The short answer
Yext counts how often each location's listing was found, viewed and acted on, plus its reviews by publisher and star. It gets no Google Analytics data. Link each location's Yext data and website to DuckDive, and your AI agent reads listings, reviews, GA4, Search Console and Ads for that location in one answer.
- Calls and direction requests never touch your website, so a site report alone misses half of local demand.
- Reviews arrive as a current breakdown with no date range. Save one on the 1st of each month to see the trend.
- Replying to reviews, reading review text and editing listings stay in Yext.
Listings actions and site sessions: two halves of local demand
A hungry commuter searches for a bakery, sees your nearest listing and taps Directions. That person never loads your website. Yext records an action; GA4 records nothing. A second person taps Website on the same listing and lands on your order page. Now Yext counts a click and GA4 counts a session.
Local demand splits at the listing. Yext's Location Listings Metrics reference defines Total Listings Actions as clicks and taps across publishers, "such as requests for directions, website clicks, and calls" (checked November 2026). Only one of those three ever reaches your site. So a location can gain walk-ins while its traffic chart sits flat, or lose phone orders while the site looks healthy.
| Step | Yext listings | GA4 on your site |
|---|---|---|
| Listing shows in Maps or Search | Yes, as a view or impression | No, nobody visited yet |
| Customer taps Call | Yes, as an action | No, the phone rings instead |
| Customer taps Directions | Yes, as an action | No, they drive over |
| Customer taps Website | Yes, as an action | Partly, a session, often filed under Google organic |
| Customer places an order online | No, it happens off the listing | Yes, if it is a conversion |
Sources: Yext listings metric definitions and Google's URL builders help page, checked November 2026.
That fourth row is why tagged links matter. Google's URL builders page says to always use utm_source, utm_medium and utm_campaign when you add parameters to a link. Put them on each listing's website link and GA4 can split listing visits from ordinary search visits.
DuckDive is the place where the two halves meet. For each location's website it stores the Yext listings and review figures beside that site's traffic, organic search, paid search and Local Services Ads numbers. Claude, ChatGPT, Cursor or any MCP client then reads all of it through DuckDive's MCP server, so "did the listing traffic reach the site?" becomes one question rather than two exports and a lookup formula.
Ask your agentFor each location, put Yext listing views and actions over the last 30 days next to GA4 sessions and conversions.
Tools it calls list_websites get_yext_summary get_ga4_overview
What Yext reports and what it never sees
Yext draws its own boundary in writing. Its help article on Google Search Console compares three Google services and states: "Yext does not receive any analytics directly from Google Analytics." Listings figures come from Google Business Profile. Search Console figures feed Yext's Pages product, the location pages Yext hosts. If your store pages live on your own CMS, their traffic never shows up in Yext's reports.
A listing view is not a Search Console impression
The same article explains the difference. Listings metrics count the whole profile when it appears, in Google Search and Google Maps. Search Console counts a URL when it appears, in Google Search only. The two numbers measure different events, so never add them up or read the gap between them as lost traffic. Ask your agent to keep them in separate columns.
Similar names, different publishers
Yext's metric reference has several "was seen" counts with different reach. Its Searches metric leaves out Yelp, Facebook, Bing and Google. Profile Views leaves out Yelp, Bing and Google. Total Listings Impressions covers Google Maps, Google Search, Bing, Facebook and the wider network. Know which one you hold before setting it beside a Google number.
- Listing status
- Whether the location's listing is in good shape on each publisher Yext syncs to, as
get_yext_summaryreports it. - Searches, views, actions
- The three activity counts
get_yext_summaryreturns over time: the listing turned up, someone looked at it, someone acted on it. - Review count
- How many reviews the location's Yext record holds, split by publisher and by star rating, from
get_yext_reviews. - Average rating
- Yext's reviews reference calls it "the cumulative average of the ratings your business has received."
- Sessions and conversions
- Visits to the location's website and the conversions recorded on them (key events in GA4), from
get_ga4_overview.
Ask your agentOver the last 90 days, show Yext listing views beside Search Console impressions and clicks for every store, as separate columns. Which stores gained on the listing but not in search results?
Tools it calls list_websites get_yext_summary get_search_console_summary
If listing views fall while organic clicks hold, the cause often sits in local rank rather than the website. Our guide to telling a map pack slip from an organic one walks through that split, and Local Falcon grids in the same agent show where in town the listing lost ground.
