The Math behind a successful MBA Admission Campaign

Every year, as a b-scgool, you are faced with the biggest challenge of strategically choosing communication channels for your Admission purposes, trying to ensure that you get it RIGHT.

I am aware that over 98% of our b-schools get it WRONG every year. There is a reason they repeat their mistakes, sub-consciously at most times.

You measure the results that come from each marketing channel you choose rather than a multi-channel strategy and measurement. You use certain channels incorrectly, without completely understanding the characteristic consumption pattern of your Applicants in that channel. You define usage of channel by the same prejudices that are wrong and need to be corrected.

So,

How will you figure out muti-channel attribution to your Applications each year?

How will you strategize for something that you don't completely understand?

How will you change from a broken system to a correct-system as soon as possible?

How will you improve effectiveness of your Admission campaign and therefore improve application numbers and maintain that for the coming years?

How will you stop random buying and learn to methodically identify channels that work for you, and the integration of such channels?

Consider the following multichannel strategy used by a b-school

 

 

This means that the bschool is using 6 different but integrated channels for its marketing program.

If we consider that Results (Overall Results i.e.) is a function of engagement and action taken, then we can represent it as:

Result = fn(Action, Engagement)

Now, according to integrated marketing assumptions, the action and engagement can happen in two different channels at a given point in time "t".

Which means, that for a result Rt, the set of action and engagement can be (Pagalguy.com, Database Calling); wherein the action happened in PG but the engagement happened over database calling.

Similarly, at a time "t+1", the R(t+1) can be (Database calling, Pagalguy.com); wherein action happened in database calling but engagement happened in Pagalguy.com.

Since it is impossible to attribute specific inputs to specific channels for results, the best can we can predict is the probability of engagement of a particular platform at a certain time for a result "R".

Mathematically, this would mean-

P(PG) = Probability of student engaging with Pagalguy.com

P(C) = Probability of student engaging over a call

R = Result from the overall marketing system

The Probability of a student engaging with a particular platform is directly proportional for the amount of time he spends on that platform, or the "exposure" to that platform. Therefore, lets assume for our calculation's sake that a student spends 40% of his time on Pagalguy.com and the 20% of the time he is exposed to Calls from b-schools.

Then P(PG) = 0.4, and P(C) = 0.2

We want to identify what will be the probability of engagement on Pagalguy, when there is a Result from the marketing system. This can be denoted as - P(PG | R)

To find this probability, we will use the Bayes theorem (since there are multiple events associated with this probability)

Therefore,

P(PG | R) = [P(PG) * P(R | PG)] / [ [P(PG) * P(R | PG)] X [P(C) * P(R | C)] ]

Now,

We will have to calculate: the probability of a result for engagement happening in PG as well as calls.

P(R | PG) = Time x Activity x (1/Total number of options at a given time)

= 4 x 5 x (1/50)  [Considering 4 hours total time, 5 pages viewed each time, and 50 B-schools as option during his visits]

= 0.2

P(R | C) = 0.1 x 10 x (1/100) [Considering 0.1 hours call time, 10 calls, and 100 bschools calling]

= 0.01

Now, 

The P(PG | R) = [0.4 x 0.2] / [ [0.4 x 0.2] + [0.2 x 0.01] ]

= 0.08/0.082

= 0.97

or 97% probability of Pagalguy engagement when Result is "R" from the marketing system.

Similarly, we can calculate and attribute different probability scores to different sets of "(Action, Engagement") to check how effective is the engagement of a particular medium and how it plays out for the marketing system.

Since, the college in this case is using 6 different channels, and given that there are two factors, the number of sets will be 6P2 (permutation), giving us about 30 such sets that will contribute to the overall results.

These sets need to be measured if one needs to figure out the Effectiveness of Every Channel used. But this is hard and near impossible due to complexity of the relationships and theimprobability to tag an action to a particular channel. All we know is that together, they form a cohesive system that provides the advantages of an "integrated" marketing system and allows us to measure the integrity of "Communication" that actually affect the efficiency of these 30 sets.

 

Start practicing the the most scientific and right strategies for your Admission campaign design, else you will keep on doing the same ad-hoc usage of communications channel and Measure them incorrectly - leading to bad results.