NextGen Healthcare | Search & Traffic Report Card

NextGen Healthcare | Digital Team

Search & Traffic Report Card

Google Search Console and Google Analytics, side by side. Same months, same math, plain answers.

View

Time range | trend charts

Traffic source | GA4 detail panels

Data

How to read this page. Google Search Console (GSC) counts what happens on Google before anyone visits: how often we appeared (impressions) and how often someone clicked. Google Analytics (GA4) counts what happens on our website: visits (sessions) and people (users). Simple view gives you the answer to each question in one line and one picture. Technical view opens up the full tables, footnotes, and export details. The grey chips under each heading tell you exactly which export the numbers came from. Anything marked Demo data is a placeholder waiting for a real export and should not be quoted.

★

Start here: the answer for leadership

Verified numbers

"GA says we're up, GSC says we're down, Salesforce says leads are down. Is the website performing or not?"

Website_Traffic_Exclusing_errors_25-26.xlsx | raw tabsNextGen_Channel_Analysis_1.xlsx | GSC Daily Raw Data tabSalesforce report1785188370846.csv | Type + MGL DateFull-month YoY pairs, named per tile

More people visit the website than last year, but fewer of the right people. That is why leads are down.The extra visitors come from ads and unknown doors. The visitors who used to fill out forms, from Google searches and emails, shrank. The website still works; the mix of visitors changed.

Total traffic is up year over year, but the honest number is about +8%, not the +38% July prints. Roughly 112K of July's sessions arrived in a 6-day late-July surge of untagged Direct traffic aimed at the homepage and privacy policy page, a bot signature; excluding it, July matches June's +8%. So "traffic is up" and "leads are down" are both true. The growth is bought and untagged: Paid Social ramped 4 to 5x and Direct grew, while organic fell about 38% and email audiences shrank about 70%. Web form fills are down about 40% from two real causes: fewer high-intent visitors, and a softer fill rate per visitor (down about 30%, starting December 2025, before any of our 2026 form changes). The store analogy: the store is busier than ever, but the extra crowd is bussed-in window shoppers, fewer people walk in after asking Google for directions, and the ones who do come in buy a little less often than last fall. Separately, the biggest lead source overall (email) fell off a cliff in October 2025, which drives most of the total MGL decline.

Leads per month, by where they came from

Email leadsWebsite leadsEverything else

The lead cliff is mostly an email cliff. Email leads collapsed from 641 in Sep 2025 to a 66 to 238 range. Website leads softened but never cratered. First question for the room: what changed in the email program in October 2025?

Why "GA4 says we're up" is a broken ruler

The report showing +104% clicks and +72% impressions is GA4's Search Console-linked view. Its previous-year baseline is missing months of data: it claims ~1.6M impressions per month for Aug 2024 to Aug 2025, but native Search Console recorded 2.0M to 3.0M per month in that same stretch, counted the smaller way. A baseline that small is not possible with complete data. Every green arrow in that report measures how many months of data exist, not performance. Do not use that report for year-over-year. Confirm in GA4 Admin, Search Console Links: the link creation date will postdate Aug 2024.

The late-July surge is not real visitors

Jul 26 to 31 ran at 30,523 sessions per day against an 11,909 baseline for Jul 1 to 25: about 112K excess sessions in six days, almost all untagged Direct traffic to the homepage and /privacy-policy (54,626 "new users" hit the privacy page on Jul 30 alone). That pattern is bots or an unattributed automated source, not buyers. Quote +8%, not +38%, as the year-over-year traffic number. Half-month totals from the split exports made this visible.

One honest caveat on "GSC impressions down 47%"

In mid-September 2025 Google removed the num=100 results parameter, erasing millions of phantom impressions from rank-tracker bots across the whole industry, and making average position "improve" overnight. Our impression cliff starts exactly then. So part of the GSC decline is phantom impressions disappearing, not real visibility loss. Clicks and sessions are the honest metrics, and those are genuinely down for organic.

"Is the website converting?" in one line: web form-fill leads per 1,000 organic sessions

Around 2.7 to 2.8 through September 2025, then a slide to roughly a 1.9 average in 2026 (trough: 1.27 in February). The dip begins December 2025 (red dashed line: the December 2025 Core Update), months before the May to July 2026 form and tracking changes, so those changes did not cause it. Some of the winter dip may be seasonal; we need last winter's baseline (the 24-month exports) to separate that.

