Lessons learned from writing a bespoke tracking platform
For a recent project I wanted to develop a tracking platform which was specifically designed to suit my needs. It needed to be simple to use, capable of handling seven figure audiences daily and built on technology I understand so it can be customised on the go.
I decided to build it arround php / mysql giving each days data a new table. I also found that alot of the information stored in traditional tracking systems wasn’t needed so I stripped it down to just the basic data I needed to optimise, some c1-4 variables and a few other bits and pieces.
I found that by using separate tables for each days data it stops alot of the problems with slow servers and having to delete data on a daily basis. Don’t get me wrong if I want to run off a report for a weeks worth of data it is still slow, so slow in fact I had to adjust the settings at which php timeouts, but I don’t think there is any way around that when analysing large volumes of information.
Dayparting can be done on a country basis or accumulated so you can see data across multiple countries, adjusted with local timezones.
Trying to keep the code as simple as possible helped avoid duplicate lines in the source. Probably a takeaway for any coding project.
The benefit of this is I now have a tracking platform which is designed to cope with high volume campaigns, which stores only the data I need and which I fully understand so if I want to analyse the data in a new way I can quickly add a new script to do that.
All in all I’m pretty happy with my new toys and would recommend it as a project for anyone in the online marketing game with a decent knowledge of php/mysql.