Worked example: listing views up, website visits flat
A site-only monthly report misreads this shape every time. It shows flat sessions and says "no change" while the store manager watches a busier counter.
DuckDive's data tools take fixed windows, not custom dates, so the clean comparison is the last 30 days against a typical month: the last-90-day total divided by three. One caution before reading the result. The 90 days include the latest 30, so a real jump looks smaller this way than it would against the previous month alone. Treat anything above 10% as worth a look.
For each Hearthline bakery, put Yext listing views and actions for the last 30 days next to GA4 sessions and online orders. Measure each one against a third of its last-90-day total.
list_websitesfinds the three bakery websites, each with Yext and GA4 linkedget_yext_summaryeach bakery, once for last30days and once for last90daysget_ga4_overvieweach bakery, sessions and conversions for the same two windows
Marlowe Street is the one to look at. Its listing drew 9,840 views and 1,312 actions in the last 30 days, about 14% above a typical month in the last 90 (8,610 and 1,150). Its website barely moved: 2,410 sessions against a typical 2,380, and 188 online orders against 185. The extra demand stayed on the listing, most likely as calls or direction requests.
| Bakery | Views | Actions | Visits |
|---|---|---|---|
| Marlowe Street | +14% | +14% | +1% |
| Quayside | +1% | -2% | -4% |
| Sallow Bridge | -1% | +18% | +8% |
Sallow Bridge shows the opposite shape. Views held near 4,400, yet actions reached 590 against a typical 500, and sessions 1,190 against a typical 1,100. More of the same viewers did something, and part of that reached the site. Check what changed on that listing this month. Quayside held steady across all three.
Next step: split Marlowe Street's actions by type in Yext. If direction requests drove the rise, no website report was ever going to show it.
Bing runs weekly
Yext's metric reference notes that Bing sends search data in full weeks rather than single days. A window that ends midweek can trail the other publishers, so wait a few days before you call a short dip on Bing a trend.
Ask your agentWhich shops had more Yext listing actions in the last 30 days than in a typical month of the last 90, while their organic sessions went down? Use the GA4 medium split, not total sessions.
Tools it calls list_websites get_yext_summary get_ga4_channels
When the question is phone volume rather than footfall, the guide to finding locations that lost calls lines listing actions up against Local Services Ads leads and Google Ads conversions.
Find the stores whose demand never reaches the site
Link one location's Yext data and website, then ask which half of its demand moved this month.
Free for 2 sites · No credit card
Reviews by publisher and star rating
get_yext_reviews returns, for the website's Yext location, its review count and average rating split two ways: by publisher (Google, Yelp, Facebook or whichever sites your Yext account collects from) and by star. It takes no date range. You get the breakdown as it stands at the moment the agent asks.
That is thinner than Yext's own review screens, yet it answers a regional manager's first three questions: which store is slipping, on which site, and how much of it is one- and two-star reviews.
Read the shape, not only the average
On a store with several hundred reviews, a bad month barely dents the average. The share of one- and two-star reviews moves sooner, so watch it as the early signal. Two stores with the same average can also differ: one full of fours, the other split between fives and ones.
Weight each publisher by its count
To rebuild a store's overall rating from the publisher rows, multiply each publisher's average by its review count, add the results and divide by the total count. A site with a handful of reviews can show a perfect 5.0 and mean nothing. Ask the agent to print the count next to every average so nobody quotes the tiny one.
Ask your agentRebuild Marlowe Street's overall Yext rating from its publisher rows, weighting each average by its review count, and flag any publisher with fewer than 25 reviews.
Tools it calls list_websites get_yext_reviews
Tracking review trends from snapshots
With no date range on the review tool, this month's answer quietly replaces last month's. The fix is low-tech: save the breakdown on a fixed day and compare. Any two saved breakdowns give you the change between them, per star and per publisher.
What one snapshot should hold
One row per store per publisher, with the date, the count at each star from one to five, the total and the average. Keep raw counts, not percentages. You can always turn counts into shares later, but you cannot rebuild counts from a share.
Ask your agentIt's the 1st. Give me every store's Yext review counts by star and by publisher as one table I can save, with today's date in the first column.
Tools it calls list_websites get_yext_reviews
Where to keep it
Anywhere your agent can open again next month: a spreadsheet tab, a dated note in a project, or a CSV in a repo folder if you work in Claude Code or Cursor. DuckDive does not keep these snapshots for you.
