Understanding User Screen Views Before Login: Insights Needed | Mixpanel Community

Understanding User Screen Views Before Login: Insights Needed

I'm trying to understand how often users see the two screens before they log in. I'm using page load events to track this, but I'm seeing some confusing numbers. For the first screen, 23 users landed on it but weren’t identified — we only have their device ID. On the second screen, 51 users appear as not identified. That doesn’t quite add up, since only 23 unauthenticated users were on the first screen. This suggests the numbers are inflated, likely because we're tracking based only on device ID. I’d like to know: is it possible to measure how many times a user sees each screen before login? And if not, what’s the best way to interpret these results?

2 comments

kapa.ai
APP
·
·
Hi event, Please stand by as I search the Mixpanel knowledge base for an AI-powered answer. I normally take about 15 seconds or so. In case AI cannot answer your question, feel free to reach out to Mixpanel support at mixpanel.com/get-support. If you have follow-up questions for this thread, please tag me again with your question: @kapa.ai <follow-up question> 👋
Feel free to ask a follow-up question once I've answered by mentioning me with @kapa.ai
kapa.ai
APP
·
·
You're correct that tracking page views before login using only device IDs can lead to confusing or inflated numbers. Here’s why, and how you can interpret or improve your measurement:
Why the Numbers Don’t Add Up
- Device ID ≠ Unique User: Before login, Mixpanel uses the device ID ($device_id) to identify users. If a user clears cookies, uses incognito mode, or switches devices/browsers, they’ll get a new device ID. This means the same person could be counted multiple times as different "anonymous users" across sessions or devices. - Session Multiplicity: If a user visits the first screen multiple times (e.g., refreshes, navigates back and forth), each visit is counted as a separate event tied to the same device ID, but you may not be able to distinguish unique "users" from unique "sessions" without additional logic. - Screen Flow Drop-off: The fact that you see more unidentified users on the second screen than the first suggests that some users are skipping the first screen (e.g., deep linking, bookmarks, or navigation quirks), or that device IDs are being reset or duplicated.
Can You Measure Unique Pre-Login Views Per User?
- With Only Device ID: You can only measure unique device IDs that saw each screen before login, not unique people. This is a limitation of anonymous tracking—device ID is a proxy for a user, but not a perfect one. - After Login (ID Merge): If you implement Mixpanel’s ID Merge correctly, when a user logs in and you call .identify(user_id), Mixpanel will retroactively link their pre-login (device ID) and post-login (user ID) events. This allows you to analyze, for logged-in users, how many times they saw each screen before logging in. However, for users who never log in, you’ll always be limited to device-level granularityIdentifying Users (Simplified).
Best Practices & Recommendations
- Call .identify() on Login: Always call .identify(user_id) as soon as the user logs in. This links their anonymous (device ID) and identified (user ID) events, letting you analyze the full pre- and post-login journey for each userIdentifying Users (Simplified) – Best Practices. - Interpretation: When looking at pre-login screen views, treat the numbers as "unique device IDs" rather than "unique users." Expect some inflation due to the reasons above. - For True User-Level Analysis: Focus on users who eventually log in. After ID Merge, you can query: "How many times did each user see screen X before their first login?" This gives you a more accurate picture, but only for users who authenticate at some point. - QA Your Implementation: Ensure your ID management is set up correctly to avoid duplicate profiles and to link pre- and post-login events as intendedIdentifying Users (Simplified) – Best Practices.
Summary
- You cannot reliably measure unique people seeing each screen before login—only unique device IDs. - Numbers may be inflated due to device ID resets, multiple devices, or session quirks. - For more accurate analysis, rely on ID Merge and focus on users who eventually log in, as their pre-login activity will be linked to their user profile.If you need to analyze the full journey, ensure ID Merge is implemented and use queries that look at pre-login events for users who have logged in at least once. For anonymous-only users, accept that device ID is an imperfect proxy for a person.For more details, see the documentation on Identifying Users (Simplified) and Identity Management Best Practices.
kapa.ai is still learning and improving, please let me know how I did by reacting below.
ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help.

2 comments

Why the Numbers Don’t Add Up

Can You Measure Unique Pre-Login Views Per User?

Best Practices & Recommendations

Summary