The challenge of web analytics

Do not put content

The Challenge of Digital Analytics & Intelligence

Giovanni Lorenzoni, CEO Seville, 22/09/2015

2 BitBang S.r.l. © 2015 – All rights reservedDo not put content

BitBang – Overview

§ Digital Analytics, Digital Intelligence, Customer Intelligence, Advanced Analytics, Optimization & Customer Experience Management consultancy company

§ Business Analysts, Data Engineers, Data Scientists

§ More than 10 years of experience in visitor acquisition, conversion and retention

§ Multiple industries

§ Marketing, Business, IT


Turn Data Into Business Insights

More Deeper For Everyone


Digital Data constitute the context in which event-level aggregation, through the application of a data model, pinpoints the added value that can

be drawn from visitors’ touchpoints

Digital Data is a Paradigm of Big Data

§ Digital Data are characterizedby:§ Millions of activities per day§ Many variables for each activity

§ Flow of information not always clearly aggregated

§ High cardinalityof the variables (page name, product sets, referrers, campaigns, etc.)

§ A revealing use of this information depends on the understanding of visitors’ behavior, which can not be inferredfrom a single record representing an outlying activity on the site (i.e. a page view)


What are the most common use cases in Big Data?

1. Solutions that focus on efficiency and risk

2. IT analytics solutions that focus on security and application performance

3. BT solutions that help understand and serve customers

6 © 2015 BitBang Srl – Riproduzione vietataDo not put content

Actionable Marketing Analytics

Helps marketers tell the story of what’s happening in their business.

With data.

And promote data-driven actions to impact business results.


What do Marketing Analytics Do for the Business

Increase Conversions

Optimize Campaigns

Increase Revenue / Bookings / Orders

Increase Traffic

Increase Average Order Value

Increase Engagement

Improve Sentiment

Reduce Customer Churn

Reduce Time To Market

Analytics uses data to guide how digital marketers…


Big Data Has Potential Positive Impacts Across The Customer Life Cycle

Source: Forrester Research



Types of Data to Deal With

DataWeb analytics

CRMBig

SocialPerformanceDigital Attribution

DMP

Multichannel

Behavioural MachineWarehouse

Customer

Voice of Customer

Target

Profiling

Back-office


Digital Data Inputs

Business Data Inputs Data

Processing Data Hub

Reporting &Analysis Action

Digital Intelligence Architecture


Challenge 1The Unknown Destinations


Challenge 2Data, Data Everywhere



The “Always Addressable Customer”

Source: Forrester Research, “The Always Addressable Customer”


The Lack of A Holistic Approach

§ Data siloed

§ Difficult joins

§ Disconnections beetween online behaviour and marketing campaigns

§ No Single Customer View

§ Inability to activate “intelligent” campaigns

Single Customer View

CustomerDB

Email Marketing

Social

Catalog Digital Behavior

Orders


Challenge 3Technology Abundance


CMO-centric

conversion / optimizationcross platform

& campaigntracking

customerexperiencetracking

dmp

testing / personalization

seo / sem

segmentation

attribution

tag management

customerinsights

CIO-centric

Legacy BI and Data Warehouse

A handful of Legacy Tools

Hundreds of Vendors

Marketing Analytics Marketplace

Source: VB Insight, August 2015



Use of Data-Related Tools and TechnologyWeb analytics

Campaign automation

Statistical package

Session monitoring

Data management platform (DMP)

NoSQL database

Demand-side platform (DSP)

A/B or multivariate testing

Relational (SQL) database

Business intelligence

Tag management

Attribution

Mobile analytics

Social analytics

Excel

Currently using Planning to use No plans to use

10%

12%

13%

30%

31%

33%

33%

43%

49%

49%

51%

52%

65%

90%

96%

21%

16%

31%

35%

24%

40%

48%

31%

33%

18%

34%

31%

27%

3%

3%

69%

71%

56%

35%

45%

27%

19%

26%

19%

33%

15%

17%

9%

7%

Source: Econsultancy, November 2014


“Technology is only one part of the equation, the otherpiece is human”

Brian Solis, Digital Analyst


Challenge 4The Shortage of Talent


The United States alone faces a shortage of 140,000 to 190,000 people with analytical expertise and 1.5 millionmanagers and analysts with the skills to understand and make decisions based on the analysis of big data.