The arithmetic between two snapshots
Subtract one from the other to get the net change at each star. For the rating of what arrived in between, work out star points (each star value times its count, added up) for both dates. Divide the change in star points by the change in review count. That figure tells you how the last month felt to customers, which the lifetime average smooths away.
The word "net" matters here. Yext's May 2026 release notes say a review's Last Updated Date now reflects changes the reviewer makes to rating or content, and that a log of those changes is kept per review. A customer who edits a four into a one adds a one-star review to your counts without adding a review. A removed review can take one away. If any star count goes down between snapshots, you are looking at an edit or a removal.
A monthly snapshot routine
- Same day each month. The 1st is easy to remember and lines up with monthly reports.
- Both splits. Save counts by star and by publisher, because a fall usually lives on one site.
- Never overwrite. Add new rows under the old ones; the history is the asset.
- Check the sums. Star counts should add up to the total. If not, ask the agent why before saving.
- Flag any falling count. It means edits or removals, which deserve a look in Yext.
- Pair it with demand. On the same day, save each store's 30-day listing actions so reputation and footfall sit together.
Be fair to Yext here. If you have its Reviews product, its Report Builder can already chart Average Rating by month (the reviews reference lists months as a dimension), and the June 2026 release notes added scheduled review exports by email. Use those when the trend only needs to live in Yext. Snapshots earn their place when you want the review trend in the same answer as listing actions and GA4.
Worked example: a location whose one-star share doubled
A quarter of a star in a month looks minor on lifetime numbers. Two snapshots show otherwise.
Ask your agentCompare today's Yext review breakdown for every location with the snapshot I saved on the 1st, and tell me where the average fell.
Tools it calls list_websites get_yext_reviews
Compare Quayside's Yext reviews today with the October 1 snapshot in my reviews file. What changed, and on which site?
list_websitesfinds the Quayside bakery websiteget_yext_reviewsQuayside's current counts and averages by publisher and star (no date range)
Against your October 1 snapshot, Quayside's one-star share doubled, from 6.0% to 11.9%. Its total rose by 70 and its one-star count by a net 32. The average fell from 4.38 to 4.14, and the net change since October 1 works out to 2.79 stars per added review.
| Stars | Oct 1 | Nov 1 | Change |
|---|---|---|---|
| 5 stars | 270 | 290 | +20 |
| 4 stars | 70 | 82 | +12 |
| 3 stars | 24 | 27 | +3 |
| 2 stars | 12 | 15 | +3 |
| 1 star | 24 | 56 | +32 |
| All reviews | 400 | 470 | +70 |
It is a Google problem. Google reviews went from 310 to 372 and Google's average from 4.45 to 4.13. Yelp (60 to 66) and Facebook (30 to 32) barely moved, and their averages held near 4.0 and 4.4.
I only have counts, not wording. Read Quayside's new one-star Google reviews in Yext next, and check its review history in case some of the 32 are edits of older reviews. Then ask the store whether anything changed around the start of October: staff, hours or the menu.
Franchise systems can add a reputation column to ad fund reporting by location. The wider setup for brands with many stores is on the multi-location page.
When Yext's own tools are the better route
Yext runs its own MCP server, and for several jobs it is the right tool. Its MCP overview (checked November 2026) lists five areas: Knowledge Graph, Listings, Reviews, Scout and Account Tools. It says you need a subscription to each product to use that product's MCP features, and that writes are switched on by default for Knowledge Graph edits and review responses. Yext's June 2026 release notes announced the server.
| Job | DuckDive | Yext's MCP and app |
|---|---|---|
| Reply to a new one-star review | No, counts only | Yes, Reviews tools post responses |
| Read what reviewers wrote | No, no review text | Yes, review text by location |
| Change holiday hours or a description | No, reads only | Yes, Knowledge Graph edits |
| Compare a store with similar brands nearby | No | Yes, Listings benchmarks |
| Listing actions beside GA4 sessions | Yes, same store, one answer | No, Yext gets no GA data |
| Listings beside Search Console, Ads and LSA | Yes, linked per website | Partly, Search Console for Yext Pages |
| Review rating trend by month | Partly, from your saved snapshots | Yes, Report Builder by month |
Sources: Yext MCP overview, Yext on Search Console and Google Analytics and Yext reviews metrics, checked November 2026.
