Let’s review some of the scenarios we deal with CRM?
- How to convert a Lead to a Contact?
- How to convert an opportunity to invoice?
- How to resolve a case raised by the customer?
- Who do I send out the marketing materials?
Don’t these questions look so 2010-ish?
We are close to a decade now and 2020 deserves some good questions out of your applications.
Let’s ask these questions slightly in a different perspective?
- Which lead is most likely to get converted to contact?
- Which opportunity is most likely to get converted to an invoice?
- What is the next case that this particular customer will raise?
- Which segment of my customers will purchase if I market?
These questions make a lot of sense now as we have been gathering so much information in the past about leads, contacts, opportunities, invoices, cases etc.,
What we are essentially doing with the Machine Learning is that we are predicting the outcome much far in advance so we concentrate our energy on the correct set of data, rather than diluting on all the data sets – In short, it improves our productivity.
In this post, I am going to talk about an Idea – an idea to transform the Dynamics 365 Customer Engagement platform to something more powerful by essentially morphing it as a doorway to bring Machine Learning to the mainstream population.
Microsoft CRM was introduced to manage customer relationship and the community introduced xRM (Extended Relationship Management) in the past which cut open the CRM like a super man’s jacket and scaled the CRM platform to new heights with its unification strategy.
And, it has had its extra ordinary days.
The term xRM was coined brilliantly in my opinion. It clearly indicates that it is not about trying to add layer upon CRM to make the CRM evolve to new business requirements rather than just being a Customer Relationship management software – but it is about transforming the native capabilities of the relationship management and scale to any business requirements, thereby completely replacing CRM with much more powerful idea.
This time around, what if we scale the CRM with “Intelligence” to make it more relevant to the current business practices.
Let’s introduce something called iRM – Intelligent Relationship Management.
What is this iRM?
Dynamics 365 Customer Engagement already offers a few scenarios that provides a sampler of machine learning and its utility.
Some of the preview features include;
Automatically suggesting Knowledge base articles to your customer service representatives so they’ll spend less time searching for the right KB article to help their customers.
Suggest similar cases to your customer service representatives, so they can resolve the incoming cases quickly.
There are a couple of other functionalities which are going to be turned off in the future (or it may already have been) as the underlying cognitive service that powers these functionalities are being turned off.
These functionalities that even though built in the CRM, it was more of an experimentation at this point.
With the current trajectory of business requirements and the technological innovations, it is not terribly complex to build a machine learning platform on top of the Dynamics 365 to enhance the business productivity.
CRM was originally introduced with an idea to build a system that helps the business with “Lead to Cash” model. The idea is still valid, and looks like the current Customer Engagement platform does exactly that, but in retrospect, there is a piece of pie missing, which is the predictive behaviour of the platform.
Almost all the scenarios that CRM offers can be augmented with the machine learning models.
The idea here is to build and maintain a free catalog of machine learning models for all of the customer relationship scenarios. During the implementation, a developer can pick and choose the models that their customer is interested in and easily package them as ‘Intelligent Module’ for each of the CRM’s out of the box modules like Sales, Service, Marketing etc., and deploy.
Sales module to have capabilities like
- Lead classification and conversion prediction
- Improve sales forecasting
- Opportunity prioritization
Service modules to have capabilities like
- Automatic case prioritization
- Predicting field equipment failures
- Product pairing/recommendation strategy
Marketing modules to have capabilities like
- Lead nurturing
- Churn prediction and steps to retain the customers
- Targeted advertisements
These are few of the functionalities that can be packaged as part of the “intelligent” Customer Engagement modules alongside CRM’s generic module that are being shipped currently.
It certainly does not have to stop here. With the MLaaS (Machine Learning as a Service), we could use predefined models to easily develop our corresponding module for the custom functionalities (at this point we are actually scaling xRM with intelligence).
Every major technology company has a suite of Machine Learning platform that helps abstract the complexity of building models by providing a set of very commonly used algorithms so that we can pick and choose the appropriate algorithm, build machine learning model and deploy.
If you are a fan of Azure, it is a no brainer to opt for Azure Machine Learning to build the intelligent modules. It provides a versatile and long list of pre developed models and caters to almost any problem you throw at it. You are welcome to go crazy and build your own model from the scratch if you want to optimize the algorithm to launch a car into the space.
Overall, it seems absolutely necessary to imagine the technical design and the business processes that supports the implementation of machine learning in the future. So, data transformation that usually a lot of people spend a lot of time before even getting into the machine learning models can be reduced significantly. It becomes our responsibility to educate our customers about the tools that are available readily in the market to take their business to the next level.
We aren’t talking about adding new functionalities to the CRM platform at this point. We are just talking about the next generation CRM that fully understands itself and transform it as iRM that greatly improve productivity, decrease cost and delight customers, which is what CRM is set out to do in the first place in my opinion.
Long Live the Machines!