Every agency has case studies. Most of them are unverifiable. A headline like “300% increase in organic traffic after our SEO-optimized redesign” sounds compelling until you ask: from what baseline, over what period, measured how, and could you speak to the client? Case studies are marketing documents written to impress — not to inform. That doesn’t make them useless, but it does make them something you need to read with specific questions in hand.
What Most Case Studies Leave Out
A case study that is designed to be persuasive rather than informative will consistently omit the following:
The baseline. A 300% traffic increase from 100 visitors per month is a gain of 300 visitors. A 30% traffic increase from 50,000 visitors per month is 15,000 new visitors. Without the starting point, percentage gains are meaningless. Always ask for absolute numbers alongside percentages.
The time frame. “Significant traffic growth” after a redesign sounds great. If it happened over 24 months with strong content investment during that period, the redesign’s contribution is unclear. If it happened in 90 days, that’s attributable. Ask for the specific window.
The contributing factors. Did traffic grow because of the new design, or because a new content strategy launched at the same time? Was the conversion rate improvement from a new checkout flow, or from a price change that happened to coincide with the redesign? Agencies rarely isolate variables in their case studies. You should ask them to.
What went wrong. A case study with zero friction is a story, not a report. Real projects encounter scope changes, technical constraints, client feedback that shifts direction, and unexpected platform behavior. An agency that can talk about what went sideways and how they handled it is more credible than one presenting a frictionless narrative.
Whether the client still uses the agency. Check if the relationship is ongoing. A long-term client who has continued to invest with an agency is a stronger endorsement than a one-off project presented as a success story.
The Anatomy of a Trustworthy Case Study
A case study worth trusting has these characteristics:
A specific, named client
Anonymous case studies — “a regional retail brand” or “a professional services firm” — are hard to verify from the page alone. The client could be anything, the numbers could be anything. Named clients with live sites you can test are the easiest to check.
Anonymized work is normal when clients ask for confidentiality, though. What matters then is whether the agency will walk you through the real work and its numbers on a call. If it won’t, ask why.
Before-and-after metrics with source data
Not “traffic grew significantly” — “organic sessions went from 4,200/month to 11,800/month over the six months following launch, per Google Analytics.” Specific numbers, specific data sources, specific time frames. If the agency can’t produce this, they either didn’t measure it or the numbers don’t support the story they want to tell.
A description of the actual work, not just the result
A good case study explains what was done. Not just “we redesigned the site” but “we migrated from Squarespace to custom WordPress, restructured the URL taxonomy across 80 service pages, and rebuilt the checkout flow to remove three friction points.” The specificity of the work description tells you whether the agency has genuine technical depth or is using impressive-sounding output language to paper over shallow execution.
Performance metrics alongside design metrics
Beautiful redesigns that don’t improve performance aren’t wins. Any case study for a web build should include PageSpeed scores before and after, if the claim involves performance improvement. Conversion rate changes should be included if the case study claims UX improvements. If the only metric is “the client loved the new look,” that’s a design testimonial, not a business case study.
Questions to Ask About Specific Case Studies
When you find a case study that looks relevant to your situation, ask:
“Can I speak with this client?” This is the fastest verification. A yes is a good sign — it means the agency is confident the client will corroborate the story. Hesitation or deflection is informative.
“What was the site’s PageSpeed score before and after?” Run the live site yourself at pagespeed.web.dev. If it’s significantly below 80 on mobile, ask what happened to performance since launch. “The client added plugins after delivery” is a reasonable answer. “We don’t track that” is not.
“What did the client invest overall — agency fees plus their time?” A $15,000 investment that generated $200,000 in additional revenue is a strong case study. A $15,000 investment that required 60 hours of client-side project management, multiple rounds of revision, and a delayed launch is a different story. The total cost — including the client’s time — matters.
