Digital Marketing - Reference Architecture

Digital Marketing has been a hot topic of discussion for quite some time and yet it is difficult to find a Reference Architecture for Digital Marketing which is industry agnostic. Many players in the ecosystem such as product firms, consulting firms, etc have attempted to draw out a diagram from their perspective and in this article I put forth a reference architecture which may be helpful to anyone trying to understand the complexities in this space and the challenges that need to be addressed in each layer of the platform architecture.

1.????Data Layer

This comprises of all types of Customer related information that is collected from different front office systems or as alternate data from external partners. Some examples of what this Customer data comprises of are the profile related information, transaction behavior, products availed, spending pattern, digital activity, etc. The challenge in this layer is to create a 360 degree view of your Customer and various organizations have implemented CDP or MDM solutions in this regard. If you’re building this layer in house, you need to ensure that you have created a star schema based data model with Customer dimension table and other relevant tables which can give you this 360 degree view. In large organisations with many subsidiaries, such information about your Customer is scattered in multiple systems for their respective products and getting this information in one place is a foremost requirement and many firms have implemented an enterprise data lake for this important purpose.

2.????Analytical Layer

This is the intelligence layer that every firm needs to build inhouse using the raw data that you can collect from various sources in your underlying data layer. This is how every firm can differentiate itself from the peers. Here you use segmentation techniques for example to create personas of your Customers, you can model their behavioral pattern, arrive at the propensity score for each product or a combination of products, pre-approved offers, etc.

3.????Personalization Layer

This is the rule engine which decides which nudge should go to each Customer at what time. Gone are the days when you had a standard set of nudges for a large set of Customers. In today’s digital & hyper personalized world, you need to have a carefully created message for each Customer based on their needs and context. We should also consider the sequence of nudges pushed to a Customer to ensure that the Customer has a smooth onboarding journey without being bombarded with random campaigns. This layer basically consists of two important functions

??Nudge Manufacturing – Here the main focus is the creation of the right nudge for each Customer based on the insights derived from the analytical layer and also based on any specific goals that you have set for the organization.?

??Nudge disbursal – Here the main focus is on arriving at?the right channel for each Customer based on their past behaviour as well as on the budget / cost constraints that may be there at any point of time.

4.????Distribution Layer

This is the channel end point layer where the decisions to push the nudge to different channels are taken. If you have a custom channel or a banner on your mobile app landing page, then you can take decisions like the top 5 nudges to be shown to each Customer at any point of time. Various organizations use CMS solutions to manage & publish content on their websites and mobile Apps. Some organizations have implemented real time website / app monitoring tools which can send real time nudges whenever there is a drop-off or a certain user activity as a trigger.

5.????Channel layer

This is the actual interface with your Customer and this can happen broadly in two ways

a.????Outbound / push channels

???????????????????????????????????????????????????i.????SMS

??????????????????????????????????????????????????ii.????Email

????????????????????????????????????????????????iii.????Mobile App Push

????????????????????????????????????????????????iv.????Social

??????????????????????????????????????????????????v.????WhatsApp

b.????Inbound / Pull channels

???????????????????????????????????????????????????i.????Mobile banners

??????????????????????????????????????????????????ii.????Call centre

????????????????????????????????????????????????iii.????Kiosks

????????????????????????????????????????????????iv.????Websites / landing pages

??????????????????????????????????????????????????v.????Mobile App

The cost associated with campaigns on each of the above channels is different and hence you need to pick your right channel to optimize your costs and get the bang from the buck. ?

6.????Performance Layer

Any campaign or personalized nudge is of no use if you can’t measure its effectiveness. There is a lot of effort organizations put towards creating an infrastructure to collect the Customer response data from the front end interfaces / touch points. For example, the Click stream data is collected on real time basis and analysed to ascertain which campaigns are getting the maximum clicks and which aren’t. Many tools exist today to collect the Customer response data on real time basis and as part of this layer you need to connect the dots between your various campaigns and the final conversions and this will give you an insight around the approach to be taken for future marketing activities.