Digital marketing analytics
Lecture 1: introduction
What is digital marketing?
Marketing is the activity, set of institutions, and processes for creating,
communicating, delivering and exchanging offerings that have value for
customers, clients, partners and society at large.
Online marketing refers to the internet and e-mail based aspects of a
marketing campaign. Can incorporate banner ads, e-mail marketing,
search engine optimization, e-commerce and other tools. -> narrow: on
communicating part; what kind of channels, how do you advertise, how in
contact with consumers, how find consumers you
Digital marketing is the marketing of products or services using digital
channels to reach customers. The key objective is to promote brands
through various forms of digital media.
Digital marketing extends beyond internet marketing to include channels
that do not require the use of the internet. It includes mobile phones (both
SMS and MMS), social media marketing, display advertising, search engine
marketing and any other form of digital media.
Digital marketing is an extension to traditional marketing, now we can
create new kinds of products and services, like online banking, streaming
videos and music.
Also for the communication with the consumer, there are more channels
available now, like social media. Communication is more direct, consumer
can easily communicate with company. Better targeting specific groups of
customers.
New ways of delivering services, like online services.
Same with exchanging offerings, like new payment systems.
Definition of marketing has not changed due to digitalization, but
more/new opportunities have become available.
History of computers
Started with the basic wooden counting window, then more mechanical
calculators, machines that can calculate (computers) since founding of IBM
in 1911. Then computers that can play chess for example (1949). Big
frame computers since 60’s and e-mail since 1971 followed by founding of
Microsoft and apple (1976). 1991 the world wide web launched, followed
by google, social media, ipads/iphones.
,Current status and future: self driving cars, generative AI, blockchain,
social media, home devices connected to phone. Advanced systems that
are almost as intelligent as humans.
Why constantly faster developments:
Moore’s law states that the number of transistors in a dense integrated
circuit doubles approximately every two years.
Exponential growth in the capability of computer chips. Also applies to
many other (digital) electronics, such as quality-adjusted microprocessor
prices, memory capacity, sensors and even the number and size of pixels
in digital cameras. Digital possibilities grow exponentially.
Eg floppy disks (old USB) decreased in size, grew in storage over time.
Then CD then DVD then Blu-ray, which has increasingly more storage for
information. Prices lowered.
Similar with the exponential development of AI: more storage, lower
prices, more data. More complex networks.
First AI was 70 years ago, just after the first computer. Then deep blue
where computer won from a world champion chess player. Then Alexa.
Then last year ChatGPT for public. Intelligence grew exponentially.
Speed and reliability of the network improved and still is, and therefore we
can see and stream everything. High internet penetration over the world,
first in developed countries, later developing regions follow. Also related to
the internet speed that increases exponentially. More information (like
videos) is being transferred and generated. Much online activity =
enormous data.
Big data analytics: continuous stream of new data is being generated,
analyze to get useful insights. A lot of data of the consumer available for
companies, used to get insides on how to influence or reach customers.
Was campaign useful, changes on products?
Data is not a goal in itself, how do you use the numbers to get insights
Transform data like what do people search for, what do they discuss on
social media? To improve marketing, advertising.
,Market share of online retailing growing. Online retailing wins market share
from traditional retailing. Large differences between categories and
countries.
Online advertising market grows, now more then 2/3.
More and richer data leads to more research possibilities and forms.
Traditionally data collected through store scanners of via surveys,
disadvantage is impact of marketing decisions (like pricing, displays,
advertising) only investigated at aggregate level. Therefore amount of
usable observations is limited. Online individual customers, the
(marketing) contracts they had and their shopping baskets can be
investigated. More closely investigate which customers have been reached
and how this changed their behavior.
Customers can be tracked individually through their different stages
through cookies and other tracking techniques (eg clickstream data).
Relatively easy to do large scale A/B testing (eg customers see different
versions of website, get different emails or different displays). Empirically
investigate word-of-mouth through social media, mobile communication.
Lecture 2: online advertising
Many different forms of online advertising:
- Email advertising (spam, newsletters) is cheap and still popular. You
reach the people that are already customer.
- Display banner advertising
- Search engine advertising
- Retargeting
- Affiliate programs. For visibility
- Online video ads.
