"The aim of marketing is to know and understand the customer so well the product or service fits him and sells itself"
Peter F. Drucker
Analytics For Marketing
Analytics for marketing brings some of the most widely used machine learning tools to your portfolio to help optimize your decision making and strategies. Being able to fine-tune your customer's behavior and predictive the right offerings will establish a competitive advantages and transform your organization to a data-driven organization.
Customer Lifetime Value (CLV)
What is CLV?
The basic definition of customer lifetime value is the expected return of a customer over their lifetime; the fundamental purpose of CLV is to create a unique view of each customer's value and derive a ceiling to customer acquisition spend. The analytical derivation of 'value' is likely to vary across business types and the intended use. A retail business may may have a completely different equation for CLV than a hotel chain.
Regardless of industry, everyone is looking to gain insights into their customer-base to grow and nurture their business. Below are some other great resources for CLV overview.
How Can You Use It?
Sure, it's great to have the numbers, but execution is key; to many executives, CLV is just another buzzword or thinks it's too difficult to implement.
Data mine the most (and least) profitable customers
Knowing the most profitable acquisition channels
Manage your database as a portfolio of wealth
Forecast future revenue based on current customers
Predict a customer's value from the onset
Proactively target risks and opportunities
Personalize marketing campaigns
Customize rates and terms
Correlate with third-party data to drive insights to margin
View Sample Playbook Execution
Pockets of Growth / Decline.
What are Pockets of Growth/Decline?
Pockets of growth can be viewed as an early indicator or an algorithmic way of managing the lowest levels of the business. Most organizations will track and assess the fundamental market segments of their business, for example, Direct, Third-Party, Web, etc..., or by product; however, being able to understand potential risk and opportunities, understanding the trends of the all permutations of your business is crucial.
That is a lot of data and trends to really look at, and we get it, not enough time or bandwidth to really understand all subsegments of the business. That's why we use computer power and statistics to really narrow down on major segments of the business that are trending up and that are trending down.
Product Offer-up & Anticipation
What is Product Offer-up?
If you are a business that have customers purchasing multiple items, there opportunity to increase revenue through cross-selling, recommendations, promotions, and product placement.
Just like Amazon and Netflix offers you items they think you'll like or items that are commonly purchased, you can have access to that level of insights and algorithms for simplified cost.
By 'offering-up' recommendations, you can increase your average order value (AOV) and items per ticket. Furthermore, this can be used to optimize your bundling strategy as well as creating great talking points for your call center.
By reverse engineering who are buying what combinations of high frequency set of combined products, we can help you understand who they are and what they do. These results can improve your target-marketing efforts, yield higher conversion rates, and provide valuable insights to value proposition.
How Does It Work?
One of the most underutilized tools in machine learning is a method called market basket analysis, this is essentially how supermarkets learned to optimize their layouts and learned men who buys diapers generally buy milk and beer with that purchase.
We aggregate all your customers with all their purchased items on their unique tickets and understand the patterns of which items are often purchased with another items.
What are propensity models?
Generally speaking, propensity models are probability models that helps identify who among your audience is most susceptible to an event. Or what combination of elements can lead to maintenance failure.
The outcome of these models are contributors or influential factors that helps define the why and how sensitive certain parameter are to the resulting event.
The outcome are contributers or influential factors that helps define the why and how sensitive certain parameter are to the resulting event.
The definition of an event can vary:
The event of a customer:
stopping payment (probability of default)
making a purchase
accepting an offer
yielding low margins