Slides of the course on big data by C. Levallois from EMLYON Business School.
For business students. Check the online video connected with these slides.
-> Definition of data integration / fragmentation in a multichannel marketing environment. Explanation of the business stakes of data integration.
Student Profile Sample - We help schools to connect the data they have, with ...
The business stakes of data integration
1. MK99 – Big Data
1
Big data & cross-platform analytics
MOOC lectures Pr. Clement Levallois
2. MK99 – Big Data
2
Integrating data in a multichannel environment
1.Data: you don’t get it on tap
2.A multichannel environment increases data fragmentation
3.The business stake of data integration – why should we care?
4.Practical aspects of data integration
3. MK99 – Big Data
3
Example: Customer data.
Source: UNICA Corporation, in Multichannel Marketing, by A. Arikan (2008).
1. Data: you don’t get it on tap
Take away: data is fragmented by nature.
You construct customer profiles by joining and assembling different sources of data into a meaningful synthesis.
4. MK99 – Big Data
4
2. Data gets even more fragmented in a multichannel environment
•Basics:
–Distribution, information and ad channels keep diversifying
•POS, print, TV, radio, outdoor posters, mobile apps, mobile sites, emails, SMS, APIs, social networks, search engines, e- commerce platforms, e-commerce websites, blogs, content channels, …
–Connections between these channels intensify and complexify
•Social TV is TV delivered with Internet services, user profiles created on one platform are imported on another, orders taken online can be picked up on a variety of POS, ads circulating through one channel replicate on other channels, …
–Underlying technologies evolve and fragment quickly, across channels
•Cookies, SaaS, APIs, retargeting, HTML, Android, etc.
–The expectations of customers on the quality of service elevate (realtime, seamless experience)
-> Business stake: how to manage the complexity of this multichannel environment to deliver value to the market?
5. MK99 – Big Data
5
Example: French bank Societe Generale, up to early 2000s
POS
Outdoor
Call center
Radio, TV, Print media
One-way communication, analog.
1. No digital data collected
Two-way communication, analog.
2. Little (but important) digital data collected
6. MK99 – Big Data
6
Mobile app
Twitter account
POS
Youtube channel
Google Plus
Online banking
LinkedIn profile
Facebook page
Instagram
Call center
Outdoor
Print media, including online version
TV (including online TV)
Example: French bank Societe Generale, in the 2010s
One-way communication, analog and digital.
1. Digital data collected in large volumes
Two-way communication, digital.
2. Digital data collected in large volumes
Two-way communication, analog.
4. Little (but important) digital data collected
Two-way communication, digital.
3. Digital data collected in large volumes
7. MK99 – Big Data
7
Example: Societe Generale, in 2020?
Check the presentation on APIs to understand the stakes of this shift.
1. Very large (extensive?) amounts of digital data collected.
8. MK99 – Big Data
8
The fragmentation of channels
before
today
Source: “Multichannel Marketing Ecosystems”, by Stahlberg & Maila 2014). Chapter 1.
Data integration now:
-Offers more opportunities to create value and differentiate from your competitors
-But is harder to manage
9. MK99 – Big Data
9
Multichannel = disintegration of data
•The fragmentation of channel means the fragmentation of your data.
•Data about your customers, products and campaigns is scattered in different places (channels). You don’t know precisely:
–Who your customers are, how they behave with you
–How your products are perceived, purchased and used
–What are the results of your campaigns
10. MK99 – Big Data
10
3. Business stakes of data integration
1.Providing a consistent customer experience across channels
–How to provide the right services on the right channels
–How to integrate the experience across channels (seamless, enriched experience)
2.Managing communication campaigns
–Which campaigns for which channel(s)?
–How to coordinate campaigns across channels?
–Which budgets should be spent, and how to spread them across channels?
–How to measure the results of each campaign and the global result?
3.Generating actionable insights for business
–How to increase brand awareness,
–How to segment the customer base,
–How to improve the retention rate,
–How to better measure default risk / fraud risk / …,
–How to develop new products / services / channels
11. MK99 – Big Data
11
The pre-condition to achieve this goals
We should be able to have a picture of all different channels and how they connect.
This is hard.
12. MK99 – Big Data
12
4. Practical aspects of data integration
13. MK99 – Big Data
13
Organizational culture
Business culture
Software Application
Data type
Infrastructure
Organizational culture
Business culture
Software Application
Data type
Infrastructure
Servers in Canada
Clicks
NoSQL DB
Web agency
Startup mentality, data driven, fast to react
Servers in France
Number, duration and subject of customer calls to call center
ERP
Call center: B2B services
Mature industry, not data driven
Example: Your company has a website generating clicks from your customers. Your customers can also use your call center. How do you integrate these two datasets about yours customers?
?
Call center
Web agency
14. MK99 – Big Data
14
The layers of data integration
1.Organizational culture
–Attitudes towards data must be compatible
–Organizations / execs don’t have the same sensitivity to the priority of data projects.
2.BU
–How will different datasets be made compatible / convertible? (clicks and calls?)
–What is the revenue sharing agreement on data, if any?
–What are the acceptable levels of investment to generate / curate / share data?
3.Software application
–If the two parties agree to share data, how to do the sharing work in practice? (see video on APIs)
4.Data type
–Are datasets of a textual, numerical type? Are they time series, if so what is the frequency? Can identifiers be reconciled?
5.Infrastructure
–The volume of data, the place it is stored, and the servers and connections available: do they permit the integration to take place?
15. MK99 – Big Data
15
Options to manage the integration
Option
Pros
Cons
Delegate to IT department
- Resources already there.
- Conformity to existing procedures is assured.
-Innovation levels remain low
-Ad hoc solutions
External provider – domain specialist
-Experimented on data integration for a single domain of expertise
-Cover all layers of data integration
-Knowledgeable of your business needs
-Coordination costs
External provider –
- Large scope of data integration, across domains of expertise
-Coordination costs
-Diseconomies of scale?
-Lack of / costly customization
16. MK99 – Big Data
16
The jargon of data integration
DMP
Data Management Platform
CRM Customer Relationship Management
DSP
Demand-Side Platform
SSP
Supply-Side Platform
ERP
Enterprise Resource Planning
ETL or API
ETL or API
ETL or API
ETL or API
: you should research the meaning of these terms.
17. MK99 – Big Data
17
Next steps
•Watch the video lecture on APIs
•Go through the reading list
18. MK99 – Big Data
18
This slide presentation is part of a course offered by EMLYON Business School (www.em-lyon.com)
Contact Clement Levallois (levallois [at] em-lyon.com) for more information.