Suggested actions: 1) Take the traffic-mix table to Stacy and Rachael before anyone debates dashboards: growth is Paid Social and Direct, decline is organic and email, and leads follow the second group. 2) Ask marketing ops what changed in the email program in October 2025 (sends, list size, or scoring). 3) Kill the GA4 Search Console-linked report for YoY use; confirm the link creation date in GA4 Admin. 4) Pull the GA4 event-name exports to locate the December rate dip: form level or form-to-MGL level. 5) Ask who launched what in Jan and Jul 2026: Direct swings of 100K+ sessions need an owner or a bot explanation.

Session and organic figures: GA4 errors-excluded export; each tile names its comparison months. Branded clicks: GSC daily branded split, same months. Leads: Salesforce MGL report by Type, Jul 2025 to Jun 2026; the report window starts Jul 2025 so the pre-peak baseline is not yet visible. Known internal changes that muddy lead counts: Qualified Meeting duration change for SDRs (May 16, 2026) and LP form changes (Jun to Jul 2026), from the Digital Marketing Updates Log.

The headline numbers

Live data

How much did we show up, and how many people actually came?

PerformanceonSearch20260515.xlsx | Chart tab | Jan 2025 - May 2026Website_Traffic_Exclusing_errors_25-26.xlsx | raw tabs | Jun 2025 - Jul 2026

We showed up on Google less than last year, but the website got more visits overall.Being seen and being visited are different doors. The next sections show which doors changed.

Suggested actions: Standardize on these two rulers: native GSC for Google visibility, errors-excluded GA4 for site traffic. Re-pull the GSC site export without the Review-snippet filter so this section runs through July.

Impressions and clicks: GSC site-wide daily export rolled up by month, filtered by the range picker. Sessions: GA4 deduplicated monthly grand totals (from the 14 monthly CSV exports), Jun 2025 to Jul 2026. Year-over-year arrows compare the latest full month against the same month one year earlier. Key-event detail in the conversions section still runs on the Apr 22 to Jul 20, 2026 export until the event-name exports land.

Do GSC and GA4 agree?

Live data

Google says people clicked. Our website says people arrived. Do the two stories match?

PerformanceonSearch20260515.xlsx | Chart tabWebsite_Traffic_Exclusing_errors_25-26.xlsx | raw tabs | Organic Search rows

Google's counter and our website's counter tell the same story.They use different rulers, so the numbers differ, but they move together. That means we can trust them.

GSC organic clicks vs GA4 organic sessions, by month

GSC clicks (Google's count)GA4 organic sessions (our website's count)

Average gap between the two

GA4 organic sessions divided by GSC clicks, averaged over months where both exist. Above 100% means GA4 counts more than GSC.

Why they never match perfectly: GSC only counts clicks from Google itself, while GA4's "organic" bucket includes Bing, DuckDuckGo, and other search engines. GA4 also counts a new session if someone comes back later the same day, and loses some visits to cookie blockers. A steady gap is normal. A gap that suddenly changes is the thing to investigate.

Suggested actions: Use the steady ratio as the health check: if GA4-organic-per-100-GSC-clicks suddenly moves, investigate tagging or bot filtering before strategy. Report the ratio monthly, not the raw gap.

GSC clicks: site-wide daily export rolled up monthly. GA4 organic: "Organic Search" channel from the monthly channel export. Months are only compared where both sources have full data.

Branded vs non-branded

Live data

Are people finding us because they already know our name, or because we won a search we had to earn?

NextGen_Channel_Analysis_1.xlsx | GSC Daily Raw Data tab | Feb 2025 - Jun 2026searchenginejournal.com algorithm history

Most clicks come from people who already know our name.Green line: searched for us by name. Orange line: searched for the kind of thing we sell.

Clicks per month

Branded (searched our name)Non-branded (searched the topic)Google update

Impressions per month (how often we appeared)

BrandedNon-branded

Average position on non-branded searches (lower is better)

Suggested actions: Branded clicks fell 46% YoY: run the branded-SERP audit (do AI answers, sitelinks, or the portal answer brand queries before our pages?). Non-branded clicks were never material, so fund AEO/GEO citation work (ungated answers on nextgen.com) rather than chasing rankings that no longer pay clicks.

