Strong customer relationships are important to the success of an e-commerce business. That's why the e-commerce personalization features are so important.
I went to a neighborhood grocery store, of which I am a regular customer, and asked for a bottle of juice. He gave me a bottle of orange juice with lots of pulp, exactly the one I like.
You are standing in a digital electronics store, wondering which video game to buy for your son - Sony PS3 or XBox 360 or Wii? You are inclined to purchase Sony PS3 but minutes later, a store executive walks up to you and points to XBox 360 console and motivates you to give it a try. Nothing more is said and you walk out of the store with XBox 360 in your hand, passing by the smiling store executive.
These are just two forms of personalization out of many. Personalization is easy to achieve in the physical world. In the virtual world, it is a tad hard. There are no store executives to help, yet you must understand the preferences, needs and intentions of the shopper and help him transact. Sounds easy? I will say may be yes but mostly no.
Let's get to product recommendations, a very interesting form of personalization. The phrases that e-retailers use goes something like this:
What is relevancy? To a shopper, it could mean related products around same price point, similar features and possibly a fast moving item.
This mobile phone costs Rs 2999 and the recommended mobiles are around the same price point. Since the shopper is looking for dual SIM phone, 2 dual SIM phones are recommended. All these 4 products are liked by the users as inferred from reviews and ratings.
Let us take another example where an attempted product recommendations goes bad.
Anomalies are easy to spot - Samsung laptop is similar to Sony PS2 and gaming laptop worth Rs 76,099 does not come close to Sony PS2 costing Rs 5690. And most interestingly, people who bought Sony PS2 also bought a geyser. That is very insightful!
Mostly, product recommendations are generated using rules or computer algorithms. If algorithm driven recommendations are not meaningful, consider manually linking related products or use paid service. It is a safer way to tread. Today's shoppers are smart and they can smell bad recommendations from miles away.
Product recommendations are very useful tools for personalization. If done correctly, it can significantly improve customer experience, conversion rate and repeat purchase.
There are many more decision to make before serving good recommendations. I will cover these in my future blog posts.
More blog posts on E-Commerce:
Social Validation and E-Commerce
Social Commerce: Part 3
Gamification
I went to a neighborhood grocery store, of which I am a regular customer, and asked for a bottle of juice. He gave me a bottle of orange juice with lots of pulp, exactly the one I like.
You are standing in a digital electronics store, wondering which video game to buy for your son - Sony PS3 or XBox 360 or Wii? You are inclined to purchase Sony PS3 but minutes later, a store executive walks up to you and points to XBox 360 console and motivates you to give it a try. Nothing more is said and you walk out of the store with XBox 360 in your hand, passing by the smiling store executive.
These are just two forms of personalization out of many. Personalization is easy to achieve in the physical world. In the virtual world, it is a tad hard. There are no store executives to help, yet you must understand the preferences, needs and intentions of the shopper and help him transact. Sounds easy? I will say may be yes but mostly no.
Let's get to product recommendations, a very interesting form of personalization. The phrases that e-retailers use goes something like this:
- Similar products
- Recommended products
- People who viewed this product also viewed these products
- People who bought this product also bought these products and more
What is relevancy? To a shopper, it could mean related products around same price point, similar features and possibly a fast moving item.
 |
| Viewed product |
 |
| Recommended products |
This mobile phone costs Rs 2999 and the recommended mobiles are around the same price point. Since the shopper is looking for dual SIM phone, 2 dual SIM phones are recommended. All these 4 products are liked by the users as inferred from reviews and ratings.
Let us take another example where an attempted product recommendations goes bad.
 |
| Viewed product |
 |
| Recommended products |
Anomalies are easy to spot - Samsung laptop is similar to Sony PS2 and gaming laptop worth Rs 76,099 does not come close to Sony PS2 costing Rs 5690. And most interestingly, people who bought Sony PS2 also bought a geyser. That is very insightful!
Mostly, product recommendations are generated using rules or computer algorithms. If algorithm driven recommendations are not meaningful, consider manually linking related products or use paid service. It is a safer way to tread. Today's shoppers are smart and they can smell bad recommendations from miles away.
Product recommendations are very useful tools for personalization. If done correctly, it can significantly improve customer experience, conversion rate and repeat purchase.
There are many more decision to make before serving good recommendations. I will cover these in my future blog posts.
More blog posts on E-Commerce:
Social Validation and E-Commerce
Social Commerce: Part 3
Gamification
No comments:
Post a Comment