Source: McKinsey & Company


In what areas are the biggest skills gaps?

Producing reports

Using digital analytics tools

Statistical modelling

Conversion rate optimisation

Using business intelligence tools

JavaScript programming

Stakeholder management

Querying database

Quality assurance (QA)

Project management

Requirements gathering

First choice Second choice Third choice

Source: Econsultancy, November 2014

3%

5%

7%

5%

6%

6%

7%

12%

13%

17%

20%

6%

7%

5%

6%

8%

5%

8%

13%

12%

10%

16%

8%

7%

6%

7%

5%

8%

5%

11%

12%

13%

11% 47%

40%

18%

18%

18%

17%

36%

21%

19%

19%

37%


The Black Hole Between Data and Knowledge


Challenge 5Domain of Expertise


WebAnalytics

DigitalAnalytics

MarketingAnalytics

*CustomerIntelligence

Marketing Customer Data Evolution

Source: VB Insight, August 2015


What Skill Sets Do I Need for Analytics?

Data Management

Business Analysis

Analytics Modeling

Understand Data§ Integrate and Manipulate§ Structured, Content and Hybrid Data§ Quality Assurance§ Big Data

Know Analytics

§ Appropriate Techniques§ Interpret Data and DiagnoseModels§ Meet Business Requirements

Focus on the Business§ Goals§ Constraints§ Decisions§ ChangeManagement§ Communication of Results

Source: Gartner, 2013


Challenge 6Governance


Organizations Need Data Leadership – The CDO Emergence

Source: Forrester’s Global Business Technographics® Data And Analytics Survey, 2015


The Need for Legal and Regulatory Compliance


Some Use Cases


Web AnalyticsPublic Web Site

Enterprise Dashboard

Web Site Performance Monitoring

BehaviouralTargeting

Usability Analytics

Social Intelligence

Customer Experience

Management

Mobile Analytics

MultiChannelAnalytics

Web AnalyticsHome Banking

Attribution Analysis

A/BTesting

Tag Management


Frequent Pattern Mining

Objective

Hypothesis

Methodology

Find behavioral patterns in the buying experience to determine effective up-sell and cross-sell promotions.

Classification algorythms continously process data to define and select association rules between devices and products identifying the proper customized experience for each user.

The propensity to buy is correlated with technology specifics and features.

Pattern 1: abito + phonePattern 2: pantaloni + pc

Grado di associazione

Pattern 3: pantaloni + scarpe + tablet


Use Case - Touch Point Analysis

>5 Touch points

3 Touch points

4 Touch points

5 Touch points

61%17%

5%

3%

2%

12%

1 Touch point

2 Touch points

61%

24%

69%

72% 71%71%

MAY JUN JUL MAY JUN JUL MAY JUN JUL

44% 47%

13%

66% 64% 63%

>5 Touch points

3 Touch points

4 Touch points

5 Touch points


Use Case - Touch FlowAssist AssistAssist AssistAssist Closer

Paid Search Brand

Organic

Paid Search No Brand

Direct

Paid Email

Remarketing

Email

Paid Social Ad

Display

Affiliate

U.V.

Sono esclusi gli u.v. che hanno utilizzato 1 solo canale

Non Paid

Paid


Use Case - Touch Points Alluvial Diagram

Periodo temporale di riferimento 05/2013 - 07/2013


Portfolio Performance: Attribution Analysis

Once the business objective shifts toward the consideration phase, our next step is to identify the best media mix, which requires the application of a customized attribution model.The custom model highlights the inadequacy of the last-touch one and the key role of the lastly-reconsidered DEM channel to achieve the objective.

Organic

Organic

Paid

DEM DEM+33% ROI


x

x

Community Performance

Do not put content

Thank you

[email protected]

Documents

The challenge of web analytics