Desire to understand the effect of customer touchpoints
Introduction
Knowing the correlation between customer journey and related conversions. Correlations that randomized, prepared experiments never respond.
Lets attribute
Let us begin with the subject of interest. It is the customer journey through available touchpoints. These customer journey touchpoints become attributes assigned. Attributes are properties describing a specific purpose e.g. spends per touchpoint, the number of conversions per touchpoint.
Game theory
It is about playing the marketing game of involved players, our customer journey touchpoints, and decisions affecting the whole game, our customer journey conversion. Assuming the purpose of optimizing aka maximizing the conversion. In-game theory, this approach is called a cooperative strategy.
Markov Chain
One way of measuring the maximum conversion output is to calculate the probability that the conversion happens. Let us assume two customer touchpoint journeys.
Now we the Markov Chain the probability movement through the customer journey is mapped.
Note: The probability can be determined in two ways
- Logically via distribution e.g. two outgoing arrows is a 50/50 probability
- By measuring e.g. the sketeched 40/60 distribution
Removal effect
The probability that a conversion, coming from T0, happens is
P( Convert ) = P( T1→ T2 → T3 → Convert) + P( T2 → T3 → Convert) = 0.5 x 0.5 x 1.0 x 0.6 + 0.5 x 1.0 x 0.6 = 0.45 = 45%
Now we have to determine the contribution of each involved customer journey touchpoint or the impact of its removal:
T1: P( T2→ T3 → Convert) = 0.5 x 1.0 x 0.6 = 0.3 = 30%
T2: P( T3 → Convert) = 1.0 x 0.6 = 60%
T3: P( Convert ) = 0.6 = 60%
Now the contribution margin
T1: 0.3/ 0.45 = 0.66 = 66%
T2: 0.6/ 0.45 = 1.33 = 133%
T3: 0.6/ 0.45 = 1.33 = 133%
This implies that the most marketing spending is to invest in T1, because it has the most loss.