Upgrade to remove ads
CHP 11: Digital Analytics
Terms in this set (84)
tremendous amount of data available to us
its like trying to take a drink from a fire hydrant
The explosion in digital marketing and analytics is partly fueled by the
the data is
objective is -
As with most analytical techniques, we first must consider what our _____ and that applies to digital especially
-this helps us narrow what data and metrics we want to track and use
Then we design our process
either allocated or treated as a "fixed" expense.
Marketing effectiveness was somewhat hindered in the past and a lot of marketing expenditures were
example, if i was running 3 different ads, in the past we didnt even know what was actually causing buyers to spend
We literally can track effectiveness using analytics down to a "click", "forward", "share", "view", etc.
With todays data-rich digital environment marketing expenses are more and more categorized as a
Because of this improvement in analytics, marketing is becoming more and more .
1 Website Effectiveness
2 traffic and Audience Data
-Websites, Mobile, Social Media
3 Effectiveness of Marketing Campaigns
-Website and Mobile
4 Metrics Unique to
There Are Various Measures of Effectiveness for Digital Analytics
we pick analytics based on campaign objective
You need to consider customers perspecitve with usability.
You also need to consider enterprise's perspective, we want to know traffic, audience and campaign measurement
-server request logs
-tagged web pages
-server error logs
Example, evaluating website effectiveness
Tagged Web Pages Track
A Few Lines of Code on Each Site Page to be Tracked
Requires Cookie to be Set on User's Browser
Good for Only 1 Visit
Good for a Specified Period of Time
-these can be set for months and the people can literally track what you are doing with your browsing on ANY site
cookies Set by a Marketing Services Provider
cookies Set by the Website
Activity data produced by tagged web pages (see chart)
cookies on user computer, tags on web pages, metrics platform all work to help see what is going on with customer
-the metrics platform is needed to analyze
Cookies >Tags >Metrics Work Together
Traffic Data Measure
Each File Requested on Given Page
-Of Little Use to Marketer; Term Often Used Incorrectly
Number of Times Ad Is Requested by a Browser
-Does Not Say That Ad Was Fully Loaded or Seen
Page Delivered to Visitor
-Not Same as Page Being Seen by Visitor
Amount of Time User Spends on Site
Number of Times a Link Is Clicked
Audience Data about People Who Visit Site
Number of People Who Visit During Specified Time Frame
Total = All Visits Including Multiple By Same Visitor
Unique = Number of Different People Who Visit
Unidentified vs. Identified
-Number of Page Views
-Path Through Site
-Number of Shopping Carts Abandoned
-Entry Page to Site
Metrics by Stage in Customer Journey
WE MUST KNOW OUR OBJECTIVES BEFORE WE DECIDE METRICS
number of impressions
reach and frequency
first website vist
subscribers to newsletters, blogs, and other content
repeat website visits
inbound calls or emails
social media comments and questions
ROI per channel
number of repeat purchases
The Only Answer Comes From
the Stated Objectives of the Marketing Activities
Which Metrics to Choose?
Valid Metrics Framework for Social media marketing
Categories in Valid Metrics Framework
Opportunity to See (same as in Advertising)
Amount of Time Spent
Ability to Affect Attitudes and Behavior
Likelihood That One Person Will Recommend Brand to Another
Engagement = Likes + Shares + Comments
Engagement = Comments (only)
Metrics depend on the platform
For facebook engagement is
for instagram, engagement is
Read Metrics Definitions Carefully!
-Especially Good to Reach Young, Affluent Viewers
-Ones That Are Hardest to Reach on Television
-Branded Content Can Inform While it Entertains
-Video Content Is Inherently Engaging
Video Content Booming
Video Metrics Based on Time Spent
number of views
click through rates
percent of video viewed
time spent on page
Designing Digital Experiments
I've mentioned a number of times the challenges that we marketers face in today's business environment
We are required to report the effectiveness of our marketing campaigns and respond to measures such as ROI, etc.
are an excellent way to determine the effectiveness of what we are doing
In your book Chapter 15 the question from a CFO to a brand manager is "You have requested $500,000 for a national campaign. Is that the right amount? Can you get the same result for $250,000?"
1 Launching a marketing campaign clearly increases sales
2 Not launching a campaign causes no change to sales
3 Launching a marketing campaign today affects sales in subsequent time periods
4 There were no other external factors (competitive action, government action, 3rd parties, etc.) affecting sales
With the previous question in mind, there are four key rules to determine a causal relationship between two variables or factors
To clearly determine whether the targeted sales increase could be achieved without spending marketing dollars, or whether any marketing spend is warranted, we have to establish .
all other factors and observe changes such as "sales" or "brand awareness"
An experiment provides a mechanism to manipulate one or more input factors while controlling _____ and observe _____
is an experiment where you measure sales among participants of the experiment only after they are exposed to your marketing campaign
A very basic experiment that can be used is the after-only experiment
In the experiment example in the book, the brand manager recruits 1,000 customers...
