Privacy and how IP Frog handles speed-test data
What stays in your browser, what an anonymous test event contains, what IP Frog deliberately leaves out, and how future aggregate reports will work.
A useful speed test needs measurements. A useful ISP report needs enough measurements to find a pattern. Neither goal requires publishing a trail of individual IP addresses.
IP Frog separates the result you see, the history saved on your device, and the small anonymous event used for aggregate analysis.
Your detailed history stays local
After a completed test, IP Frog keeps up to ten recent results in your browser's local storage. That makes the comparison cards and charts work when you return on the same device.
The locally saved entry can contain the headline summary, chart samples, quality classifications, test mode, and completion time. It is not synchronized to an IP Frog login because there is no user account. Clearing the test history removes those entries from that browser.
The analytics event is intentionally smaller
A completed-test event may include:
- Rounded download and upload speed
- Latency and jitter
- Test duration and light/full mode
- Streaming, gaming, and video-call quality labels
- ASN and network organization
- Approximate city, region, and country supplied by Cloudflare
- The Cloudflare edge location that handled the event
This is enough to answer product and aggregate questions: Are tests finishing? Which networks have enough samples for a regional summary? How do typical results change over time?
What is deliberately excluded
IP Frog does not send the following fields with the PostHog speed-test event:
- Public IP address
- Latitude or longitude
- Postal code
- Reverse-DNS hostname
- Full browser user agent
- Raw chart sample series
- The encoded shared-result URL
Each run gets a random per-test identifier. The start and completion of that test can be connected, but the identifier is not reused as a permanent browser profile. PostHog is also instructed not to create a person profile from the event.
Why include approximate city and network?
Internet performance is local. A nationwide average can hide regional capacity, routing, technology, and congestion differences. ASN and approximate region make more honest grouping possible later.
Approximate IP-based location is not GPS. It can point to a nearby city, provider gateway, or corporate/VPN exit rather than the device's physical location. That limitation belongs in any report using it.
How public reporting should work
Individual rows are not the product. Grouped patterns are. Before publishing a provider-and-region segment, IP Frog plans to require a minimum sample count, show that count, use robust statistics such as medians and ranges, state the time window, and exclude obvious VPN or cloud-hosting networks where possible.
Small groups will stay unpublished. Reports will be described as observed IP Frog results, not a promise about every customer or plan.
Advertising is a separate choice
If Google AdSense is activated, Google's certified consent platform will handle required advertising choices in supported regions. Advertising cookies are separate from the per-test analytics design described here. The footer provides links to the full Privacy Policy and, when active, privacy and cookie settings.
Privacy is not one switch. It is a series of decisions about which details are actually necessary. IP Frog's default is to keep that list short.