100 Questions to Ask Your Google Analytics

Most people open Google Analytics with no idea what they're looking for. They stare at a wall of cards, nod like they understand it, screenshot something, and close the tab. That's not analysis. That's tab-hopping with extra steps.

The fix isn't a fancier dashboard. It's better google analytics questions. Data doesn't hand you insight — questions do. A good question turns a sea of numbers into a decision: raise the budget, kill the campaign, rewrite the landing page, fix the checkout.

So here are 100 of them. Real questions you'd actually type, grouped into ten themes of ten. Skim to the section you need. Steal the ones that fit. Use them as your monthly reporting checklist, your QBR prep, or your "what the hell happened last week" starting point.

And then — because pulling 100 answers by hand is a special kind of misery — I'll show you how to just ask them against live GA4 data instead of building reports.

Why google analytics questions beat dashboards

A dashboard answers questions someone thought to ask three months ago. Your client's business moved on. The question you have today — "why did organic conversions drop last Tuesday?" — isn't on any card, so you go digging, exploring, exporting, pivoting in a spreadsheet, and an hour later you have one number.

Questions are portable. They survive GA4's next UI redesign. They work whether you're in GA4 reporting, Looker Studio, or a chat window. Master the questions and the tool becomes interchangeable.

Two quick definitions before the list, because they trip everyone up:

• What is a session in Google Analytics? A session is a single visit — one continuous group of interactions a user has with your site within a time window (GA4 ends it after 30 minutes of inactivity by default). One user can have many sessions.
• What is bounce rate in Google Analytics? In GA4, bounce rate is the inverse of engagement rate — the percentage of sessions that were not engaged. An engaged session lasts 10+ seconds, has a conversion event, or has 2+ pageviews. So a "bounce" in GA4 means none of that happened. It's not the old Universal Analytics single-page metric, and confusing the two is the fastest way to misread a report.

Now, the questions.

Acquisition & traffic questions

Where people come from, and whether more of them are showing up.

1. How much total traffic did we get last month vs. the month before?
2. Which channels drove the most users last quarter?
3. Is organic search traffic growing or shrinking year over year?
4. What are my top 10 referral sources right now?
5. How much of my traffic is direct, and has that share changed?
6. Which social platform sends the most engaged sessions?
7. What's my new vs. returning user split this month?
8. Which countries or regions are growing fastest in sessions?
9. How much traffic comes from mobile vs. desktop vs. tablet?
10. Did any single source spike or crash week over week?

Engagement questions

Whether the traffic you're paying for actually does anything.

11. What's my overall engagement rate, and how does it compare to last month?
12. What is bounce rate in Google Analytics for my top landing pages?
13. Which pages have the highest average engagement time?
14. How many engaged sessions per user am I getting?
15. Which channel produces the most engaged sessions per dollar?
16. What's my average session duration trend over the last 90 days?
17. Which content keeps people on the site longest?
18. Where do engaged users drop off before converting?
19. Are mobile users less engaged than desktop users, and by how much?
20. Which campaigns bring high traffic but low engagement (aka wasted spend)?

Conversions & revenue questions

The only google analytics metrics your client actually cares about at renewal.

21. How many conversions did we get last month, and from which channels?
22. What's my overall conversion rate, and is it trending up?
23. Which landing pages convert the best?
24. What's the revenue per channel over the last 30 days?
25. Which campaign has the highest return on ad spend?
26. How long is the typical path from first visit to conversion?
27. Which keywords or search terms lead to conversions?
28. What's my cost per conversion by channel this quarter?
29. Which conversion events are declining and need attention?
30. How much revenue is attributed to organic search vs. paid?

Landing page questions

The doorways. Most agencies never audit them properly.

31. What are my top 20 landing pages by sessions?
32. Which landing pages have the worst engagement rate?
33. Which landing page drives the most conversions per session?
34. Are any high-traffic landing pages converting poorly?
35. Which landing pages lost the most traffic month over month?
36. What's the entrance-to-conversion rate for my paid landing pages?
37. Which landing pages have high traffic but zero conversions?
38. How do landing page metrics differ by device?
39. Which blog posts act as landing pages and then send people deeper?
40. What's the average load-affected bounce on my key landing pages?

Audience & retention questions

Who these people are and whether they come back.

41. What's the age and gender breakdown of my converters?
42. Which audience segments have the highest lifetime value?
43. What's my 7-day and 30-day user retention?
44. How many users came back more than once this month?
45. Which cohort of new users retained best over 8 weeks?
46. What interests do my highest-value users share?
47. Which cities or regions have the most loyal returning users?
48. Are returning users converting at a higher rate than new ones?
49. What's the churn pattern for users who converted once?
50. Which acquisition channel brings the stickiest audience?

E-commerce questions

For anyone selling online, this is the section that pays the invoice.

51. What's my total e-commerce revenue this month vs. last?
52. Which products generate the most revenue?
53. What's my cart abandonment rate, and where do people drop?
54. What's the average order value by channel?
55. Which products get viewed a lot but purchased rarely?
56. What's the checkout completion rate step by step?
57. Which promotions or coupons drove the most orders?
58. What's the purchase conversion rate for returning vs. new customers?
59. Which product categories are growing fastest in revenue?
60. How does mobile purchase behavior differ from desktop?

Channels & campaigns questions

Where the budget goes and whether it's working.

