When you're dealing with small visitor numbers individual user tracking might seem very important. However, aggregate behavior is more insightful as it also smo

When you're dealing with small visitor numbers individual user tracking might seem very important. However, aggregate behavior is more insightful as it also smooths out outliers. Eg. 3% of all iPhone users traffic reach goal X every month; versus, this 1 user that visited the site 25 times in the last six months, 3 times through a VPN, 17 times from public, open access Wi-Fi, 5 times from a different country, reached goal X after spending 27 seconds on page Y. The first one is directly actionable insight. Increase your reach to iPhone users, as more iPhone traffic leads to bigger conversion numbers. While the later doesn't lead directly to any actionable insight, but a bunch of theory crafting, while stripping away any form of privacy (and being a juicy target for intelligence agencies and hackers). Now, that is of course my opinion. Having touched projects that had collected both "lightweight" analytics data, and everything the user touches. I'm sure the later could lead to more experimentation, but in the context of that project it amounted to just hoarding. Instead, for other types of actionable insight I would suggest surveying converted users. Of course there can be selection bias, and unless the UI is done right can be obtrusive UX. This is the exact information more data would help you answer (in theory), but rather than asking the user you're trying to read their minds by tracking everything about them. At least, again, my opinion based on experience.