $1,500 is using AI without telling you, because the math does not work otherwise.

Question 6: What happens if the website needs to be rebuilt in three years?

The question is really about lock-in. Some agencies build websites that only they can maintain, on platforms that make migration expensive. Others build on standard technology any competent team can take over. The difference matters for your long-term cost of ownership.

Good answer.

They build on standard, well-documented technology (Next.js, SvelteKit, Sanity, Vercel, similar). They commit to giving you full repository access and admin credentials. They specify in writing that the code is yours. They have a documented handoff process for clients who move to a different agency.

Red flag.

They built it on a proprietary platform only they support. They retain hosting in their own account. They will not release admin credentials without a fee. They use language like "fully managed solution" that means you cannot leave without rebuilding from scratch. This is the most expensive trap in the agency-client relationship.

Question 7: How will you preserve our existing SEO during the rebuild?

If you have any organic traffic at all on your current site, the way the rebuild handles redirects, content migration, and structured data will determine whether you keep that traffic or lose 30 to 60 percent of it in the first 90 days post-launch.

Good answer.

They describe a specific process. Full URL inventory crawl before the rebuild starts. A redirect map with every old URL mapping to a new URL. Pre-launch redirect testing to confirm 301s are correct. Sitemap submission to Google Search Console on launch day. A 90-day monitoring window where they track Search Console performance and fix issues fast.

Red flag.

No mention of redirects. Vague language about "SEO best practices." Promises that the new site will rank higher without a plan for preserving existing rankings. Any agency that has not done a real migration before is going to learn on your traffic.

The specific playbook we use for migrations is detailed in our writeup on why your website is not generating leads. The technical foundation work is non-negotiable if the rebuild is supposed to lift conversion.

Question 8: What does a typical optimization look like in month four?

The first three months of an agency engagement are usually heavily structured. Discovery, design, build, launch. Most agencies have that part figured out. The harder question is what they actually do in month four, five, and six, when the website is live and the work is supposed to compound.

Good answer.

They describe specific recurring activities. Weekly review of Search Console performance. Monthly content publication on a defined editorial calendar. Quarterly conversion rate testing on key pages. Ongoing ad spend optimization with specific keyword-level cuts and additions. They name the specific tools they use to do this and the cadence at which decisions get made.

Red flag.

The retainer covers "ongoing maintenance and updates" without specifics. They will publish "monthly content" but cannot tell you what it will be about or how it will be selected. They charge $3,000 to $5,000 per month for an undefined activity set. This is how most agency engagements coast for months without producing measurable change.

Question 9: Can I talk to one of your current clients without you on the call?

Reference calls with the agency on the line tell you nothing. The references will say what the agency wants them to say. The useful question is whether the agency will let you talk to a current client alone.

Good answer.

They volunteer two or three current clients you can reach directly. They send the introduction by email and stay out of the conversation. The references know they were going to be contacted and have agreed in advance to speak candidly.

Red flag.

They offer references but only on calls they will join. They give you contact information for clients who left the agency a year ago, not current ones. They cannot provide references in your specific category or revenue range. They get defensive when you ask.

Question 10: What happens when something breaks at 11 pm on a Friday?

Websites break. Forms stop sending. Payment integrations fail. Servers go down. The question is whether the agency will fix it fast or whether you will wait until Monday morning to even hear back.

Good answer.

They describe a specific support process. A monitoring system that catches problems before clients notice. A clear escalation path for emergencies. A response time commitment in writing for production issues. They name the actual people on call and how to reach them.

Red flag.

Support is "business hours, Monday to Friday." There is no clear protocol for emergencies. The contract has language about response time but no commitment to resolution time. The support person is a junior account manager who will need to escalate to engineers who you never get to talk to directly.

Question 11: How do you decide what to prioritize when there is more work than budget?

Every engagement runs into trade-offs. The question is how the agency makes decisions when something has to give. The answer reveals whether they are optimizing for your outcomes or their convenience.

Good answer.

They describe a framework based on impact on closed revenue. Specific examples of trade-offs they made on past projects, with the reasoning written down. A willingness to push back on you when you ask for something that will not move the metric they were hired to move. An honest acknowledgment that prioritization is harder than it sounds and they will not always get it right.

Red flag.

They commit to everything in the proposal without discussion of trade-offs. Or they accept every client request without questioning whether it serves the underlying goal. Or they describe "agile delivery" without explaining what gets cut when the calendar slips. Agencies that cannot say no to features end up building expensive websites that do nothing useful.

Question 12: What would you do differently if we hired you a year from now instead of today?

This is the question that gets to whether the agency thinks about your business strategically or just executes briefs. The good answer reveals their actual perspective on your market, your competitors, and your readiness for the work.

Good answer.

They have a specific point of view. Something like: "If you waited a year, I would push you harder on positioning before the rebuild. Your current positioning is unclear and we would be designing a site for a buyer who does not yet exist clearly. We could either fix that in the first phase of the engagement or you could spend 60 days on it before we start." A real perspective, not flattery.

Red flag.

Generic answers about "the market changing" or "AI evolving" without specifics about your business. Reluctance to give you any feedback that might disqualify them from the engagement. They are selling you their package, not advising you on your business.

How most agencies will respond to this list.

If you send this question list to three agencies and watch how they respond, you will get a clear ranking. The first agency will be annoyed and tell you they prefer to walk through their standard process. The second will answer four or five questions clearly and pivot to generalities on the rest. The third will engage with the whole list and tell you which questions they have strong answers on and which ones they are still working on.

