How to read your marketing report in 2026: nine metrics and five questions
A Decatur family practice's Google numbers for July 23 to October 14, 2026 could be written up two ways. One says: "Your visibility is up 36%." The other says: "Your clicks are down 2%." Both are true. Google showed the practice 36,377 times, against 26,782 in the same dates of 2025. People clicked through to the website 2,120 times, against 2,154. (The 2026 window has 81 days of Search Console data against 84 in 2025, so the comparison is slightly conservative.)
Which sentence a report leads with tells you a great deal about the report. In 2026 the numbers in a marketing report no longer move together the way they used to, and an owner who reads only the biggest number can be misled in either direction.
This guide is for owners who receive a report from an agency, an SEO tool or a website platform and want to know what to trust. It adapts the "how to read your own numbers" chapter of our Search Everywhere Field Study, an observational study of three DIWYM clients. We wrote it with our own reports in mind. If any of these checks make our work look worse, good: you should know.
Why reports are harder to read in 2026
Customers now find, judge and call local businesses without visiting the website. Pew Research Center (July 22, 2025; 900 U.S. adults, March 2025 browsing) found people clicked a traditional result on 8% of visits when Google showed an AI summary, against 15% when it did not. SparkToro's 2024 study (Datos panel, September 2022 to May 2024, U.S.) found that 58.5% of U.S. Google searches end without a click to the open web.
So a business can be seen more and clicked about the same. A report built only on website visits will undercount that. That is why the first thing to ask of any report is what it measures, and the second is whether the numbers are comparable. The wider picture is in our pillar post on what three North Alabama businesses taught us about search and in seen more, clicked less.
Nine metrics, and what to watch out for
Metric | What it tells you | Watch out for |
Impressions | How often you appeared in Google Search | Counted differently before and after mid-September 2025 |
Clicks | How often someone clicked through to your site | Falls when Google answers on the page; flat can hide growth |
Average position | Your average rank across all searches | Jumped for many sites in September 2025; blends best and worst searches |
Click-through rate | Clicks divided by impressions | Drops when new pages add impressions at low positions |
Business Profile views | How many people saw your profile in Search or Maps | Google shows only six months; save it regularly |
Profile actions | Calls, direction requests and website clicks | Often the best sign of intent; still not customers |
Website visits (sessions) | Visits to your website | Seasonal; compare the same months year over year |
Direct visits | Visits with no recorded source | Likely includes apps, Maps and AI assistants, not just typed addresses |
Indexed pages | Pages Google has stored and can show | A page that exists is not necessarily indexed |
Here is how we read each one.
1. Impressions
An impression means your business appeared in a Google Search result, whether or not anyone clicked. It is the best measure of how visible you are. It is also the metric that changed meaning in mid-September 2025, when Google stopped allowing 100 results per page (the num=100 setting, dated by industry coverage to September 10 to 14, 2025). Rank trackers had been loading deep results and inflating impressions. Search Engine Land's analysis of 319 properties found 87.7% lost impressions and 77.6% lost ranking keywords. If a report compares impressions from before and after that date, the comparison is unreliable.
2. Clicks
Clicks are real visits from Google to your website, and they were not affected by the September 2025 counting change. They are also the number that falls when Google answers a question on its own page. For the family practice, impressions rose 36% (July 23 to October 14, 2026 against 2025) while clicks were flat, down 2% (2,154 to 2,120). For a Decatur equipment rental company, impressions more than doubled (8,148 to 16,873, up 107%) and clicks rose a modest 16% (886 to 1,024). We state flat clicks plainly because a report that spins them as growth is not worth trusting.
3. Average position
Average position is your average rank across every search you appeared in. It sounds precise and is easy to misuse, because one very good result and many poor ones blend into a middle number. In the week of September 1, 2025, the practice's average position was 14.2. Two weeks later it was 8.0. For a hometown pharmacy it went from 19.7 to 8.5 over the same weeks. Neither business had changed anything on its site. The jump was Google's counting change, and we do not count it as a result. Any chart that shows average position moving across that fortnight is showing a change in measurement.
4. Click-through rate
Click-through rate (CTR) is clicks divided by impressions. It falls naturally when you add pages that start out appearing at lower positions, because they add impressions faster than clicks. A falling CTR in a report is not automatically bad news. Ask whether impressions grew at the same time.
5. Business Profile views
This counts how many times your Google Business Profile was seen in Search or Maps. For our three clients, about 9 in 10 views came from Google Search and about 8 to 9% from Maps (May to mid-October 2026). Google shows only six months of profile data at a time, so if nobody saves it each month, you lose the history. Ask your agency whether they do.
6. Profile actions
Actions are calls, direction requests and website clicks from the profile. For the practice (May to mid-October 2026), about 45% were calls, 13% directions and 42% website clicks. They are often the best sign of real intent, because someone did something. They are still not customers or patients, and a good report never calls them that.
7. Website visits (sessions)
A session is one visit to your site; it ends after 30 minutes of inactivity. Visits are seasonal, so compare the same months year over year. At a Decatur equipment rental company, April 2026 had 777 visits against 529 in April 2025 (up 47%). Compare April with last April, not with February, because spring is that business's busy season. Also check whether your platform measures visits at all: the practice's website visits were not measured in our study, so we could not compare them with its profile.
8. Direct visits
Direct means no recorded source. It is not the same as "typed the address." It likely includes links from texts, email, social apps, the Maps app and AI assistants. At a Hartselle hometown pharmacy, direct was about 50% of visits from January to September 2026, and only five visits since May 2024 carried an AI label. We cover what may be hiding in that number in direct traffic is rising.
