Saturday, 30 March 2013

Case Study: Customer Analytics for Telecom Operator to Cross-Sell and Up-Sell

How to find products/services to Cross-Sell / Up-Sell using Customer analytics / NBO

The Telecom operator has a good customer base and wants to drive more sales using existing customers. The operator decided to craft a NBO strategy for increasing sales from existing customers (this is not just retaining the customer - by reducing Churn)

As with any NBO - the first step is to determine the Objective

Step 1: Define The Objective

The objective here is - Increase sales from existing customer through cross-sell or up-sell

A customer with two or more services is four times less likely to Churn than a customer with a singl service - this means that cross-sell not only increases the incremental sales but also increases Loyalty. The strategic objectives like these should be driving the NBO objective

With the goal being defined very clearly, the next step is to collect required data for the NBO

Step 2: Collect Data

The Telecom operator has data on the services/products that are used by each customer, the plan, the usage history, subscription data,  the CDRs , Customer Demographics data and Internet usage data (for the subscribers) . This data holds wealth of information and needs to be mined.

The entry channel (the channel through which the customer was acquired) would be a vital information.

The Location data also plays an important role in understanding the customer, Location history would explain if the customer is a frequent traveller, when and where he/she makes uses the service.

If the internal data collection is done, one can look for Syndicate data or External Data (Payment Card data)

Step 3: Analyze, Model  and Execute

Once the Data is available, it needs to be cleaned and normalized. Then the customers are segmented based on their purchase behavior, subscription plans etc. There are also industry specific segmentation standards like Prizm. This should be a good input for rolling up data

Understanding the customer is a first stage in NBO. The customer's behavioral attributes are
analyzed using statistical analysis (clustering, link analysis, decision trees), predictive modelling etc.

It's just not about understanding the customer, the Telecom company needs to analyze the product/service that can be offered to a particular segment. Understanding the Service/Product Offering is the next stage .

For example,
  • Wireless devices
  • WiFi Services
  • VOIP Call services
  • Local call services
  • Broadband Direct connection
  • loud hosting

For example, a propensity model can be created for knowing the probability of the customer for purchasing/trying a particular product / service. For example, a Broadband plan for a customer who is not using the services from the operator (and using the one from a Competitor) with competitive pricing / value add would entice the customer to switch-over.

The propensity of a customer who is currently using the service is set to Zero. This will avoid irittating the customer with the product/service she already uses.

The propensity of each service is ranked along with the best suited time of  sending the offer. The Offer is then sent to the customer through his prefered Channel and at the appropriate context.

Context is always the key - if someone has a 4G-enabled phone, he has a fair chance of upgrading to a 4G network. Knowing the user's device information is helpful here.

Tools like SAS and SPSS can help in achieving this faster. There are Open Source Software - like R, which is catching up fast

Step 4: Measure and Recalibrate

Every offer that is sent is an Test Offer. An offer can be accepted/ignored/rejected. The best NBO strategy is to measure the response of the offer and take the learning back to Analysis stage.

The offer response data is fed back to the database. If there is any need for further data points, it needs to be collected. The model is then refreshed based on the new analysis and deployed.

Next Best Offer Solution Framework

This is a continous cycle and the role of Statisticans, Business Leaders play an important role

Thursday, 28 March 2013

Customer Attrition Modeling (Analytics)

Churn or Attrition Modeling

Customer is the key for any business. As Don Peppers once said - the only value that a company adds during the lifetime is the Customer.

Losing customer is losing business. But, how to avoid them moving away?

How to measure Customer Attrition

How to know if the customer is leaving . Probably if he/she is not shopping / visiting / interacting / subscribing like the past, it is an indication that he/she might leave.

Good datamining should measure the customer attrition easily. Predicitve models can be built on data for Customer attrition.

The key data points to be considered for Attrition modeling are the Recency, Latency, Average Spend per Visit, Average Time spent on Website, Duration of Visit (Website) etc

The predictive model then attaches a probabily score to each customer. The least churner to the most ..

How to retain the customer?

Starbucks has its own Loyalty program -My Starbucks Rewards™ . One in four customers use the Loyalty card, which provides a wealth of data for Starbucks analyitcs team. The customers are then analyzed and segmented. The customers who are more Loyal are ignored! Yes, the ones who are about to Churn are Rewarded with appropriate offers. The offers are based on the past purchase history!

Payment Card Industry - Crucial for Real-Time Next Best Action (NBA) - Big Data

How to Collect Data (Big Data) for Real-Time Next Best Action (NBA)

Data is more vital for sending Promotional Messages, Recommendations or Any Action for any Customer / Segment

The main focus for any retailer/bank/airline/hospital is to build a comprehensive customer signature. The customer signature provides vital information of the customer in the form of :

  • Demographics
  • Geography
  • Behavioral
  • Location
  • Purchase Behavior
  • Social behavior
Collecting this data and rolling up accounts to 70% of Customer Analytics. However, there is always a limitation to it. The customer might shop only specific items with the Retailer / avail certain services from the Bank. If the Retailer / Bank is a specialized one (like Apparel/Jewelery) the information for individual customer will be limited to few tens / hundreds of records /year.
Payment GatewayLocation Based Offer / Message using NBA
To overcome this a Bank / Retailer can have alliance with Loyalty firms / use syndicate data etc. The Credit Card / Payment gateway company holds the key.

Take a case of the customer shopping in a high-end Apparel retailer, a Jewelery shop and have paid using Credit card in both places.

