Diagnosing Sudden Traffic Spikes on Tuesdays: Causes and Solutions
We have sudden traffic spikes every Tuesday on pages that we generated programmatically. How would you diagnose this? Is this bot behaviour and where can it come from?
Hi Thibaut N., 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> đź‘‹
Feel free to ask a follow-up question once I've answered by mentioning me with @kapa.ai
To diagnose sudden traffic spikes on programmatically generated pages every Tuesday, and to determine if this is bot behavior, we can follow these steps:1. Analyze the traffic using Mixpanel's Insights: First, create a visualization of the traffic pattern over time, focusing on Tuesdays. This will help confirm the spike pattern Identifying growth KPIs that are leading indicators of product success.2. Use Anomaly Detection: Mixpanel's Anomaly Detection feature can help identify unusual patterns in your data automatically Introducing Anomaly Detection and Root Cause Analysis.3. Investigate user properties: Break down the traffic by various properties such as operating system, browser, and versions. This can help identify if the traffic is coming from specific sources that might indicate bot behavior Debugging in Mixpanel.4. Check for bot indicators: Look for signs of bot traffic, such as unusual user agent strings or patterns in the behavior. Mixpanel's JavaScript SDK automatically filters out some known bots, but you may need to implement additional filtering Block unwanted bot traffic.5. Analyze user behavior: Examine the actions these users take on your site. Bots often have different behavior patterns compared to real users. You can use Mixpanel's funnel analysis to see if these users are following expected user flows Debugging in Mixpanel.6. Check for recurring tasks: Since the spikes occur every Tuesday, investigate if there are any scheduled tasks, cron jobs, or automated processes that might be triggering this behavior.7. Examine traffic sources: Use Mixpanel's UTM tracking to see where this traffic is coming from. If it's bot traffic, it might not have typical referrer or campaign data Traffic Attribution: Analyze your marketing channels.If you determine that this is indeed bot traffic, it could come from various sources:1. Web crawlers: Search engines or other indexing services might be crawling your programmatically generated pages. 2. Scrapers: Bots designed to extract data from your pages. 3. Automated testing: If you or a third party are running automated tests on Tuesdays. 4. Malicious bots: Attempting to exploit vulnerabilities or perform DDoS attacks.To mitigate bot traffic, you can implement additional filtering in your Mixpanel tracking code, as described in the Block unwanted bot traffic documentation. You may also want to implement server-side measures to identify and block suspicious traffic patterns.
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.