Running both in one agent works: Yext's connector for replies, wording and edits, DuckDive for questions that span sources. For analysis-only sessions, Yext's overview says you can switch Allow User MCP Writes off under Account Settings, Account Features. DuckDive only reads your Yext figures; its optional actions change your DuckDive account, never Yext.
What DuckDive does not do with Yext
- No review text, no replies, and no edits to listings, hours or photos.
- No custom dates: listing figures use fixed windows from 7 days to 12 months, and reviews have none.
- No Google Business Profile connection of its own.
- One Yext location per website.
If your job is managing the review queue and you never look at website numbers, Yext alone covers it. DuckDive pays off when someone asks why a store's reputation, listing activity and website tell three different stories. Agencies running that check for many local clients can see how on the local SEO agency page.
Connecting a Yext location to a website
DuckDive organises everything by website. Each store's website is the anchor, and its Google sources and Yext location hang off it.
-
Add each store's website
Select Continue with Google and add the site in the app, or let your agent do it. Your agent signs in through a DuckDive page where you click Continue with Google once; it never asks for your password.
Give your agent this lineSet up DuckDive for my website. Follow https://analytics.duckdive.ai/
agents.md -
Link that store's Google sources
Hook up the store's GA4 property and Search Console site first. A store that advertises also gets its Google Ads or Local Services Ads account. Doing this before Yext means a dip in listing views can be set against sessions and clicks for the same store from day one, instead of sitting alone. Per-source detail is in the GA4 source guide and the Search Console source guide.
-
Connect Yext from the Integrations page
Yext is not part of the Google sign-in, and no agent tool links it, so this step happens in the DuckDive app under Integrations. Connect it only for the stores that use Yext; the rest work without it.
-
Watch the import finish
Linking Yext kicks off its history import, typically a few minutes of waiting. Ask the agent for
get_sync_statusand read the Yext row: ready, importing or failing, with a note of how recent its numbers are. If the searches, views and actions fromget_yext_summarylater look stale, an admin may point the agent'srefresh_website_datacall at theyextsource with a look-back of anywhere from 1 to 30 days. DuckDive takes one such request per website each quarter hour, so settle the source and the window before asking. -
Save your first review snapshot
Run the snapshot prompt the day Yext shows as ready. The review tool cannot reach back for a starting point later.
One domain, many stores
The review tool reads "the website's Yext location", singular. If every store lives under one domain with a page per address, that domain is one website in DuckDive. Ask us how to set that up before you link anything.
Subscriptions, review data and snapshots
Do I need my own Yext subscription?
Plan on it. DuckDive reads listings and review figures from Yext; it does not sell or include Yext. Which Yext products the connection needs is still being confirmed, so email hello@duckdive.ai before buying anything just for this. Stores without Yext still get GA4, Search Console, Google Ads and Local Services Ads.
Can my agent read the text of my reviews?
Not through DuckDive. The review tool returns counts and average ratings by publisher and by star, with no wording and no replies. For the text, use Yext's review screens or Yext's own MCP server, which its help article says can pull review text and post responses.
Why is there no date range on reviews?
The get_yext_reviews tool takes none. It returns the breakdown as it stands when your agent asks. To see a trend, save that breakdown on the same day each month and have your agent compare two saved copies. Listing searches, views and actions work differently: pick one of the preset ranges, anything from a single week up to twelve months.
Doesn't Yext already have my Google Analytics numbers?
No, Yext says in writing that it gets no analytics directly from Google Analytics; see what Yext reports and what it never sees.
Can a later Yext import change a snapshot I already saved?
No. A snapshot is a file you keep outside DuckDive, so nothing imported afterwards can rewrite it. Only the live breakdown moves. That is the point: the saved copy stays fixed as your baseline, and the gap between it and today's answer is the net movement, edits and removals included.
Sources
- Location Listings Metrics, Yext Help Center, checked November 16, 2026.
- URL builders: Collect campaign data with custom URLs, Google Analytics Help Center, checked November 16, 2026.
- Google Search Console, Yext Help Center, checked November 16, 2026.
- Reviews Metrics and Dimensions, Yext Help Center, checked November 16, 2026.
- May '26 Release, Yext, checked November 16, 2026.
- June '26 Release Notes, Yext, checked November 16, 2026.
- Yext MCP Overview, Yext Help Center, checked November 16, 2026.
Last reviewed . Numbers in EXAMPLE DATA blocks are fictional.