½ are exposed to a new advertisement
½ are exposed to an old advertisement
Cereal purchases by the 1,000 customers is tracked
The difference in sales between the two groups will indicate the magnitude of the potential sales lift provided by the new advertising campaign
digital experiments are extremely easy cheap and faster, there is not reason for digital marketers to NOT use them
Sales increase in the short term
Sales increase in subsequent periods
SEE after only experiment diagram
This after-only design satisfies two of the four conditions for causality:
the new advertisement
the after only design experiment cannot indicate whether the increase might have occurred without
critical to remove a bias
Deciding how to distribute customers between the "test" group and the "control" group is
There are two primary ways to select control groups
allocating participants randomly between the treatment group and the control group. Works well with a large enough sample size to provide fairly homogenous groups
Attribute matching -
selecting sample groups based on similar demography, geography, income, etc. Works well in offline for TV advertising in different markets. You want those two markets to be similar in attributes - "matched".
This experiment design requires you to measure the output of interest both before and after participants have been exposed to the inputs
expose both groups to the same old advertisement first to see if there is any difference between the groups you have divided
new advertisement, and 500 to the old advertisement
Using the previous experiment example, with the before after design you would randomly divide the 1,000 participants into test and control groups just like the after-only experiment
The difference here is you would
Then expose 500 to the
before after experiment is very effective in controlling for all external differences between test and control groups.
100 units without changing anything
In this "Before-After" example you can see that when both groups are exposed to the old ad before the new ad is implemented, the test group had a sales lift of
new base for comparison to the new ad
this means that Something else caused that sales lift and those 100 units are now used to set the
Instead of starting with 1,000 units as a "base" measure we now start with
200 units to 100 units
Now the lift in sales from the new ad drops from ___ to ____ units
In this example, if the product sold to retailers for $1.59 and the cost of goods is 99 cents then the unit contribution is $0.60.
If the sales lift in the prior example was truly $60 ($0.60 * 100) and the cost of an email sent to the customers is $0.10 per email to 500 customers then the total cost is $50.
$60 in sales lift... $50 cost in email campaign... ROI of 20%.
marketers and make a difference in our organization.
Metrics... analytics... effectiveness... efficiency... this is what we do in Marketing to be better
As demonstrated in the previous example, digital (online / web) experiments have gained a significant edge over traditional offline field experiments
-purchasing of spots in different channels in the test markets requiring significant lead time
-Once the video is shot, it can be difficult, time-consuming and very expensive to change it
-The cost of the experiment increases with each new version of the video or commercials you would like to test
Consider the difference between TV and e-mail advertising campaigns
easier and less expensive for testing different versions.
Email advertisement can be created much more quickly and at much lower cost making it
easily test the simultaneous influence of multiple inputs
The faster execution, lower cost and quicker results of digital experiments allows marketers to
tv advertising (offline) could take weeks to test
email advertising (digital experiment) could take a day for results
The explosion in digital marketing and analytics is fueled by the tremendous amount of data available to us
-Easily captured, tremendous detail, timely, relevant, quicker results, cheaper
-Digital would not be a good platform IF your target market is not online!
The more relevant and timely a marketing campaign is the more effective it is!
-Consumers don't mind ads, etc. that they expect or enjoy
-To make our marketing more relevant and timely we can track so much detail about our customers with cookies (session, persistent, 3rd party, 1st party) and technical markers like IP address, first time visitor, returning visitor, entry page, exit page, time on page, forwarding of emails, video viewing time - it's endless.
-We have discussed Web metrics, email metrics, social media metrics and video metrics. All are heavily used in marketing analytics.
Digital experiments are used heavily today. We must design experiments that answer the "cause / effect" relationship between our campaigns and the consumer response.
THIS SET IS OFTEN IN FOLDERS WITH...
EXAM 2 Analytics
CHP 13: Website Analytics
Big Data 101
CHP 1: MARKETING ANALYTICS
OTHER SETS BY THIS CREATOR
EXAM 3 Study Guide
EXAM 2 REVIEW
CHP 13: PR and sponsorships
CHP 14: Display Ads
OTHER QUIZLET SETS
MGT 340 Exam 3 Vocab Summer Session
sophia us history I unit III practice milestone
Team Trust Harvard
*Patient Care Ch 17 test review