Source: daily branded vs non-branded GSC split from the channel analysis workbook, rolled up monthly. Branded = queries matching the brand regex (nextgen, mirth, medfusion, healthfusion, meditouch, pxp, and common misspellings). Rank improving while non-branded clicks stay flat is the expected effect of Google's AI Overviews answering questions on the results page; a measurement change, not a performance failure.

Inside Google's AI answers

Live data

When Google writes the answer itself, do our pages get pulled into it?

Performance-on-Search-Generative-AI-Features-2026-07-22.xlsx | Chart + Pages tabs

Google's AI answers use our pages, and it is growing.When Google writes the answer itself, it borrows from our homepage and Mirth pages most.

Weekly impressions inside AI features (AI Overviews and AI Mode)

Pages Google pulls into AI answers most

Suggested actions: Keep the cited pages (home, Mirth, Integration Engine) ungated and current; they are our AI shelf space. Re-export this report monthly and watch for Google adding query data; fold it into the TIGER report card.

Source: GSC "Performance on Search Generative AI Features" export. Data begins May 18, 2026, when Google started reporting it. Google reports impressions only for AI features: no clicks, no queries yet. When query data ships, this section grows a query table.

What people searched

Live data

The actual words typed into Google before finding us.

GSC_Branaded.xlsx + GSC_Non_Branded.xlsx | Queries tabs | Apr 20 - Jul 19, 2026PerformanceonSearch20260515.xlsx | Queries tab | 16 mo

These are the real words people typed before finding us.Flip the tabs to see name searches vs topic searches.

Suggested actions: Branded list: protect login and portal queries with clean titles and sitelinks. Non-branded list: map the top 20 to pages, then write direct-answer sections for the question-style queries; those are the AI fan-out targets.

Where people come from

Live data

Which doors people walk through to reach the website.

Website_Traffic_Exclusing_errors_25-26.xlsx | raw tabs | Jun 2025 - Jul 2026

Most visitors come straight to us or from ads. AI visitors are few but growing fastest.Each color is a different door into the website.

Sessions by channel, per month

Organic search vs Direct, sessions per month

Organic SearchDirect

All paid ads vs Direct, sessions per month

Paid Search + Paid Social combinedDirect

Top traffic sources

Top referrers (sites that linked people to us)

Sessions in the "AI Organic Traffic" channel, per month

Suggested actions: 1) Audit Direct: the Jul 26 to 31 surge (30.5K sessions/day, homepage + /privacy-policy) and the mid-Jan spike need owners or bot exclusions; get an IT/bot filter ruling before these pollute another month's totals, and tighten UTM discipline with Jeff's fix. 2) Ask for Paid Social lead quality before celebrating its 4-5x volume. 3) Add the AI Organic Traffic channel to Nikki's monthly KPI pull; the May-Jul surge is the trend to track.

All panels: GA4 export, Jun 2025 to Jul 2026, full 14 months. Channel and source figures are page-weighted sums (a visit touching 3 pages counts 3 times), which overstates absolute session counts by roughly 1.1x to 1.2x but tracks trends faithfully; headline totals elsewhere use GA4's own deduplicated grand totals. Tables react to the traffic source filter in the header. The Total users view comes from the workbook's "Total Users by Channel" tab. "AI Organic Traffic" is the custom channel group defined in our GA4 admin; it counts sessions the channel rules catch, so treat it as a floor, not a ceiling.

Top pages, by what kind of page they are

Live data

Which solution, service, specialty, company, and blog pages pull their weight?

Website_Traffic_Exclusing_errors_25-26.xlsx | raw tabsPerformanceonSearch20260515.xlsx | Pages tab | 16 monextgen_website_sitemap | URL classification

These are our most visited pages.Use the tabs to see each family of pages.

Top pages people landed on from organic search (GA4, Jun 2025 to Jul 2026 | always organic, ignores filter)

Suggested actions: Prioritize the SEO/AEO brief pipeline by this table: highest-session solution pages with weak GSC clicks are the optimization shortlist. Use the organic landing table to pick the next webinar recap and FAQ targets.

GA4 columns (sessions, users): Jun 2025 to Jul 2026 window, errors excluded, and they react to BOTH the time range picker and the traffic source filter (users show n/a under a source filter, where only sessions are stored per month). GSC columns (clicks, impressions): full 16-month aggregate per page and always reflect all Google traffic; GSC cannot be filtered by our GA4 sources, so those columns hide when a filter is on. Categories assigned from URL path.

