22
Harnessing the Power of Big Data to Run Data Driven eCommerce Company

Harnessing the Power of Big Data to Run Data Driven ...2016.icoict.org/wp-content/uploads/presentation/Tutorial-Blibli-ICoICT-2016.pdfSource: Blibli.com Internal Data 5x YoY Growth

Embed Size (px)

Citation preview

Harnessing the Power of Big Data to Run Data Driven eCommerce Company

Was founded in July 25th , 2011

BLIBLI.COM

Our Product Categories

Our Business Models

2,200,000++

Members

788,000++

likes

334,000++

followers

Our Members & Fans

1,500,000++

Subscribers

*) Data per Apr 2016. Source: Blibli.com Internal Data

29,000++

followers

Merchants & Products

8,800++ 640,000++ 9,300++

*) Data per Apr 2016. Source: Blibli.com Internal Data

InternetBanking

Debit / Credit Card

Transfer Method

Electronic Money

COD Payment

Our Payment Methods

ATMConvenience

Store

Wide Distribution Network

ALL OVER INDONESIA

ALL KINDS OF

PRODUCT*

NO MINIMUM PAYMENT*

*

*) Except products that need special treatment**) Except diapers & milk product, min. IDR 350k

FREE Shipping

Our Shipping

Traffic Growth

4.35xYoY Growth

(Q1 2016 vs Q1 2015)

Traffic Growth Mar 2016 vs May 2011272x

900.000++ Sessions / Day*

3.300.000++ Pageviews / Day*

650.000++ Users / Day*

*) Data per May 2016. Source: Blibli.com Google Analytics

2011 2012 2013 2014 2015 2016

2011 2012 2013 2014 2015

Sales Growth

*) Data per Apr 2016. Source: Blibli.com Internal Data

5xYoY Growth

(2015 vs 2014)

Sales Growth FY 2011 vs 2015240x

Data Ingestion

Store Front Events

• Transactions

• Cart, Wishlist

• Page views, sessions

• Banner impression & click

• Product search

• Social media sharing

Back Office Events

• Order validation

• Payment

• Fulfilment

• Product upload, approval

• Inventory movement

• Customer service interaction

Current ingestion rate of

10M+ records daily

System Architecture

Sample Use Cases

Recommendation

Search Analysis

Fraud Detection

Logistics Selection

Customer Segmentation

Merchant Analysis

Recommendation

Product Recommendations

• Customers who viewed X also viewed Y

• Customers who bought X also bought Y

• Customers who viewed X eventually bought Y

• Similar products

• Frequently bought together

Search Recommendations

• Top searched items

• Search box auto-complete

• Related searches

Newsletter Personalization

• Personalized content

• Personalized promotional offer

• Personalized products

Recommendation

• Gender

• Age

• Location

• Attributes

• Category

• Price

• Wishlist

• Cart

• Order

• Product view

• Promo view

• Banner clicks

Clickstream Transactions

User ProfileProduct

Search Analysis

Searched text

Search results

Search result clicks

40% of all searches return 0 results

WHY?

• Typo in search

• Product not available

• Search improvements

Fraud Detection System

Logistics Partner Selection

Logistics Performance

Capa-city

Price

{From, To, Weight}

Customer Segmentation

Merchant Analysis

• Out of stock

• Returned product

• Late delivery• SKU uploaded

• SKU published

• Sales

• Product review

THANK YOU