61. Which marketing channel has the best conversion rate right now?
62. How is my Google Ads traffic performing vs. organic?
63. Which paid campaign has the lowest cost per conversion?
64. Are my UTM-tagged campaigns tracking correctly?
65. Which email campaigns drove the most on-site engagement?
66. How does paid social compare to organic social for conversions?
67. Which campaign had the biggest week-over-week swing?
68. What's the assisted-conversion contribution of each channel?
69. Which referral partners actually send converting traffic?
70. Where should I shift budget based on last month's return?

Content questions

For content and SEO teams — pair these with your keyword data.

71. Which blog posts drive the most organic traffic?
72. What content leads to the most conversions downstream?
73. Which pages have high impressions but low engagement?
74. What's my most-read content that has no call to action?
75. Which older posts are quietly declining and need a refresh?
76. What content do returning users read most?
77. Which pages have the highest exit rate mid-funnel?
78. What topics keep people clicking to a second and third page?
79. Which content performs on mobile but flops on desktop, or vice versa?
80. What's the traffic-to-lead ratio of my pillar pages?

Technical & site speed questions

Because a slow site quietly deletes your conversion rate.

81. Which pages have the slowest load times?
82. Is site speed hurting conversions on any key page?
83. Are there pages with a sudden spike in errors or 404s?
84. How does page performance differ across browsers?
85. Which device types have the worst engagement due to speed?
86. Are Core Web Vitals affecting my highest-traffic pages?
87. Did a recent deploy change bounce or engagement anywhere?
88. Which pages have broken or missing event tracking?
89. Is my tag firing correctly across all templates?
90. Are bot or spam sessions inflating any of my numbers?

Anomalies & alerts questions

The "wait, what just happened" section. Ask these weekly.

91. Did any metric drop more than 20% week over week?
92. Which page had an unexpected traffic spike, and why?
93. Did conversions fall off a cliff on any single day?
94. Is any channel suddenly sending unqualified traffic?
95. Did a traffic source disappear entirely overnight?
96. Are there seasonality patterns I should plan around?
97. Which anomaly this month is worth flagging to the client?
98. Did a competitor campaign or news event move my numbers?
99. Is revenue tracking diverging from my payment processor's totals?
100. What's the single biggest change since last month I can't explain yet?

You could pull all 100 by hand… or just ask

Read that list again. Every question is answerable in GA4. And every answer costs you the same tax: open the property, find the right report, set the date range, add a comparison, apply a segment, maybe drop into the Explore module, export to Sheets, pivot, screenshot, repeat. Times 100. Times every client.

That's the job most agencies quietly hate. It's not analysis — it's data janitor work. And it's exactly what ARLO exists to delete.

ARLO is the MCP layer for agencies. It connects Claude Desktop directly to every client's live marketing data — GA4, Search Console, Google Ads, Meta, YouTube, Google Business Profile, Shopify, Stripe — through the Model Context Protocol. Which means you stop reading Google Analytics and start talking to it.

You type question 27 — "which search terms led to conversions last month?" — into Claude, in plain English, and ARLO passes it through to the live GA4 API and hands back the answer. No report to build. No CSV to export. No tab to hop to. It's a pure pass-through connector, so it stores nothing — your client data never sits on someone else's server. One Google OAuth grant connects every client, and every team member gets their own MCP token.

That's how to read Google Analytics in 2026: you don't. You ask, and the data reads itself to you.

Need the answer to live somewhere your client can see it? ARLO also pushes data out to destinations — live Looker Studio dashboards, Google Sheets, BigQuery, Slack digests, and branded PDFs — so the same question that answered you in chat can populate a client-ready report on a schedule. Ask once, deliver everywhere.

If your team lives in search, this pairs naturally with the way we think about ARLO for SEO specialists — and if you're building out an AI-first workflow, our sister guide on how to use AI for SEO picks up exactly where this leaves off. Under the hood it's all standard Google plumbing; if you want to see what GA4 actually exposes, the Google Analytics developer docs are the source of truth.

FAQ

What is a good bounce rate in GA4?

There's no universal number, because GA4 bounce rate is just 100% minus engagement rate. A rough rule: engagement rate above 60% (bounce under 40%) is healthy for most content and landing pages. But context wins — a blog post that answers a question fully might "bounce" a lot and still be doing its job. Compare against your own historical baseline before you panic.

What questions should agencies ask GA4 monthly?

At minimum: traffic vs. last month by channel, conversion rate and total conversions, top and worst landing pages, revenue by source, any metric that dropped more than 20% week over week, and the single biggest change you can't yet explain (question 100). Those six cover 80% of what a client wants to know at a monthly review.

How do I read Google Analytics without getting lost?

Start with a question, not the interface. Pick one thing you're trying to decide — where to move budget, which page to fix — and only pull the google analytics metrics that answer it. Ignore every card that doesn't serve the decision. Better yet, ask the question in plain language and let ARLO fetch just the relevant numbers instead of scrolling reports.

Can I ask these questions across multiple clients at once?

Yes — that's the point of building for agencies. With ARLO, one OAuth grant connects your whole portfolio, so you can ask a question of one client's GA4 or compare the same metric across all of them without logging in and out of properties. Your team members each get their own token, and client data stays isolated.

Start asking — it's free right now

The list above is a checklist you'll use for years. The faster way to work through it is to stop pulling answers by hand and start asking your data directly.

ARLO is completely free while we're in early access — every connector, every destination, no card required. Connect a client's GA4 in a couple of minutes and ask your first ten questions before your coffee's cold. Start free at /welcome and see what "ask, don't export" actually feels like.