9. Indexed pages
A page can exist on your site and still not be in Google. "Indexed" means Google has stored the page and can show it. For the practice, Google's URL Inspection tool showed all 50 new pages as submitted and indexed on October 14, 2026. Ask whether your report checks indexing page by page, because a page that is not indexed cannot appear at all.
Five questions to ask about any marketing report
These five catch most of the problems we see. You do not need to be an analyst to ask them.
1. Window: are the two periods the same length and the same season?
A comparison of a 12-month stretch with an 8-month stretch, or a winter with a summer, proves little. The pharmacy's Google Search model showed impressions at 5,042 a month from October 2025 to September 2026 against 3,843 a month from February to September 2025, up 31%. That is a fair statement only with its windows attached, because the two periods are different lengths and different seasons. When we compare the practice and the rental company, we use the same dates in both years. Ask: what exactly are we comparing?
2. Discontinuity: does the comparison cross mid-September 2025?
Google changed how impressions and positions were counted around then. Any comparison that spans that date, especially for impressions and average position, mixes two different measurements. Search Console keeps only 16 months of history, so a comparison starting from June 2025 can still cross it. Ask: does this chart cross September 2025, and if so, why is it being used?
3. Brand: how much of the search traffic is people already searching your name?
Searches for your own name show people who already know you. They are valuable, but they are not discovery. In our study, the share of search clicks from name searches was 98% for the pharmacy, 84% for the practice and 37% for the rental company (clicks with a reported search term, June 13, 2025 to October 14, 2026). Our rule of thumb is that above about 80%, the website mostly serves existing customers. Ask: what percentage is brand, and how much of the growth comes from searches without my name? We walk through the check in name searches vs discovery.
4. Beyond the website: does the report include your Business Profile, calls and directions?
If a report covers only website traffic, it can miss most of the picture. For the two sites where we measured visits, the profile was seen 2.0 to 4.5 times as often as the website was visited (May to mid-October 2026; the pharmacy figure counts one of its two stores). Ask: where are my profile views, calls and direction requests?
5. Proof: can I trace each number to a source I could check myself?
Every number in a good report should point to a tool you can open: Search Console, the Business Profile Performance page, your website analytics. If a number cannot be traced, it cannot be trusted. Ask for the source and the date range behind every chart, and ask to look at it yourself once.
Put the questions to work this week
Open your most recent report and find the date range of each comparison.
Circle any chart that includes September 2025 and ask why.
Find the brand share. If it is not in the report, ask for it.
Ask where your Business Profile views, calls and direction requests are tracked.
Ask the agency to name one metric that did not improve and what they plan to do about it.
That last step is the one we pay most attention to. A report that cannot name a number that did not move is probably not reading the data honestly. We tell clients when clicks are flat, when a ranking "improvement" was Google's counting change, and when a result cannot be separated from the other work done at the same time. If you want more on how we approach local visibility, see our local SEO and AI enhancements service.
Frequently asked questions
Why are my impressions up but my clicks flat?
Google is showing your business more often but answering more questions directly on its own page, so fewer people need to click. A Decatur family practice had impressions up 36% (26,782 to 36,377) and clicks down 2% (2,154 to 2,120) from July 23 to October 14, 2026 against 2025. Flat clicks with rising impressions can still mean growing visibility, and a good report says plainly that clicks did not grow.
Can I trust average position in my report?
Be careful with it. Average position blends your best and worst searches into one number. Around September 10 to 14, 2025, it jumped for many sites because Google stopped allowing 100 results per page, not because of anything the sites did. Never accept a comparison of average position that crosses mid-September 2025 as evidence of improvement. Compare clicks and profile actions instead.
What is the September 2025 change in my report?
Around September 10 to 14, 2025, Google stopped allowing 100 results per page. Rank trackers had been loading deep results, which inflated impressions. Search Engine Land found that 87.7% of 319 properties lost impressions afterward. A report that compares before and after is comparing two different ways of counting. Clicks were not affected, so they remain comparable. Ask for impression comparisons that start after the change.
What is a good share of branded searches?
There is no universal number. Our rule of thumb is that above about 80%, your website mostly serves people who already know you. Our three clients ranged from 37% to 98% (June 13, 2025 to October 14, 2026). A high share is not a failure, particularly for a long-established business. Ask what share of your growth comes from searches without your name.
Why should my report include my Business Profile?
Because customers use it as the front door. Across the three profiles we studied, about 9 in 10 views came from Google Search and about 8 to 9% from Maps (May to mid-October 2026), and for the two sites where we measured visits, the profile was seen 2.0 to 4.5 times as often as the website was visited. A report without it misses the largest piece of local visibility. Also ask the agency to save the six months of profile history Google shows.
What should I ask my marketing agency at my next meeting?
Ask five things: what two periods are being compared, whether anything crosses September 2025, what share of search is your name, where your Business Profile views, calls and directions are, and where each number comes from. Then ask which metric did not improve. An agency that answers that last question openly is usually one you can trust with the rest. Write the answers down so you can compare them at the next meeting.
Keep reading
What three North Alabama businesses taught us about search in the AI era: the study behind every example here.
Seen more, clicked less: why impressions and clicks now move apart.
The ranking jump that wasn't: Google's September 2025 reporting change: how to spot it in your own Search Console.
Name searches vs discovery: how to work out your own brand share.
Zero-click search: why businesses get fewer visits even when they rank 1: the national picture behind flat clicks.
A good report should leave you knowing what to do next, not just what happened. If yours does not, we will read it with you. Call (256) 274-1289 or message us for a free visibility check.