Know where your customers are coming from. 6 sigma approach in data-driven marketing.

Hey, PH folks! Today I wanna share some useful info about how you can analyze the data using 6 sigma approach and find strong/weak points of your lead-gen channels. There's an awesome approach in marketing and product management called "6 sigma". This rarely used by companies approach is a very powerful tool to analyze and improve your traffic/lead generation system. Imagine you're working in a game development company and running ads (or any other traffic/lead generation source like loyalty programs, influence marketing, collaborating with Twitch streamers, whatever). You've set all the necessary analytics to track your KPIs and are now able to track the data changes from day to day. Let's say, you're analyzing your installs. 3 days ago you had 150 installs, 2 days ago - 140, yesterday - 170, today - 120. Can you objectively evaluate how well your traffic/lead generation works according to such a chaotic graphic? You'll probably say no. As usual, you are not able to assume if the fall/rise is caused by a quality of your traffic-generation system, or it's just a typical market fluctuation caused be many-many external factors. Should you be nervous about the fact, that you had 120 downloads today? Or should you be happy of 170 download you had yesterday? Such graphics include 2 types of information: 1. How our traffic/lead generation really works + 2. Some external factors in particular day or period of time. But none of the existing analytics tools are able to split them out to give us a better understanding about the external factors. Having an information about them is essential, because you will be able to catch positive external factors and include them in our system on a permanent basis and, on the other hand, exclude negative factors. Here's the photo of the chart: https://ibb.co/XWGf43S 6 Sigma approach: 1. Calculate average value. Sum all the values and divide by the days/months in the period. 2. Calculate the range between the maximum and minimum value. 3. Do the same as in step 2 for other periods/months. 4. Sum all the ranges and divide by number of months. It's called "average span". 5. Use Shewhart charts to find the corresponding row in d2 column according to the amount of months you are analyzing row 6 in d2 column, if you're analyzing 6-month data: 6. Divide "average span" from step 4 by d2 coefficient. The result you'll get is called "sigma" Σ. 7. Let's say, your sigma is 60. Now, use your graphics and put 3 sigmas above your average span value, and 3 sigmas below. At the point where the graphic touched one of the upper lines – positive external factors played a role. At the point where the graphic touches one of the lower boundaries, negative external factors played a role. Let's say, you've found out, that at the day when graphic touched upper sigma line, you've been hosting a tournament in your mobile game. Now you know what to do - host tournaments more often to increase your installs! Analyze that specific factors and include/exclude them from the system to improve your traffic/lead generation process. Be a sigma 😁