2014年10月31日星期五

Amazon's recommendation secret



In recent classes, we are discussing the recommender systems. In my mind, the most successful company using the recommender system is Amazon. All Amazon's users often receive emails, which recommend the items users are likely prefer and even buy.


At root, the retail giant’s recommendation system is based on a number of simple elements: what a user has bought in the past, which items they have in their virtual shopping cart, items they’ve rated and liked, and what other customers have viewed and purchased. Amazon calls this homegrown math “item-to-item collaborative filtering,” and it’s used this algorithm to heavily customize the browsing experience for returning customers. A gadget enthusiast may find Amazon web pages heavy on device suggestions, while a new mother could see those same pages offering up baby products.


Judging by Amazon’s success, the recommendation system works. The company reported a 29% sales increase to $12.83 billion during its second fiscal quarter, up from $9.9 billion during the same time last year. A lot of that growth arguably has to do with the way Amazon has integrated recommendations into nearly every part of the purchasing process from product discovery to checkout. Go to Amazon.com and you’ll find multiple panes of product suggestions; navigate to a particular product page and you’ll see areas plugging items “Frequently Bought Together” or other items customers also bought. The company remains tight-lipped about how effective recommendations are. (“Our mission is to delight our customers by allowing them to serendipitously discover great products,” an Amazon spokesperson told Fortune. “We believe this happens every single day and that’s our biggest metric of success.”)

2014年10月17日星期五

Social Media Analytics Are Transforming Digital Marketing

It has already passed 7 lectures. As a business student, I'm struggling in the matrix, the models and the programming of the course. But I still keep thinking about what should I learn from these course except the technic analytics methods. Oh yes, how to take use the social media analytics in digital marketing. Today let's talk about how the social analytics transforming digital marketing.

Does social media analytics mean looking at one-off engagements - fans, follower, likes - and using this information to make major campaign decisions? Not exactly. The realms of both digital marketing and customer care have moved beyond the web to social media, and marketers have incredibly advanced technologies at their disposal. Social media reveals not just who is engaging with your brand and what they're engaging with, but why they're engaging. The result creates a real value by measuring whether a campaign worked based on data and ROI.

Most companies already pay attention to customer feedback via tweets, Facebook comments and so on. That's important, but next step is learning what motivates customer purchase. Progressive marketers already use tools to influence point-of-purchase strategy, brand positioning, product development and more.. For example, by analyzing chatter across social channels, one consumer packaged goods company discovered that one of its beverages was routinely perceived as a good workout. This led to product placement near the exit of health clubs.

Moreover, by monitoring the competitors and their audience, companies can react to what's going on in an effort to capture market share. For example, if the competition releases a new product and customers are dissatisfied, this is an opportunity to target those users. It's possible to know who has the highest share of voice, the happiest consumers or the biggest impact from promotions, PR or launches.

The metrics to measure social media customer satisfaction

As a business guy, my primary attention may not the technic sides, but the analytics part, which can help the business decision-making. Social media is a important way for companies to promote their business. I'm really interested in how they use social media and how to analyze it.

These days, I read some articles on social media marketing. Provide excellent customer service is one of the top priorities of a successful business. For example, Zappos, one of the first retailers that uses social media as a customer care channel and sees them continuing to try out different ways to delight their customers. “The lifetime value of customer is a moving target that can increase if we can create more positive emotional associations with our brand", the CEO of Zappos said.

I summarize the main metrics to measure the customer satisfaction from some articles and share with you here.

1. Inbound volume
How many incoming messages you're receiving via social media. This can demonstrate the need for additional resources and/or indicate more customers are choosing to use social as their preferred method.

2. Response volume
The total number of responses that the company has issued or the total number of enquiries dealt with.

3. Keyword brand mentions
The total number of mentions of the brand (including the misspellings) not specifically directed at a company's official social media account. This metrics is most useful when looked at alongside sentiment metrics to identify the nature of comments.

4. Nature of enquiries and mentions by topics
By grouping enquiries and comments into categories, you can spot trends in types of problems and identify common queries that may then be resolved faster in the future via searchable help pages or FAQs. This metric can also help to detect potential reputation management problems.

5. Percentage of enquires responded to
How well the company is responding to its customers via social media.

6. First response time
Response speed.

7. Average response time
How quickly, on average, the team or individual replies to a customer's messages over the course of a specific query.

8. Average handling time
How long, on average, does it take for issues to be resolved compared with other channels like phone or email.

9. Abandonment rate
The number or percentage of queries the customer abandon, without a specific resolution.

10. Deflection rate
The percentage of issues moved to another channel rather than being resolved directly within social.

11. Sentiment
This provides an overall rating of positive versus negative sentiment in relation to the brand. It can be measured on comments addressed directly to company's social accounts, as well as general brand mentions across social.

12. Influencer sentiment
The opinion of your more influential customers is likely to have a greater impact on others. So providing a better service to people deemed to be more influential.

13. Competitor sentiment
In order to provide context, it can be useful to measure your sentiment alongside that of your competitors.

14. Customer satisfaction Index (CSI) or Net Promoter Score
Ultimately, this may be your most important metric. How do your customers feel about you and your brand when you ask them?

2014年9月21日星期日

Let's begin Social Media Analytics

Well, it's the first blog for my social media analytics journey. I'm a Business Analytics student from CUHK Business School. Hope to make friends with all of you, IE buddies.😄

First, I must advertise my forum - www.ba-china.org - an online communication about big data, business analytics and data science. If you are interested in BA, please join in BA-China! Click the URL, then register! Share your resource and opinion about analytics in the Big Data area! I will also share my Social Media Analytics journey in BA-China.

Second, let me talk about why I select the course - Social Media Analytics, even though I don't have any programming background. Social Media opened the area of Web 2.0. Plenty of data is generated through social media every second. Why people are addicted in social media, how to analyze their behavior, how to predict the effect of social media marketing, and how to take use of social media analytics to promote our business are the questions that I'm interested in. In addition, I can hold the opportunity to learn Python, a popular programming language used in big data. No Zuo No Die why you try. No Try No High Give Me Five! (a very hot sentence on the internet in China).

Finally, share with you some useful study resource:
1. Coursera: Social Network Analysis by University of Michigan
2. Udacity: Programming foundations with Python
3. Python教程中文版