Why screenshots alone prove nothing
A ranking screenshot is the easiest thing in marketing to fake, crop, or cherry-pick. It shows one keyword, one moment, one search location, with no way to check the traffic that came with it, or whether the client's phone actually rang. An agency can screenshot a ranking for "gutter repair 32801" the day it happened to peak, then never mention it fell back to page three a month later.
Case study pages are marketing collateral. They're written to sell you, not to inform you, which isn't automatically dishonest but it does mean the agency chose which numbers to show and which to leave out. A single before/after graph with no axis labels, no date range, and no named client tells you the agency is comfortable with vague. That's a signal on its own.
The same problem shows up in stock "before and after" website mockups. A slick redesign comparison proves a designer can make a page look better. It proves nothing about whether that new page ranks, loads fast, or converts a visitor into a phone call. Plenty of agencies are strong at design and weak at the technical and content work that actually moves rankings, and a pretty screenshot is exactly how that gap gets hidden from a buyer who hasn't asked the right questions yet.
What actually proves an agency did the work is data you can check independently: the client's business name so you can pull up their current map pack position yourself, a shared Search Console property (even view-only) showing real click and impression trends over months, or GA4 showing lead-form and call-tracking numbers tied to specific campaign dates. None of that is hard for a legitimate agency to hand over. If a rep gets cagey when you ask for the client's name and city, that's the tell, not the screenshot.
- Ask which keywords the screenshot is for and whether they're branded, low-competition, or the actual money terms your business needs.
- Ask for the date range: a screenshot with no timeframe could be from three years ago.
- Ask whether the number is rank position, or clicks and leads. Rank without traffic doesn't pay bills.
The verification checklist: what to actually ask for
Before you sign anything, put the agency's sales rep on the spot with specifics. A legitimate shop answers these without flinching. A shop running on smoke stalls, redirects you to a generic FAQ page, or says the client data is "confidential" when you're only asking for a name and city, not their financials.
| Ask for | What it proves | Red flag if refused |
|---|---|---|
| Client business name + city | You can independently check current map pack and organic position | "We can't share that" for a public-facing ranking claim |
| Search Console or GA4 screen-share or read access | Real traffic trend, not a single cropped moment | Only offers a static PDF or slide deck |
| Start date of the engagement | Lets you judge the timeline against realistic ranges | Vague answers like "a few months" |
| Whether the number is organic or paid | Paid rankings vanish the day the ad budget stops; organic compounds | Agency conflates the two or dodges the question |
| A live call-tracking or lead-count number, not just impressions | Impressions and clicks don't always turn into a ringing phone | Only shows top-of-funnel metrics, never leads or calls |
None of this is an unreasonable ask. A firm that's actually done the work for a real plumbing or roofing client has this data sitting in a dashboard already. Handing it over costs them nothing. Refusing costs you a bad hire.
What realistic contractor marketing timelines actually look like
This is where most fabricated case studies fall apart, once you know what to expect. Google Maps and local pack movement can show up in 30-90 days for a business with a clean, complete Google Business Profile in a moderately competitive metro. Organic rankings for the terms that actually carry volume, things like "[trade] + [city]" or "[trade] near me" in a metro with active competition, typically build over 4-9 months of consistent work: content, technical fixes, citations, and link building compounding together.
A roofer in a mid-size metro competing against three franchise operations and a dozen local crews is not going to rank for "roof replacement [city]" in three weeks. If an agency's case study implies that, one of three things is happening: the keyword shown has almost no competition (long-tail, low-volume, doesn't move the revenue needle), the win is paid traffic mislabeled as organic SEO, or the timeline in the case study is quietly longer than what's advertised in the pitch.
Trade matters here too. A niche trade in a smaller metro (say, a marine lift installer covering three coastal counties) can see faster local pack movement than a roofer competing in a metro with heavy franchise ad spend, simply because there's less competition to out-rank. An honest agency adjusts the timeline talk to your actual trade and market, not a one-size-fits-all promise.
- 30-90 days: Google Business Profile optimization, review velocity, map pack movement for less competitive terms.
- 4-9 months: organic rankings for competitive service-area keywords, with the pages and links built to support them.
- Ongoing: rankings need maintenance. An agency that treats SEO as a one-time project rather than a monthly engagement is setting you up for a slow fade six months after they stop.
Anyone promising page-one placement in two or three weeks for a genuinely competitive term is either selling you paid ads under an SEO label or setting up a bait-and-switch you'll discover after the deposit clears.
Red flags in how an agency presents its own results
Beyond the individual claim, watch the pattern. A single sketchy stat might be a one-off exaggeration. A pattern of vague, unverifiable, or unfalsifiable claims across an entire pitch deck tells you how the agency operates day to day.
- No client names, ever, on any case study. NDAs are real and some clients don't want to be named publicly, but a firm with dozens of case studies and zero named, checkable examples across all of them is worth a second look. A reasonable shop can point to at least some named local clients you can call.
- Percentage-only claims with no baseline. "300% more leads" from a business that had one lead a month before is a different story than one that had thirty. Ask for the actual before and after numbers, not just the multiplier.
