You’re in the quarterly review. Every chart is green, every arrow up, and your booked jobs are flat. You’re not the first owner in that chair.
The graph is always green because the ground beneath it has gotten cheaper to fill. Content has never been cheaper to publish, tracked keyword lists have never been longer, and AI Overviews have changed how impressions are counted. The whole ecosystem now manufactures up-and-to-the-right by default. Your agency doesn’t have to lie to send you a report that looks like growth.
What follows is how to read an SEO report critically: why it looks like growth even when nothing did, what a deeper read of the same data actually shows, and the five questions to bring to your next agency call.
Up and to the right is the default
Most sites trend up over time, even ones doing little or no active SEO work. Three forces push the line up automatically. None of them require your agency to have done anything.
The first is site age. A trades website that’s been online for three years has more pages indexed than the same site at year one. It has internal links connecting those pages. It ranks for hundreds of long-tail variations of queries it was never built to target. Aging takes care of itself.
The second is brand search. Even if you’re not running marketing campaigns, you’re branding. Wraps on the trucks, yard signs at the job, business cards in the kitchen drawer, logos on the uniforms, door hangers, equipment stickers, the sign you bought on the kids’ baseball outfield. All of it pushes your company name into someone’s head, and that name comes back out as a Google search when their AC dies.
Your existing customers do the same thing because they don’t have your phone number memorized. Both kinds of brand search land in your organic numbers. Your SEO line goes up because the trucks are on the road and the customer base is growing, not because the SEO agency did anything.
The third is the impression inflation that came in with AI Overviews. Google now pulls citations into the AI summary block at the top of a result. According to Google’s own documentation on impression counting in Search Console, being cited as a visible source inside that block counts as an impression even when the user reads the AI’s answer and never scrolls to your link. Most trades site owners noticed their impression counts went up sometime in 2024 or 2025 while their click numbers stayed roughly flat. The impressions are real. The exposure they represent is not what it used to be. We covered the broader shift in our piece on AEO and SEO, because the same forces show up in the AI visibility pitch arriving in your DMs.
Three forces, none of which require any work to produce, all of which manufacture growth that doesn’t produce booked jobs.
The four moves that dress up the report
Beyond the natural drift upward, there are four specific things agencies do that make the report look better than the business is doing. None are technically dishonest. All are designed to be nodded at, not questioned.
The first is tracked keyword list inflation. The agency tells the rank tracker which keywords to monitor, the tool reports how those keywords rank. Adding keywords to the tracker is sometimes strategic, when the business is pursuing a new market or service line, and sometimes clerical, when the report needs to show movement. If your agency adds two hundred new keywords this quarter and 120 of them turn out to already be ranking on page two or three, the next report shows “120 new keywords are now ranking.” That looks like momentum. It isn’t. The keywords were always ranking. The honest cut tells you which keywords were added for strategy and which were added for the row count.
The second is targeting long-tail keywords that nobody searches. Ahrefs ran a study comparing keyword volume estimates to actual Search Console impressions and found that zero-volume keywords get roughly the volume the tools predict, which is to say, near zero. Agencies still target them, because ranking on a no-volume keyword is easy and produces a report that says “top three ranking for thirty new keywords.” The report shows the rank. It does not show the traffic, because there is no traffic.
The third is geogrid greenwashing on those same niche queries. The geogrid is the map of red and green dots showing how your business ranks at different physical locations around your service area. Red means low rank, green means high rank. A geogrid that went from mostly red last quarter to mostly green this quarter looks like momentum. Then you read the keyword the geogrid was measuring. “Ac repair maintenance company near you.” Half of that string is filler. It has no monthly search volume. The geogrid measures rank, not search volume, so the green is technically accurate and operationally meaningless.
The fourth is aggregate organic traffic without service area segmentation. “Organic traffic up 18% YoY” sounds great in a slide. Then nobody asks where the traffic came from. A plumber in Albany doesn’t get paid by DIY readers in Phoenix, or by bot farms in the Philippines. Most trades agencies report total organic traffic because the unfiltered number is bigger than the filtered version. Filtered to the service area, the same report might show flat or down.
The brand search problem and what you can actually separate
Brand search and non-brand search are two different products. Telling them apart is the difference between knowing whether your SEO investment is working and knowing whether you’re paying SEO for traffic the wraps, the yard signs, and the existing customer base already produced.
Brand searches land in your organic numbers. All of them. The new customer who saw your truck this morning, the existing customer who lost your number, the referral who looked you up before calling. When your agency reports total organic traffic and total organic conversions, they’re taking credit for every dollar of branding the business has ever done. This is the single most common way SEO performance is overstated for trades businesses.
Now the technical part. Brand versus non-brand at the traffic level is fully separable today. Search Console has a query report that shows every term that drove a click. Any agency can filter that report by your company name, exclude the matches, and produce two numbers: brand organic clicks, non-brand organic clicks. Two columns. Five minutes of work. Most agencies don’t show you these two columns because the standard template doesn’t include them. The unfiltered number is bigger and easier to lead with.