That third agency is the one to hire. Not because they are perfect, but because they are honest. The agencies that pretend they have everything figured out are usually the ones who have figured nothing out beyond their proposal template.

The agencies who got uncomfortable with our question list were the ones who could not produce closed revenue. The agency that engaged with every question is the one we hired. That happened to be ComCreate.

A few more questions worth asking depending on your category.

If you are a home services business.

• Do you have specific experience with Local Service Ads, including license screening and review collection? (LSAs are the single highest-leverage paid channel for most contractors and most agencies do them badly.)
• How do you handle call tracking and the attribution back to specific keywords or campaigns? (Without this, you cannot tell which ads produce closed jobs.)
• What is your process for managing seasonality? (Most home services categories have 2x to 4x demand swings between peak and off-season.)

The specific LSA setup that drove Four Corners Concrete to $350K in 90 days is documented in our writeup on the Local Service Ads playbook for contractors.

If you are a healthcare or dental practice.

• Do you have experience with HIPAA-aware data handling for any form data that includes health information?
• How do you handle the structured data specific to medical and dental practices (Dentist, MedicalBusiness, Physician schema types)?
• Have you worked with practices that had to navigate Google Business Profile policy changes specific to medical practices?

If you are a B2B SaaS or technology company.

• What is your experience with structured data for software and product pages?
• How do you handle freemium funnels, gated content, and trial sign-up tracking?
• Do you integrate with HubSpot, Salesforce, or other CRM platforms in a way that maps website behavior to deal stages?

If you are an e-commerce business.

• Do you have specific experience with Shopify Hydrogen, headless commerce, or whatever platform your stack runs on?
• How do you handle product schema, review schema, and the technical SEO specific to product pages at scale?
• What is your approach to checkout optimization and reducing cart abandonment?

What this list does not cover.

The list above focuses on the agency's process and accountability. It does not cover questions specific to brand and design, which matter and are worth asking but are harder to evaluate objectively. It does not cover legal questions about contracts, IP, and termination terms, which are worth running past your lawyer. It does not cover questions about pricing structure, which we have a separate post on already and which depend heavily on your category and project.

The list also assumes you have already done your own internal diagnostic work. If you have not figured out what your current website is doing wrong, no agency will be able to fix it for you. The agencies you talk to should be able to help with that diagnostic during the discovery phase, but they cannot substitute for your own understanding of where your business is and where it needs to go.

For the diagnostic side of the conversation, our writeup on why your website is not generating leads covers the seven specific issues most service-business sites have and how to identify which ones are yours.

Frequently asked questions.

How many agencies should I evaluate before hiring?

Three is usually enough. Two if your category is narrow and one of them is clearly the right fit. More than four becomes a coordination problem and slows the decision. The point of evaluating multiple agencies is not to find the cheapest one, it is to triangulate on what good actually looks like in your category.

Should I ask for a guarantee of results?

No, and you should be skeptical of any agency that gives one. Marketing outcomes depend on many factors outside the agency's control (your operations, your sales process, market conditions, competitor activity). A credible agency will commit to a process and a measurement framework but not to a specific revenue number. An agency that promises specific results is either lying or planning to claim credit for trends that would have happened anyway.

What if the agency I want to hire fails some of these questions?

Use the failed questions to negotiate. If they cannot answer Question 8 about month-four optimization, ask for that specific accountability to be added to the contract. If they fail Question 6 about lock-in, push for explicit code ownership and credential transfer terms. The list is a discovery tool, not a pass-fail test. The point is to surface the gaps so you can address them in the contract.

What if my budget is small and I can only afford agencies that will fail most of these questions?

Then your real question is whether to hire an agency at all. Below a certain budget, you are better off hiring a senior freelancer who can answer most of these questions personally, or building in-house, or waiting until your business can support a real agency engagement. The cost of hiring the wrong agency at any price point is higher than the cost of delaying the engagement.

Can I send this list to an agency before our first call?

You can, and the response will be informative. An agency that engages enthusiastically with the list is probably worth talking to. An agency that gets defensive or asks to walk through their standard process first is showing you exactly how they handle accountability. Either response is useful information.

Will the agency be offended if I bring this list to the call?

A good agency will be impressed that you have done the work to come prepared. A bad agency will be offended. The reaction is itself a useful filter.

How much should I expect to invest in the agency relationship?

For service businesses with revenue between $500K and $10M, credible agency engagements run $15K to $60K for the initial build and $4K to $12K per month for ongoing optimization. Below that range you are paying for a junior team or a template-driven shop. Above it you are paying for enterprise overhead that is rarely justified at small business scale. The specific number depends on category, project complexity, and the depth of the ongoing work.

What to do with the list.

Print it. Bring it to your next agency call. Watch which questions make the agency leader pause and which ones make them light up. The pattern in their answers will tell you, in 30 minutes, what 6 months of working with them is going to feel like.

Most owners pick an agency based on the proposal that looks best. The owners who pick well pick based on the conversations that get specific the fastest. This list is the fastest way we know to force the conversations to get specific.

If you want to bring this list to a call with us, we run discovery conversations for service businesses that are designed to engage with exactly these kinds of questions. The first conversation is thirty minutes. No proposal pressure. We answer the list openly, including the questions where our answer is "we are still figuring that out." The owners who hire us are the ones who watched us answer the hard questions without flinching.