- Stock testimonial language. Watch for reviews that read like ad copy rather than something an actual contractor would say. Real client quotes reference specific work: a rebuilt site, a service-area page, a ranking that moved.
- Long-term contract required to "prove it works." Legitimate agencies don't need a 12-month lock-in to demonstrate results; they show you the trend line and let the numbers make the case month to month.
- Every review is 5 stars with no negative or neutral feedback anywhere. That's a sign of review gating or filtering, not a spotless track record.
- Guarantees of a specific rank position or a specific lead count. Nobody controls Google's algorithm outside of Google. An agency guaranteeing "page one in 90 days or your money back" is either quietly targeting a keyword nobody searches for, or setting up a refund clause they've already priced into what they charge everyone.
None of these alone is disqualifying. Together, in the same pitch, they add up to a pattern of an agency selling a story instead of selling verified work. Trust the pattern over any single data point, good or bad.
Questions to ask on the call, in order
You don't need a forensic audit to protect yourself here. A short, direct line of questioning on the first call separates a shop that does the work from one that's mostly selling the pitch.
- "Can you give me the name and city of a current or recent client in my trade, so I can check their rankings myself?"
- "Would you be willing to give me read-only access to Search Console or GA4 for that client, or a screen-share walkthrough?"
- "Is that ranking or traffic number organic, or does it include paid ad spend?"
- "What was the starting point before you took over? What did rankings and lead volume look like on day one?"
- "What's the realistic timeline for my specific trade and market, and what would make that longer or shorter?"
- "What happens to my rankings if I cancel? Do they hold, or fade?"
- "Who else on your team has worked a client in my trade, and what's specific about how you approach it?"
That last question matters more than it sounds. A generic pitch deck built for "contractors" broadly, with no sense of how a roofer's sales cycle differs from an electrician's or how a landscaping company's seasonality changes the reporting rhythm, is a sign the agency is running the same script on everyone. A shop that's actually worked your trade will have specifics: which keywords carry the most call volume, how emergency-service terms behave differently than planned-project terms, what a slow season looks like in the data so you don't panic over a normal dip.
Listen for specificity in the answers. "It varies" is a fine answer to question five if it's followed by an actual range and the factors that move it. "It varies" as the answer to every question is a dodge.
A shop confident in its own work will walk you through this without hesitation, because the underlying data already exists. A shop that's mostly running on testimonials and screenshots will get slippery, change the subject to the pitch deck, or push you toward signing before you can dig further. That pressure to skip the questions is itself the biggest red flag on this whole list.
What honest reporting looks like once you're a client
Vetting doesn't stop at the sales call. Once you're signed, the reporting an agency sends you monthly should look like a continuation of the same transparency you asked for up front, not a reversion to vague screenshots.
A straight-shooting monthly report includes actual ranking positions for the keywords that matter to your business (not a handful of easy long-tail terms picked to look good), organic traffic trend over time from Search Console or GA4, map pack position across the service areas you actually cover, and lead or call counts tied to the campaign, not just top-of-funnel impressions. It should also be honest about the months where movement stalled. Real SEO has flat months and even down months around algorithm updates. An agency whose reports show a perfect, uninterrupted upward line every single month for years is smoothing the story, not reporting it straight.
You should also be able to walk without a fight. Contracts with 12-month lock-ins on services that are inherently month-to-month (SEO, local SEO, AI search work) are a structural way to hide underperformance behind a term you can't exit. Fair engagements let you leave if the reporting isn't backing up the invoice.
The bar here isn't perfection. It's whether the agency treats you like someone who deserves the real numbers, both the good months and the flat ones, instead of a highlight reel that only shows the wins.
Vetting a web build vs. vetting an SEO engagement
The verification process changes depending on what you're actually buying, and conflating the two is where a lot of contractors get burned. A custom website is a deliverable you can inspect the day it launches: load speed, mobile behavior, whether the contact form actually routes to a real inbox. SEO, local SEO, and AI search visibility are ongoing processes measured in trend lines, not a single finished product, which makes them easier to fake in a sales pitch and harder to verify on the spot.
For a web build, ask to see three to five live sites the agency actually built, not just designed, and check them yourself: load time under two seconds is a reasonable bar on current hosting infrastructure, mobile layout shouldn't break or require pinch-zooming, and the site should be readable without a marketing background. Ask who owns the code and the domain once the project is done. An agency that builds you a site you can't take with you if you leave is a structural lock-in dressed up as convenience.
For SEO and local SEO, the verification standard is everything covered above: named clients, shared analytics access, and timelines that match the 30-90 day and 4-9 month ranges that hold across the industry. The two are related but distinct disciplines. A shop that builds a beautiful site with zero organic visibility strategy behind it, or one that promises rankings on a site with technical problems no one's fixed, is selling you half the job. Ask how the web build and the SEO work connect, specifically whether the site architecture (service pages, service-area pages, schema markup) is built to support the rankings they're claiming they'll get you.
- Web build: inspect the finished product directly, ask about code and domain ownership.
- SEO / local SEO / AI search: verify with shared data access and realistic timeline ranges, since there's no single finished artifact to inspect.
- Ask how the two connect: a fast, well-built site with no content or authority strategy behind it won't rank on its own.