Compensation methods
Advertisers pay the content provider for the advertisement space. In
traditional (mass) media the costs of an advertisement or commercial
depends on a few factors:
- Type of advertisement (print, sound, video)
- Size and position of the advertisement (place in commercial block,
place in newspaper, duration)
- The audience size (reach)
- The target group (eg broad vs specific; desired group costs more,
group with higher income costs more, fewer is more expensive)
- The quantity (in bulk or one per time)
, For online advertising there are a few different payment schemes. Because
easier to track how many you reach, what they do with the ad.
- Cost per mile (CPM): amount to be paid per 1000 ad impressions.
- Cost per click (CPC): amount to be paid per click on ad. So only pay
when customers clicks, not on forehand.
- Cost per action/acquisition (CPA): amount to be paid (or a share of
revenue) when a click on the ad led to a certain action (eg filling in a
form, buying a certain product of subscribing to a service – also
affiliate).
Simple online path to purchase
When having data about the amount of impressions, clicks and
conversions (eg clicks that leads to the desired action) per advertisement,
you can easily compare different advertisements and (in principle)
calculate the ROI of different campaigns.
Eg: 100 000 impressions -> 1500 clicks (so click-through rate: 1,5%) -> 75
conversions (conversion rate: 5%) -> €25 gross profit per conversion (new-
old/old)
In practice it does not matter that much if the payment is on the basis of
CPM, CPC or CPA. These numbers are easily comparable if you know the
(expected) click-through rate and the (expected) conversion.
For calculating the ROI with this data it is important that the
advertisement is solely responsible for the conversion and that the impact
is direct. Eg someone who didn’t click but then did searched and
purchased the product. Or would the made sales had happened without
the ad.
Search engines
Organic search results; by the algorithm that fits best with the keyword.
Don’t have to pay to google when you click on them.
Sponsored search results; companies pay an amount per click to google to
be shown high when searched.
Why do companies do sponsored when they are also shown as organic
result, eg mediamarkt when searching iphone 17? Most attention is given
to the first few search results. You want people on your website so you
have to score high on google. This can be reached by being very relevant,
conducting ‘search engine optimalization’ and by search engine
advertising.
Lecture 1: introduction
What is digital marketing?
Marketing is the activity, set of institutions, and processes for creating,
communicating, delivering and exchanging offerings that have value for
customers, clients, partners and society at large.
Online marketing refers to the internet and e-mail based aspects of a
marketing campaign. Can incorporate banner ads, e-mail marketing,
search engine optimization, e-commerce and other tools. -> narrow: on
communicating part; what kind of channels, how do you advertise, how in
contact with consumers, how find consumers you
Digital marketing is the marketing of products or services using digital
channels to reach customers. The key objective is to promote brands
through various forms of digital media.
Digital marketing extends beyond internet marketing to include channels
that do not require the use of the internet. It includes mobile phones (both
SMS and MMS), social media marketing, display advertising, search engine
marketing and any other form of digital media.
Digital marketing is an extension to traditional marketing, now we can
create new kinds of products and services, like online banking, streaming
videos and music.
Also for the communication with the consumer, there are more channels
available now, like social media. Communication is more direct, consumer
can easily communicate with company. Better targeting specific groups of
customers.
New ways of delivering services, like online services.
Same with exchanging offerings, like new payment systems.
Definition of marketing has not changed due to digitalization, but
more/new opportunities have become available.
History of computers
Started with the basic wooden counting window, then more mechanical
calculators, machines that can calculate (computers) since founding of IBM
in 1911. Then computers that can play chess for example (1949). Big
frame computers since 60’s and e-mail since 1971 followed by founding of
Microsoft and apple (1976). 1991 the world wide web launched, followed
by google, social media, ipads/iphones.
,Current status and future: self driving cars, generative AI, blockchain,
social media, home devices connected to phone. Advanced systems that
are almost as intelligent as humans.
Why constantly faster developments:
Moore’s law states that the number of transistors in a dense integrated
circuit doubles approximately every two years.
Exponential growth in the capability of computer chips. Also applies to
many other (digital) electronics, such as quality-adjusted microprocessor
prices, memory capacity, sensors and even the number and size of pixels
in digital cameras. Digital possibilities grow exponentially.
