gary taylor - data analytics in an e-commerce environment

Post on 20-Jan-2017

311 Views

Category:

Marketing

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

DATA ANALYTICS IN AN

ECOMMERCE ENVIRONMENT

Gary Taylor, tmwi – April 2016

Gary Taylor

Digital Director, tmwi

Follow me on Twitter @garyptaylor

2004 – 2007: Built a network of mortgage and secured

loan lead generation websites

2008: Won Young Entrepreneur Of The Year Award

2012: Helped establish the digital business of tmwi

Today: Largest independent media and digital

agency outside of London and #35 in the UK*

*Campaign Live / Nielsen 2016

My Journey…”Inspired by disruptive tech and data”

• Lectured at Birmingham City University Future Media (MA) Course on

Marketing Analytics

• Studying Machine Learning and Data Science (Stanford University)

• Proficient in Python, SamIam (Sensitivity Analysis & Inference) and

Octave (Linear and Non-Linear problems)

• Over the last 18 months, have been developing a data analytics

platform to help clients understand and answer their data

challenges

What is data?

• It’s a word used recklessly across the industry

• Without data you can't create new knowledge

• THE vital element for setting KPI’s and objectives

• BUT, if you don't understand your data... You can't

grow, can't adapt, can't change...*

*Jono’s – “It’s frameworks, not data, which drive success”

“change is the only constant…”

Plato

“99% of companies haven't

changed their marketing strategy in

the last 10 years...”

Reliable Source: A “recent” survey

A story of data…

Gateway Communications & GTV• 2003 – 2008 grew to be the largest provider of pan-African mobile

networks on the continent

• Sold to Vodafone for $675m

AFB.comCredit cards

Jumo.world• Disruptive fintech

• Mobile lending and credit

facilities to the “un-banked”

population

• Network reach of 35m users

• Issuing 45,000 loans every day!

• Rewind 60 years… CAF was formed

• Three famous African EPL footballers – Didier

Drogba (C), Kanu (A), and Quinton Fortune (MU)…

• Until the 70’s football remained relatively un-

represented

• Joao Havelange (FIFA President) who opposed

Apartheid advocated football across the continent

• In 2010 South Africa hosted the World Cup

• Today there are 230m EPL fans in Africa

Can anyone guess what AFB did?

New Product Development!

Gateway Communications & GTV• 2003 – 2008 grew to be the largest provider of pan-African mobile

networks

• Sold to Vodafone for $675m

AFB.com• A leading provider of credit cards in Africa

• Servicing 600,000 credit card holders

• 450 retail store card partners

Jumo.world• Disruptive fintech

• Mobile lending and credit

facilities

• Network reach of 35m users

• Issuing 45,000 loans every day!

Knowledge Fusion.

“processing resulting in an appearance of new

knowledge…”

Jacques Ludik – CDO at Jumo

• Ph.D in Computer Science

• 20+ years’ experience in Advanced Analytics, AI, Machine Intelligence & Data Science and its

applications

• Founded the Machine Intelligence Institute of Africa

(MIIA)

• Senior Lecturer and Researcher in the Department of

Computer Science at Stellenbosch University

• Has a patent filed for Knowledge Fusion

Why is this important in

ecommerce?

Data Challenges

Challenges

Data Capture

Data Ownership

Multiple Data

Sources

In-house Resource

No Central Database

IT Limitations

Lack of Data

Strategy

Fragmented

Market

Cost

• Not all data is captured

• Data ownership

• Multiple data sources with multiple

interfaces

• No central database…mostly held in

multiple 3rd party software and excel!

• Limited in-house data support/ resource

• IT limitations compromise data integrity

• No data strategy… what is the end goal

for the data?

• Too many vendors on the market, difficult

to be technology agnostic

• Cost of technology

“If you can’t explain it simply, you don’t

understand it well enough.”

Albert Einstein

Collection Integrity Analysis Visualisation

Gathering

Measuring

Quality

• Design

• Implementation

• Storage / Security

• Processes

• Retrieval

• Inspecting

• Cleaning

• Transforming

• Modelling

• Presentation

• Representation

• Commercial toolbox

Our SIMPLE Process For Data Analytics

Multiple sources of data (API, XML*, CSV, raw data)

Database and DMP architecture and

proprietary technology

Visual dashboard (Client front-end)

Collection

Integrity and Analysis

Visualisation (Insight)

Platform Architecture

*API – Application Programming Interface, XML – Extensible Markup Language

Real Insight…

Client Case Study

• Identified and reduced “harvesting” gaps in organic search with new SEO strategy

• Worked closely with paid media and programmatic teams to inform online

consumption patterns of users with highest propensity to convert, using 1st and 3rd

party audience data

• Provided technical support to improve mobile experience, site speed and on-page

call to actions

• Improved UK organic conversion rate from 7% to 11% (highest in European region

for the brand)…a 56% increase in transactions within 6 months

Process Expansion

“Worked closely with paid media and programmatic teams to inform online

consumption patterns of users with highest propensity to convert, using 1st and 3rd party

audience data”

Mosaic Profiling Of

1st Party DataTGI Consumption

Propensity Match

3rd Party DataInform and Refine

xxxk records of

users who had

purchased online

• Media owners

• Content

preferences

• Consumption

habits

• Site lists

Identified further

top indexing

audiences, using

cookie data

• Future content

strategies

• Programmatic

buying tactics

Content Themes

Sites Ranked By Index % Index

Children's 12 152.1

Property 19.3 131.1

Technology 23.9 123.7

Gaming 20.9 123

Beauty 14.1 122

Music 31.1 121.8

Mobile Phone Provider 27.4 121.1

Job search 26 118.9

Music download 19.5 117.7

Fashion 19.6 117.5

Utility (e.g. water, fixed line phone, TV licence) 34.8 117.1

Film & Cinema 32.7 116.1

Football (not related to one club) 21.7 115.9

Auction 33.6 115.8

Website Category Preference

Actual Websites

Sites Ranked By Index % Index

Disney (+04/14) 9.3 163.7

TEAMtalk.com 2.6 152.3

Bounty 3.2 151.7

BikeRadar.com 2.1 150.9

NetMums 10 150

Kerrang 3.2 149.8

ESPN 3.4 149.1

Kayak .8 148.2

Liverpool (Liverpoolfc.com) 4.7 147.3

Net-a-porter/mr porter 2.3 144.9

X FM 1.8 144.9

Babycentre 5.8 144.4

Sites Ranked By Size % Index

Google 94.6 112

Facebook 84.4 113

YouTube 80 112.4

Amazon 78.2 113.7

Ebay 70.3 115.7

Any BBC online site 57.7 109

Google Maps 56.8 114

Windows Live 56 114.7

Yahoo 48.2 106.1

BBC News 48.1 108.1

Tesco 46.7 114.9

Twitter 46.3 111.8

Summary…

• Set clear goals, objectives and KPIs

• Do what Jono said!

• Follow our process to address your data challenges (Collection,

Integrity, Analysis and Visualisation)

• Integrated campaigns drive efficiencies, don’t work in silos

And Finally…

The bar is open… it’s a Friday…

What are we all still doing in this bloody conference?!

Thank you.

Follow me on Twitter @garyptaylor

top related