How to create customer delight by using analytics

Technology is shaping business today. It is causing disruption and shifting profitability and creating pressure to perform or perish. The advent of digital and e-commerce companies has changed the ground rules for most businesses today. It is a new world, very different and with very different rules of the game. In this theatre, companies would want to know how data analytics and insights can help them defend growth and profits.

The first and the most important variable is data. Organisations need to streamline the collection, storage and usage of data. Many small companies collect data at multiple points – marketing, sales, operations, after sales service, amongst others. Most of this collection is being done without any specific end use in mind and is therefore at best haphazard. The approach seems to be bent on collecting whatever one could and figuring out the usage later.

My data your data

There also are cross department silos with each silo defending the data they have collected leading to overall an inefficient usage. Organisations should map the data being collected at all points and create a singular database across functions accessible to the organisation.

It is important for organisations to oversee the evolving legal framework and avoid a data backlash. Laws around various subjects are still evolving. Data being collected today can become unusable due to changes in the law in future.

It is important to implement controls at the beginning to take care of future concerns. One should therefore develop systems and marketing programs which are compliant not only with the current law, but also anticipate they will be in line with the direction in which law is evolving.

Rise of cohorts

For corporates who have individual level customer data available, they can do a lot of work in profiling customers. The data is worked with mathematical algorithms to extract identifiable groups with different characteristics. This is referred to as cohorts. Once you have identified these cohorts, it is possible to predict the cohorts with larger growth opportunity.

For example, it is possible to identify those who shop on weekends only or those who visit once a month or those who visit on sale periods or those who buy whenever a new launch happens.

Once you have identified these cohorts, you can very easily run different marketing programs for these cohorts. You can selectively contact them and run focused campaigns.

Social around cohorts 

Using customer analytics for planning digital marketing campaigns is a great idea and not being used by many organisations. A lot of social media and digital marketing campaigns are being planned without your own customer analytics.

Once you have identifiable cohorts and a plan for each of them, you can use social media to deliver these campaigns. You can run retargeting campaigns on your customers through social and digital media channels.

Social media companies also allow you to create additional audiences which are similar to your cohorts. If you have a list of people which are your best performing customers, any digital media platform can help you reach other people in your geography but with habits and disposition very similar to your customers.

At the store

Leading e-commerce channels surprise us with their shopping suggestions. They have a unique ability to know a when you are on their website? What was your past shopping behavior? How to reach you? The same ability can be created for downtown stores by using technology to your advantage.

We can use technology for on-ground stores by inviting customer to give a missed call or click a QR code inside the store and get a promotional offer. The missed call helps us in informing that a particular customer is inside the store. It also helps in identifying the particular customer as well as his past purchases and his profile.

The automatic algorithms at the backend generates specific offers for a customer as per their past history and delivers those on SMS. The engine manages to surprise the customer and suggest products and promotions leading to an immediate growth in sales for the participating stores.

Forecasting 

Forecasting is another area where technology can be used to improve efficiencies. Most companies do have historical timeseries data at a granular level. There are advanced algorithms on platforms like R and Python which can be used for increasing the efficiency of forecasts. The modelling is easy today as these algorithms can pick up multiple trends and generate better forecasts.

For any industry a good forecast can help increase efficiencies or reduce the costs. One can use these to generate daily sales forecasts and use those for planning inventories, planning manpower or even planning promotions.

Modelling behaviour 

Regression and correlation models can be created based on historical data to test the impact of promotion on volume and sales growth. A lot of businesses use promotions to drive sales. Overtime most businesses however drive promotions as a habit without a deep understanding of the direct impact of promotions.

Correlations and regressions models can actually pinpoint, if promotions result in sales growth; which product promotions lead to sales growth; extent of sales growth to decide how much promotion; does a promotion in one cannibalise something else. Most of these are advanced mathematics problems and a study is to be conducted to find directions where analytics can be used to improve profitability.

Key takeaways

• Organisations need to streamline the collection, storage and usage of data.
• Once you have identified these cohorts, you can run different marketing programs for cohorts.
• Regression models can be created based on historical data to test the impact of promotion on sales.
• Automatic algorithms at backend generate specific offers for a customer going by history and delivers on SMS.

By Manishi Sanwal, Managing Partner, Voiceback Technologies.