Brand versus non-brand at the revenue level is a different problem, and the limitation is structural. In 2011, Google stopped passing the search term in the referrer for organic clicks. When a user clicks your organic result, the session that lands on your site doesn’t carry the query that brought them there. Search Console knows the query because it watched the click happen on Google’s side. GA4 doesn’t, because the user got to your site without it.
You can know which queries drove clicks. You can know which pages drove revenue. You cannot tie a specific query to a specific dollar through organic search. That has been true for fifteen years and is not changing.
The honest workaround is page-level revenue. If your site is built with a clean service-by-city page structure (one page per service per service area: /ac-repair/springfield, /furnace-repair/dayton, and so on) page-level conversion is a directional proxy for the queries that fed it. The /ac-repair/springfield page got 47 clicks and produced four booked jobs at an average ticket of $2,300. That’s enough to know whether the page is doing work. Most trades sites don’t have this structure, which is itself a finding worth reporting.
The ServiceTitan / GA4 integration sits in this same territory. The integration documentation describes what it does honestly: when properly set up, it pushes revenue data from booked jobs back into GA4 against the source and the landing page. Organic search drove X dollars this month. The /ac-repair/springfield page drove X dollars this month. Real value, and it’s the closest thing to honest organic revenue attribution available today.
What it does not do, and cannot do, is tie revenue back to the specific search query that started the visit. The query was stripped at the front door before ServiceTitan ever saw the visitor. That’s a Google limitation, not a ServiceTitan one. Anyone selling you query-level organic revenue without explaining the model behind it is selling you a model as if it were a measurement.
What a CMO-level read of an SEO report looks like
Everything above can be technically honest. The keywords are ranking. The traffic is up. The geogrid did go green. Honesty isn’t the issue. Depth is.
An experienced reader of these reports asks for more segmentation than the standard report offers, and tracks one thing as the leading indicator: non-brand short-tail high-intent local traffic compounding over time. Traffic compounds naturally with or without effort. The specific kind that produces booked jobs does not. That’s the line on the chart that has to be going up, and that’s the line nobody puts in the standard report unless asked.
The cuts that produce that view:
Non-brand organic traffic, segmented by your service area. Out-of-market clicks don’t produce service calls. Neither do DIY readers in Phoenix or bot farms in the Philippines.
Map pack impressions and calls, separate from website performance. For most trades businesses, Google Business Profile drives more inbound contacts than the website itself. BrightLocal’s consumer search behavior research found that one in five consumers conduct local searches directly inside maps. A report that doesn’t separate map pack performance from website performance is hiding the bigger half of the picture.
Rankings on short-tail, high-intent local queries. “Ac repair springfield.” “Emergency plumber albany.” Not the overspecified strings that produced the green geogrid. The leading indicator is more of those queries ranking, and ranking higher, over time.
Page-level conversion for the pages that actually convert. Service-by-city pages should be reported one by one with their click counts and their booked jobs. Aggregate organic conversion hides which pages produced and which didn’t.
Share of local voice against your three real competitors. Name them. Three plumbers in town own the local SERP. Where do you rank against them on the queries that matter, by service area, by month?
An explicit acknowledgment of what the data cannot tell you. Query-level organic revenue. Pre-GBP-attribution conversions. Anything that requires modeling rather than measurement. The CMO-level report names the limit out loud. The standard report papers over it and hopes you don’t notice.
Be your own SEO advocate
None of the filters above apply themselves. Search Console doesn’t separate brand from non-brand by default. GA4 doesn’t surface page-level conversion by default. The map pack doesn’t separate from the website by default. Every honest version of every report sits one filter away.
Every quarter you accept the unfiltered version is a quarter you’re paying for SEO without knowing what SEO is producing. The advocacy isn’t adversarial. It’s the same standard you apply to every other line in your P&L. You don’t accept “trucks did 18% more miles this quarter” without asking which trucks, which routes, and how much revenue those miles produced. Don’t accept it on organic traffic either.
Five questions force the honest version into the open at your next agency call.
- Show me my non-brand organic traffic for my service area only. Brand-filtered, geo-filtered, last twelve months.
- Show me page-level conversion for my service pages, not aggregate. Which pages produced booked jobs and which ones didn’t?
- How many of the keywords on this report were added to the tracker in the last ninety days versus already tracked twelve months ago?
- What does map pack performance look like, impressions, calls, direction requests, separate from the website?
- Where do I rank against [your three real competitors by name] on these queries, by service area: ac repair [city], emergency plumber [city], hvac service [city]?
An agency that has the filters set up and hands them over without flinching has nothing to hide. An agency that needs two weeks and a follow-up email to produce them has been counting on you not to ask. The asking, not the answering, is what shifts the relationship.
Walk into your next quarterly review with these five questions and watch what happens. The room shifts. The deck stops landing. Someone needs a week to follow up.
The reports were always green. The filters were always one click away. Whether you’d ever ask was the only question that mattered.