Eg floppy disks (old USB) decreased in size, grew in storage over time.
Then CD then DVD then Blu-ray, which has increasingly more storage for
information. Prices lowered.
Similar with the exponential development of AI: more storage, lower
prices, more data. More complex networks.
First AI was 70 years ago, just after the first computer. Then deep blue
where computer won from a world champion chess player. Then Alexa.
Then last year ChatGPT for public. Intelligence grew exponentially.
Speed and reliability of the network improved and still is, and therefore we
can see and stream everything. High internet penetration over the world,
first in developed countries, later developing regions follow. Also related to
the internet speed that increases exponentially. More information (like
videos) is being transferred and generated. Much online activity =
enormous data.
Big data analytics: continuous stream of new data is being generated,
analyze to get useful insights. A lot of data of the consumer available for
companies, used to get insides on how to influence or reach customers.
Was campaign useful, changes on products?
Data is not a goal in itself, how do you use the numbers to get insights
Transform data like what do people search for, what do they discuss on
social media? To improve marketing, advertising.
,Market share of online retailing growing. Online retailing wins market share
from traditional retailing. Large differences between categories and
countries.
Online advertising market grows, now more then 2/3.
More and richer data leads to more research possibilities and forms.
Traditionally data collected through store scanners of via surveys,
disadvantage is impact of marketing decisions (like pricing, displays,
advertising) only investigated at aggregate level. Therefore amount of
usable observations is limited. Online individual customers, the
(marketing) contracts they had and their shopping baskets can be
investigated. More closely investigate which customers have been reached
and how this changed their behavior.
Customers can be tracked individually through their different stages
through cookies and other tracking techniques (eg clickstream data).
Relatively easy to do large scale A/B testing (eg customers see different
versions of website, get different emails or different displays). Empirically
investigate word-of-mouth through social media, mobile communication.
Lecture 2: online advertising
Many different forms of online advertising:
- Email advertising (spam, newsletters) is cheap and still popular. You
reach the people that are already customer.
- Display banner advertising
- Search engine advertising
- Retargeting
- Affiliate programs. For visibility
- Online video ads.
Compensation methods
Advertisers pay the content provider for the advertisement space. In
traditional (mass) media the costs of an advertisement or commercial
depends on a few factors:
- Type of advertisement (print, sound, video)
- Size and position of the advertisement (place in commercial block,
place in newspaper, duration)
- The audience size (reach)
- The target group (eg broad vs specific; desired group costs more,
group with higher income costs more, fewer is more expensive)
- The quantity (in bulk or one per time)
, For online advertising there are a few different payment schemes. Because
easier to track how many you reach, what they do with the ad.
- Cost per mile (CPM): amount to be paid per 1000 ad impressions.
- Cost per click (CPC): amount to be paid per click on ad. So only pay
when customers clicks, not on forehand.
- Cost per action/acquisition (CPA): amount to be paid (or a share of
revenue) when a click on the ad led to a certain action (eg filling in a
form, buying a certain product of subscribing to a service – also
affiliate).
Simple online path to purchase
When having data about the amount of impressions, clicks and
conversions (eg clicks that leads to the desired action) per advertisement,
you can easily compare different advertisements and (in principle)
calculate the ROI of different campaigns.
Eg: 100 000 impressions -> 1500 clicks (so click-through rate: 1,5%) -> 75
conversions (conversion rate: 5%) -> €25 gross profit per conversion (new-
old/old)
In practice it does not matter that much if the payment is on the basis of
CPM, CPC or CPA. These numbers are easily comparable if you know the
(expected) click-through rate and the (expected) conversion.
For calculating the ROI with this data it is important that the
advertisement is solely responsible for the conversion and that the impact
is direct. Eg someone who didn’t click but then did searched and
purchased the product. Or would the made sales had happened without
the ad.
Search engines
Organic search results; by the algorithm that fits best with the keyword.
Don’t have to pay to google when you click on them.
Sponsored search results; companies pay an amount per click to google to
be shown high when searched.
Why do companies do sponsored when they are also shown as organic
result, eg mediamarkt when searching iphone 17? Most attention is given
to the first few search results. You want people on your website so you
have to score high on google. This can be reached by being very relevant,
conducting ‘search engine optimalization’ and by search engine
advertising.