Conversions and events

Partly demo

When people did the things we care about: downloads, form fills, demo requests.

GA4_Website_Data.csv | Apr 22 - Jul 20, 2026 (older window)Event-type split | demo pending export

Fewer people fill out our forms than last fall.Fewer of the right visitors arrive, and each visitor fills forms a bit less often.

Key events by source Live

Key events by page Live

Demo data below. GA4's export gives us one combined "key events" number. The split into file_download, form_submit, asset_download, and demo_request needs the "Monthly x Event name" and "Event name x Page" exports. The numbers below are invented placeholders showing the final layout; they swap out the moment the real export lands.

Conversions by type, per month Demo data

form_submitfile_downloadasset_downloaddemo_request

Suggested actions: The event-name exports are the single highest-value pull remaining: they split form_submit from downloads and date the December dip. Until then, treat per-source "events per 100 visits" as the quality score when judging Paid Social and Direct.

Live tables: GA4 "Key events" totals by source and by page, Apr 22 to Jul 20, 2026; both react to the traffic source filter. "Key events" is GA4's umbrella count of all conversion actions combined.

Campaigns

Demo data

Which named campaigns are bringing people in?

Pending GA4 "Monthly x Session campaign" export

Demo data. Campaign names are not in the current exports; this table is a placeholder shape. It fills with real numbers from the "Monthly x Session campaign" GA4 export.

Month by month

Live data

Every month in one table, with what Google changed and what we shipped laid over the top.

NextGen_Channel_Analysis_1.xlsx | GSC Daily Raw Data tabWebsite_Traffic_Exclusing_errors_25-26.xlsx | raw tabssearchenginejournal.comDigital_Marketing_Updates_Log.xlsx | Updates tab

Each row is one month: what Google showed, who came, and what changed.Red notes are Google changing its rules that month.

Timeline: Google algorithm updates ● and what we shipped ●

Suggested actions: Keep logging ships in the Updates Log with dates; this table is the causality defense. Before attributing any swing to our work, check the Google column first, then the Direct channel, then our ships.

Algorithm dates: Search Engine Journal's Google algorithm update history. Shipped items: Digital Marketing Updates Log. GA4 columns run Jun 2025 to Jul 2026 (errors excluded); GSC branded columns end Jun 22, 2026 pending the refreshed export. Month table respects the range picker.

Where every number came from

Each dataset baked into this file, its exact window, and which sections use it. Nothing on this page is invented except the two panels marked demo.

The Top of Funnel: One Story

The direct answers to the questions asked, with the exact report path to verify each one independently. Everything here stops at form fill; Salesforce is out of scope as requested. The charts that used to conflict differed because of four measurement artifacts, not four realities; question 5 lists them. Fixed windows, no filters, so the story cannot change mid-meeting.

The bigger picture of Search: brand and non-brand, all engines and AI

What was asked

"The bigger picture of Search in general (brand and nonbrand, us vs. competitor), not just Google search, but across all engines and AI. What is the story, trends, year over year."

What this is really askingIs our search visibility actually declining, is it specific to us or the market, and is AI search replacing what we lost?

The answer

We appear on Google less than a year ago, but a large share of that drop is a Google measurement change, not lost visibility; our rank on the terms that matter improved, branded clicks fell about 46%, and our pages are increasingly pulled into Google's AI answers, where impressions are growing month over month.

Branded vs non-branded Google clicks, monthly (red lines: Google updates)

BrandedNon-branded

How to pull this yourself

GSC > Performance > Search results > Query filter > Custom (regex) with brand terms, then the inverse, 12 months with YoY compare. AI features: GSC > Performance on Search Generative AI (impressions only for now). Non-Google engines: GA4 traffic acquisition, source contains bing or duckduckgo. Us vs competitor on classic rankings needs a Semrush position export; that gap is listed at the bottom.

Website: are sessions and visits roughly the same or higher?

What was asked"It would appear sessions, visits are roughly same or higher."

What this is really askingCan I trust the number that says traffic is fine?

The answer

Yes, sessions are up about 8% year over year on GA4's own deduplicated totals; July prints +38%, but roughly 112K of that arrived in a six-day late-July surge of untagged bot-pattern traffic, and excluding it July matches June's +8%.