The Credit Card company doesn't have the Transaction level information but knows that Customer X has shopped for the following in the same locality

a) Apparel
b) Jewelery

If this information can be used prudently by the speciality footwear store in the same location. Probably the customer might be interested to drop -in provided he / she knows about the store. If the Footwear retailer forms an alliance with Credit card company and push appropriate messages, the sales would definitely go north!

+MasterCard for example collects and stores the transactional data from the customer for data analysis. Customers have the option to opt out.

 Companies like BeyondAnalysis use the Payment Card  (Visa, +MasterCard )'s Transaction data and provide the Retailers like  +Waitrose information about the customers who are purchasing from other supermarkets as well. This information would be limited to aggregate level and not at an individual level there by protecting the privacy of the individual and at the same time generating necessary customer insights

How to Deliver Next Best Action (NBAs) in Real-Time

How to create Location-based Real-Time Next Best Offers (NBOs)

NBAs are supposed to understand the customer's need and fulfil them even before he/she wants to act on it.

The main purpose of the NBA/NBO is to happen real-time. With the advent of smartphones delivering the Action in appropriate location and time is now easy.

Any Long distance passenger have to wait for few hours for the connecting flight; The Airline knows the customer’s demographic info, location, time (both are very much accurate) - a good amount of info for Analytics

The data of all passengers are fed to Analytical system and their past travel history, lounge purchases, social habits and syndicate data is crunched to figure out the need for each passenger. The needs are then matched with the Offerings of the Airline's partners (Hotels, Boutiques, Spas) in that particular airport and appropriate recommendations are created

Real-Time Next Best Offer (NBO) Flow

Timely message / offer is sent to passenger’s mobile phone (e.g., offer on antique jewelry in Airport boutique ) based on the time/location and her Inflight purchase/interaction behaviour. The last part - knowing his/her immediate purchase/interaction is more important. This is fed into the Analytical system to validate against the recommendaion that are waiting to be delivered - the recommendations are refreshed and delivered. This is what makes the 'Offer' - Real-time (Right Product/Service)

The above example uses Mobile Phone as the Delivery Channel. Please refer What are the key channels to be considered for Promotions for other Channels.

The omni-channel world requires product information to be managed consistently. The main focus areas are
• Provide consistent data across all channels and update the information real-time
• Offer product information at the right time and place;
• Compile a complete data set and deliver rich, unique data;
• Optimum utilization of social media information

What are the key channels to be considered for Promotions

How to Find Right Channel for NBO

One of the key feature of Next Best Offer is to provide the right promotion through Right Channel. It is worth to listdown the channels that should be considered normally.

  • call centers
  • direct mail
  • email
  • in-store display
  • in-store staff
  • kiosks
  • mobile devices
  • POS receipts
  • web recommendations
The Channels are not limited to the above.

Tools and Metrics to Understand Mobile Data

Mobile users contribute to more than half the percentage of traffic for many Retailers. The Mobile shopping, however hasn't grown with equal pace

Mobile Data Analytics involves collecting, understanding and reporting the insights based on Mobile data

Mobile Data Collection: 

Collecting mobile data has its own challenges. The variety of devices and the Applications prevent a standards for mobile data collection technique

Tags are key for Mobile Data Mining. Tags are small piece of code embedded inside the apps which relay information back to the provider

Mobile Websites (mobile version of websites) are similar to the normal Website but light weight and simple design for mobile.

Real-Time Next Best Action Based on Mood - Voice Analytics

Real Time Delivery of Offer is always a challenge, whatever the Channel might be. It's even challenging in Call Centrer where the CCE (Call Center Executive) is talking to the customer over phone. Now, Computational Voice analysis is helping the executives identify the mood of the customer on the other end. This is also called as emotion detection / mood detection

Identifying the mood of the customer can help drive more insights into the customer herself. Yuval Mor, the CEO of Beyond Verbal the company which has created mood detection algorithms for call centers says the program can also pinpoint and influence how consumers make decisions. “If this person is an innovator, you want to offer the latest and greatest product,” Mr. Mor says. “If this person is a more conservative person, you don’t want to offer the latest and greatest, but something tried and true.”

The company is also providing APIs for developers to develop Emotional rich apps

+Flurry, Inc., which had been an App analytics platform developer has extended it to Mobile Web as well

Email Marketing - An effective tool till date

Email marketing has been and remains a primary marketing lever for businesses," said Geoff Alexander, +iContact . It's always been a tool and with Dynamic Messaging (Hope your remember Microsoft's Bing campaign)

Wednesday, 27 March 2013

How to Find Social Influencers using analytics

How to Identify the best influencers from a given data

Who are these Social Influencers ?

One can associate a Social Influencer with :

One who has the maximum followers
One whose opinion is considered by his social circle (who can influence others easily)
One who creates and shares content regularly
One who makes most calls; accepts most calls etc

Benefits of identifying Social Influencers

Identifying influential persons within a circle will help the marketers expand the business; for example, they might be selected for a Test Drive (as Ford did), evaluate a product/service (Beta Programs from Microsoft etc), conduct surveys and thereby spread the word.

How to Identify the Influencers

Identifying Influencers is a broader statement. There are scores like Klout score etc available to weigh the influence of one in social media

The Klout score is an useful one for Social Media

However, a more deep data mining exercise would be more helpful. This would be more helpful than a specific score as the business goal of the datamining would pave the way for the most appropriate social influencers for a goal.

For example, if a Telecom group wants to acquire (poach in fact) new customers, it can identify customers who are good influencers and whose circle/friends are from different telecom network etc.

Building a robust dataset is very important. One needs to collect all possible information (from the call records - CDRs ) and roll-up the same for each customer,

Once the Data is created, one can use Supervised learning techniques or Link analysis to find the best scoring customers.

Related Posts Plugin for WordPress, Blogger...