A Case for Customer’s Network Value (CNV)

In the era of personalization and digital networking, it is imperative that impact of individual level actions is not limited to an individual alone. A customer will speak to her network about her experience with a brand in private and public forums. Businesses thus need to consider the impact of personalized actions on the network.

You, Me and Our Network: Think of a situation that both you and I take up an insurance policy with premium worth ? 10,000/-. While you made the decision totally on your own, I received advise from you on how much insurance to take and that resulted in me taking policy of 10% higher premium. Had we both not interacted I would have ended up with premium of ? 1,000/- less (10% of 10,000).

So, it is fair to say that you helped generate premium income of ? 10,000 + ? 1,000 = ? 11,000/- for the insurer, while I generated premium income of ? 10,000/-.

(we both get the credit for ? 1,000/- as it required two of us to tango for this ? 1,000/-)

Hence, in-spite of having paid the same premium, it is reasonable to rank you higher in terms of worthiness within our network.

This worthiness or the monetary value that you or I generate as customers can apply to other businesses too (Banks, Telcos, e-Retailer etc.). That my friend, is what Customer’s Network Value (CNV) is all about!

Defining CNV: Organizations today want to use Customer Life Time Value (CLTV) as a decision metric for customer management decisions. CLTV, (as most of you and wiki will say) - is defined as the monetary value of an individual customer comprising of value realized to-date and future expected value. I want us to further divide CLTV into value due to customer’s own actions and that attributable to influence of other people in the customer’s network.

If we add to CLTV the monetary value generated from other customers in the network due to actions of one individual customer, we will have the Customer’s Network Value (CNV). This can be even more impactful than using CLTV.

Approach to Calculate CNV: In a customer network, action and influence are not always uni-directional. A customer will also be impacted by actions of other customers in her network.

We need to know,

i)                   If actions of the customer did lead to action of others or was it done independent of it?

ii)                 Strength of influence that a customer had on others in her network. (Not everyone is influenced equally by the same customer.)

If CLTV(x) is the Customer Life Time Value of customer x, then an attribution model for CLTV of x can be defined as 

CLTV(x) = ∑w(i)*CLTV(x)                                                                                       …(1)    

where,

w(i) represents share in x’s CLTV due to ith customer in its network

∑w(i) = 1 with w(0) representing the share in CLTV due to own actions of customer x

CLTV(x) is the CLTV of customer x

For a customer C1 having only 1 customer C2 in its network, it can be written as

CLTV(C1) = w(C1)*CLTV(C1) + w(C2)*CLTV(C1)                                                  …(2)

The Customer’s Network Value (CNV) for customer x in a network of n customers can be written as

CNV(x) = ∑w(i)*CLTV(i)            (0 <= i <= n) …(3)

where,

CLTV(i) is the CLTV of customer i

w(i) is the % share of customer x in the CLTV of customer i

w(0) = 1 and CLTV(0) = CLTV of Customer x

For a customer C1 having only 1 customer C2 in its network, CNV can be written as

CNV(C1) = CLTV(C1) + w(C1)*CLTV(C2)                                                                 …(4)

Let us look at it with the following example,

Let there are two customers, Charu (C1) and Parag (P1) connected to each other (networked!). They both purchase a new phone of the same brand online. If the sequence of their actions were such that Charu made the purchase first and shared this information with Parag. It can be argued that Charu had an influence on the purchase decision of Parag and hence has a share in the CLTV of Parag resulting due to this purchase.

Since, Charu (C1) alone cannot take all the credit for purchase by Parag (P1), hence the need to define an attribution model for CLTV and the share of influence.

Given that there can be difference in respective strength of influencing ability, let’s assume

Charu’s actions have 10% impact on Parag’s CLTV (w(C1) = 0.1 for CLTV(P1))

Parag's actions have 0% impact on Charu's CLTV (w(P1) = 0 for CLTV(C1))

lets say due to this purchase, CLTV for both Charu and Parag increased by ? 5,000/- each.

With the above assumptions – we can write CLTV for Charu and Parag as follows (refer eq 1 and 2). (for the purpose of simplicity there is no other value to include in CLTV)

For Charu (C1),

CLTV(C1) = 1*CLTV(C1) + 0*CLTV(C1)

5000       = 1*5000 + 0*5000

For Parag (P1),

CLTV(P1) = 0.9*CLTV(P1) + 0.1*CLTV(P1)

5000 = 0.9*5000 + 0.1*5000

The CNV of Charu and Parag can be defined as (refer eq 3 and eq 4)

CNV(C1) = CLTV(C1) + w(C1)*CLTV(P1)   = 5000   + 0.1*5000 = 5500/-

CNV(P1) = CLTV(P1) + w(P1)*CLTV(C1)   = 5000 + 0*5000       = 5000/-

Hence to prioritise between C1 (Charu) and P1 (Parag), preference be given to Charu due to higher CNV in-spite of equal CLTV value for both.

This can also lead to a scenario in which a customer with low CLTV gets prioritized due to her higher CNV which will be due to her much stronger network.

Way Forward: There are complexities involved in ascertaining customer’s network and strength of influence on the network. The approach to determine weights w(i)s for CLTV attribution is something that I continue to evaluate. I will share my perspective in future posts.

Personalization helps to achieve enriched customer experience, engagement and value. Use of CNV would allow to quantify the value of ripple effect that is triggered due to an activity in the customer’s network.

Email: [email protected]     LinkedIn: www.dhirubhai.net/in/goyalgaurav