True monthly sessions (GA4 deduplicated grand totals)

How to pull this yourself

GA4 > Reports > Acquisition overview with a 12-month compare, or an Explore with Month and Sessions only. The trap that produced conflicting charts: adding a page dimension double-counts sessions by about 1.2x, and the Search-Console-linked report has an incomplete prior-year baseline; use neither for totals.

Where is traffic shifting from and to?

What was asked"Is there more direct now? More paid... less organic... where is traffic shifting from and to."

What this is really askingIf totals held, what changed underneath, and does the new mix produce the same value?

The answer

Traffic shifted out of organic search (down about 38%) and email (audience down about 70%) and into paid (Paid Social up 4 to 5x, Paid Search roughly doubled) and untagged Direct; totals held, but the visitors who historically fill forms were replaced by visitors who mostly do not.

The shift: Organic vs all Paid vs Direct, sessions per month

Organic SearchPaid Search + Paid SocialDirect

How to pull this yourself

GA4 > Reports > Acquisition > Traffic acquisition, dimension Session default channel group, plotted by month across 14 months. Channel-level figures overstate absolutes (multi-page and multi-channel counting); read them for mix and trend.

Are people engaging more or less with the site?

What was asked"Are people engaging more or less with site."

What this is really askingIs the quality of a visit holding up, or is the site itself getting worse at holding people?

The answer

On the proxies we can compute cleanly, per-visitor behavior is roughly stable; a precise answer needs one small GA4 engagement export we have requested, and we are deliberately not quoting engagement rates from the report with the broken baseline.

Sessions per user, monthly (a stability proxy)

How to pull this yourself

GA4 > Explore > Free form: Month, Engagement rate, Average engagement time, Engaged sessions, Jun 2024 onward. Five minutes of work; this is the one export that fully answers the question.

"I feel we have charts that conflict. We just need one story."

What this is really askingAlign on one narrative and explain why the numbers disagreed.

The answer

The charts conflicted because of four measurement artifacts, not four realities: a GA4 report comparing against an incomplete prior-year baseline (made everything look up), Google removing the num=100 parameter in September 2025 (made impressions look collapsed), page-level session sums (inflated totals about 1.2x), and bot surges in Direct (inflated July). Corrected for all four, every source agrees: total traffic stable to up about 8%, the mix reversed from organic and email toward paid and untagged, fewer high-intent visitors arrive, and that is the top of funnel story down to form fill.

The follow-up questions we expect

What will be asked"Direct is up, paid is up, organic consistent. So if traffic to site is consistent... is it that we have a conversion problem on the site? Is it the type of traffic and the reason it is not converting?"

One premise to fix first"Organic consistent" is true within 2026 (a 52K to 78K monthly band since January) but organic is down about 38% versus a year ago (98K in Jun 2025 vs 61K in Jun 2026). Both are accurate for their window; name the window out loud.

Answer: is it a conversion problem on the site?

Partly, and we can size it: the form-fill rate within organic traffic, our constant-intent yardstick, fell about 30% starting December 2025, before any of our 2026 form changes; but most of the lead decline comes from who arrives, not from the site's machinery. Never quote the blended site-wide conversion rate: adding low-intent paid and bot traffic crushes that average even when the site changes nothing.

Answer: is it the type of traffic?

Yes, mostly: the channels that grew convert at roughly 1 key event per 100 visits, while the channels that shrank convert at 2.4 to 8.4. Trade 10,000 email visits for 10,000 paid social visits and roughly 835 key events become 97, with the website doing nothing wrong.

Key events per 100 visits, by traffic source (Apr 22 to Jul 20, 2026)

Two footnotes: Direct is a mixed bag of returning customers, logins, and bots, so its rate alone does not make it a healthy channel. AI Chatbot visitors convert above organic; small volume, but it backs the AEO investment case with conversion data, not just visibility data.

How to pull this yourself

GA4 > Reports > Acquisition > Traffic acquisition: Sessions and Key events columns, divide. Or Explore: Session default channel group, Sessions, Key events. The remaining unknown, whether the December within-organic dip happened at the form itself, is settled by the event-name exports listed below.

What we cannot answer yet, and the two exports that close it: (1) engagement precisely: GA4 monthly engagement export as described in question 4; (2) us vs competitor on classic rankings: a Semrush position tracking export for the agreed keyword set. Everything else above is backed by data already in this file; switch to Technical view for every table and footnote.