User:Jugdev/sandbox

Attribution

In relation to programmatic media, attribution refers to how one would effectively balance and prioritise specific media inventory with reference to key performance indicators. From a business' perspective, this would mean clear reporting on the impact of media inventory using metrics such as turnover, profit, customer retention and volume of sales in addition to taking a taking a holistic approach assigning value to all parts of the purchase funnel (from research to last click and beyond).

Customised Attribution Model

The customised attribution model combines the use of either linear, first, last, time decay or position based components as a foundation whilst factoring in additional layers that would be unique to the given key performance indicator. Each, or a number of the components below, act as a pre campaign hypothesis and are tested to prove a particular assumed proposition. In some cases, however, one would be accepted as highly probable in the light of established facts (i.e. in the case of one or a limited number of touchpoints).

Linear Hypothesis

The linear attribution hypothesis serves every touchpoint in the conversion path and would commence with the idea that equal credit is to be shared for the conversion.

First Interaction Hypothesis

The first interaction hypothesis suggests that all credit is due to the first touchpoint.

Last Interaction Hypothesis

The last interaction Hypothesis would suggest that all credit is given to the last touchpoint.

Last Non-Direct Interaction Hypothesis

The last non-direct interaction hypothesis suggests that the last non-direct touchpoint is due all credit for a given conversion.

Time Decay Hypothesis

The time decay hypothesis works in a chronological manner and would commence with the idea that the touchpoint closest in time to the conversion gains all credit.

Position Based Hypothesis

In the position based hypothesis, one would commence with the idea that 40% of credit is assigned to each of the first and last interactions, whilst the remaining 20% credit is distributed evenly across the remaining touchpoints.