“What specifically drove the result — was it the design, the SEO work, or something else?” Attribution matters. If the agency claims their redesign drove a 40% conversion lift, they should be able to explain the mechanism: what specifically changed, why that change drove the improvement, and how they know that factor (rather than another simultaneous change) was responsible.
Say you’re comparing two agencies based largely on case studies. Agency A shows an impressive headline number but can’t attribute the growth specifically to the site rebuild, and acknowledges that the client also launched a significant paid ads campaign in the same period. Agency B shows a smaller number, but walks you through the before-and-after checkout analytics and explains exactly which three UX changes produced the lift. Agency B’s case is the stronger evidence.
Red Flags in Case Study Presentations
Some patterns in case study presentation should make you pause.
All results are from very small baselines. Doubling traffic from 50 to 100 visitors is mathematically impressive but practically meaningless. If every case study involves growth from tiny starting points, the agency may not have experience with larger, more complex situations.
Every case study tells the exact same story. “Client had outdated site, we redesigned it, traffic grew, client happy.” If every case study follows the same arc with different names, the agency is using a template narrative rather than reporting on what actually happened.
Technical work is vague but results are specific. “We implemented advanced SEO strategies that boosted their visibility by 180%.” Advanced how? What specifically? Vague descriptions of work alongside specific numbers suggests the numbers may not be connected to the described work.
The portfolio links are broken or redirect to sites that look nothing like the case study. Test every live link in a portfolio. A case study for a “beautiful luxury brand website” that links to a generic-looking site running a page builder is a gap between the story and the reality.
Our studio has been running since 2016 — 200+ projects, all hand-coded, no page builders. When you ask about results, we walk you through exactly what we built and what happened afterward, on a call. That’s the standard case studies should be held to.
Using Case Studies as One Signal Among Several
Case studies are useful. They’re not sufficient. Used alongside live portfolio testing, reference calls, and contract review, they give you a more complete picture of an agency. Used in isolation, they give you the picture the agency wants you to have.
The most useful thing a case study can tell you isn’t whether an agency delivered good results for someone else. It’s whether the agency thinks rigorously about outcomes, communicates specifically about what they did, and holds themselves accountable to measurable results. Those qualities — visible in how they write about past work — predict how they’ll approach your project.
If you’re ready to evaluate what a transparent, well-documented agency engagement looks like in practice, talk to us or explore our custom WordPress development service.
Frequently Asked Questions
How many case studies should I review before making a hiring decision? Two to three strong, detailed case studies are more useful than ten surface-level ones. What you’re evaluating is depth of analysis and specificity of results — not volume. Focus your energy on the most relevant examples and test them thoroughly.
Should I only look at case studies from my industry? Industry experience helps with understanding your specific context, but it’s not a prerequisite for quality work. A technically strong agency that hasn’t built sites in your exact vertical can still deliver excellent results. More important than industry match is whether their past work solves problems similar to yours — complex navigation, e-commerce checkout, content-heavy architecture, performance optimization.
What if an agency has no formal case studies? Ask for client references instead, and request that you speak with clients who had projects similar to yours. A newer agency without a formal case study library can still be evaluated — it just requires more work on your end to gather the same information through direct conversations.
Are Clutch or Google reviews a substitute for case studies? They’re supplementary, not equivalent. Reviews tell you about general sentiment. Case studies (ideally verified) tell you about specific outcomes. Use reviews to validate that the general client experience is positive, and case studies to understand what the work actually produces.
Can I ask an agency to produce a custom case study format for evaluation? You can ask them to walk you through a past project in a structured way — what was the brief, what did they build, what were the results, what would they do differently. Whether they call it a case study doesn’t matter. You want the same information: specific scope, specific results, honest reflection.
What if a case study has impressive numbers but the live site underperforms? That’s valuable information. If an agency’s case study claims a PageSpeed improvement and the current live site scores 38 on mobile, something happened between delivery and now. Ask about it. The answer might be “the client installed Elementor themselves six months after launch,” which is fair. Or it might reveal that the performance claims in the case study were never real.