Integrating online and offline worlds

A few years ago I had written a post of my experience with a general insurance company. (Same Company ...Same Customer). This post highlighted how the company was giving different customer experiences across different channels. 

I have noticed a recent trend among my colleagues and myself included. Even when one purchases a product at the store outlet, one often visits online retailers to check options, comparisons and prices. In most cases, this online research also includes the web site of the outlet from where the product is eventually purchased. 

I have been working out on how the two worlds can be integrated for a retailer. The toughest format to integrate is the super market. These are stores with a high RFM value. I had a passing mention of this in my post on customer captivity (Aim for captivity .. not loyalty).  The relevant excerpt from this post is pasted below:

"Grocery purchases are often a chore rather than fun activity. A grocery retailer could allow a customer to define her basket of regular purchase. Then have an SMS facility wherein the customer sends in her request for a particular basket and have it delivered to her home. What a convenience that would be? Would this customer want to go over the pain of defining her baskets with another retailer... highly unlikely."

While the thought started with an idea, overtime I have fleshed out the model of integration between online and offline formats. The key aspect of this integration is unified customer experience across the channels. There should be a scenario wherein the customer feels that certain activity can only be done on the online or the offline format. I will attempt to highlight key aspects of the same in the post. 

• The customer gets registered over the net or at the store. The mobile number is used as the identifier of the customer.
• The customer can create shopping list over the net and store it. She can also give it unique and meaningful names..such as weekend, monthly, household, etc. 
• The customer can also create a shopping list by SMSing the receipt number to a predefined number. The backend system would retrieve the purchase basket of the receipt number and store the basked as a shopping list against the customer account. The customer can copy or modify the list as per requirement. Also, the feature to specify brands or leave it open for each item is available to the customer. 
• When the customer needs to shop for specific items, say household items on a monthly basis, the customer will need to send an SMS to the predefined number with the shopping list name or number. The customer can also order for the shopping basket over the net. Based on the delivery preferences, the basket is then delivered to the customer. 
• Over time, the purchase pattern can be identified for each customer. Using analytics, one can predict the likely basket and the time of purchase. As this time nears, the retailer can prompt the customer for upcoming need. This could also be a reminder call for the customer who probably just needs to confirm the basket and have it delivered to her. 
• In order to increase the basket, the retailer can used market basket analysis to recommended additional products to be added to the basket. Also, personalized offers could be presented to  the customer. 
• The customer can also order a shopping list over the web and opt to have it ready for pickup at a particular outlet. While at the outlet, maybe she wanted additional items which she wants to inspect before purchase. This also gives opportunity for impulse purchase while she is at the store. The customer saves time picking up items of regular purchase in her basket. 
• On lines of fast food joints, the above basket could also be in a drive-in lane where the customer can pick it up at the take-away counter and pay for it without leaving her vehicle.
• The retailer could also offer products on trial to entice purchase. This is corollary to the product association analysis that determines the next best product. In this case we look at products that are least likely to be bought by the customer but one that if enticed the product has best chance to be bought. For example, the profile of the customer shows that she has a school going youngster in her household. Our basket analysis shows the "school" related products that the customer has bought. The disassociation analysis also shows the products that the customer has not bought and is least likely to buy. A conditional modelling can show the product amongst this second set that has the best chance of being bought if the customer is enticed with the product. Such a product can be given to the customer on a trial basis to be returned within, say, 15 days if the customer does not want it. The customer can opt to purchase it and be billed in the next purchase basket.  
• Additional service, such as toileteries and personal effects can be made available in different cities to be delivered to the customer when she is travelling. 

If any retailer wants to run a pilot on this, kindly contact me at michaeldsilva@gmail.com. Together we can define an appropriate process for blurring that online / offline demarcation and provide a unified and enhanced customer experience.