Week_11_Web Analytic Flashcards
What marketing clouds do?
Bring all kinds of data together and use them for campaign
How to personalize a person before send him a campaign?
Get data from interactions and events, then merge match and enrich the data to get this person’s interests, contexts and predictions. Use these analytic result we can know this person’s target and personal activity in the future and send them campaign.
What is marketing cloud useful for？
- Combine online and offline data.
- Personalize 1:1 customer journey.
- Leverage AI to analyze large data sets and provide insight.
- Easy to conduct large scale testing, optimization and personalization in real time.
What is digital fingerprints?
Extract data from
- browser.
- cookie
- tagging behavior on website.
- server log
- APP, GPS
What should marketer encourage visitor?
log in to the website (even delete cookie still can collect data)
What is the possible trend for mega brand?
driven by omnichannel/marketing clouds-away from web analytics packages to raw data from multiple sources then analyzed by AI
EU’s General Data Protection Regulation (GDPR)
Protect data collecting from customers in EU
What is identity resolution?
the process to match users across platforms, online and offline
What is drivers for identity resolution appear?
Marketing clouds, DMP’s and proliferation of devices, use several browsers and apps
How to determine the unique customer?
Email ID (not for last name and email address
Cookie can not track ads within App
True
Software ID and hardware ID, Which one is privacy supporting?
Software ID, can be disable and reset by consumers. Ads track rely on software ID
What are features of hardware ID?
Unique and persistent.
advertiser can see that conversion online in their Google Ad report-if Google Store Conversions- enabled
True.
Mobile GPS tracking is used by apps and also by ad platforms like Google, Bing, Facebook for online conversion
False. can be tracked offline.
Where marketer can buy aggregate segment data?
Ubermedia, it provides GPS tracking data and let marketer make contextual ads(个人化广告）
B2B marketer used to use IP tracking technology but B2C not use
True
Trinity approach of web analytics
Experience, behavior, outcome
Four purposes that web analytic can achieve
Business objective
Goal
Dimension & Metrics
KPI
Dimensions are attributes of visitors to your website, usually descriptive words
Metrics are always quantitative-numbers, percentages, ratios that measure one or more of the characteristics of a dimension
True
Metrics by themselves are not useful, need to be associated with a dimension
True
What are three categories that metric can be in?
Acquisition( new user, session)
Behavior (bounce rate, average session duration)
Conversion (goal conversion rate)
What is KPI?
Quantifiable and measurable, critical to the success of organization. limit 5-8 metrics, applied consistency.
Why management don’t want to see some reports?
No trends and insight at all. just rehashing of data.
How to avoid bad report?
Put extra data in an appendix if too much data
• Must have recommended actions-action verbs
• Mention ROI or cost savings if proposed actions taken
• Provide statistical validation
• Assumptions and context
• User segmentation-key
• Higher level metrics for context- Task Completion Rate (qualitative analysis) or Message Amplification (social media)
Three key factors from Avinash Kaushik
#1: Conversion Rate: Remember that the average website conversion rate usually is around 2% #2: Average Order Value: why? Conversion Rate could be through the roof and yet revenue could be down. #3: Days & Visits To "Purchase“: if multiple sessions lead to conversions you might not be attributing to the right campaign / acquisition drivers
Key to successful in web analytic
- Insight vs. Reporting mind set
- Understand definitions of each dimensions and metric: what they represent: a session, returning user, referrals, source/medium, etc
- Make sure you are using higher level features so as to measure profitability and find problems e.g., LTV, attribution and cohort analysis
- Learning how to combine data, export, segment, use regular expressions.
- Understand the limitations of free platforms
- Have your website tagging done.
- If you sell products/services make sure e-commerce is enabled in GA-so you can track at the product & category level
What is event?
Events are user interactions with content that can be tracked in depth. from a web page or a screen load
What is funnel?
A defined path that visitors should take to reach the final objective
What funnel analysis can do?
understanding where problems in the conversion process may reside
What is segment?
create groups of users to see if they behave or perform differently (better or worse) than the aggregate.For example you can use referral URL, landing page or connection speed to to set segment.
What is bounce rate?
how many visitors viewed only one page and left your website-usually not possible to accomplish a goal with one page- registrations, blogs and landing pages the exception
Best focus of segment？
Conversion/funnel stages
A low bounce rate implies a higher likelihood of conversion – this is true up to a point
True, normally < 40% is good.
PPC leads to low bounce rate
Wrong, usually leads to high bounce rate. 95% is bad.
What is attribution？
what credit will you give for sales, which stage you wanna focus on to improve sales
Single touch attribution
Optimizing one of the following: Last interaction (google search) Last non-direct click (social media) Last Adwords click (PPC) First interaction
Multi-touch
Linear (equal credit to each stage)
Tme decay
Position Based (weighted in different stages)
Data driven
Cross device attribution
start from phone (65%) to PC 4%
start from tablet (11%) to PC (10%)
Challenge of cross device attribution
due to the lack of a foolproof, accurate way to identify the same user across different devices
How to solve this challenge?
- Deterministic Matching, identifying the same user across different devices by connecting the same unique identifiers together. For example, let people log in on website.
- Probabilistic Matching, use anonymous unique identifiers to identify the same user across different devices, probabilistic matching, uses a range of different data sets and algorithms to make probable connections. For example, IP address, Device ID, location…
What use of website tagging?
collect and share data between your website and analytics platforms and ad networks
How it shows?
A tag is a pixel or beacon by which data is collected on a website. Can be a simple 1×1 transparent pixel or image loaded onto the web page or JavaScript code put on web pages
How to place tag?
Place on web page (header/footer) and delivered to a web browser or app when a web page loads- tag management makes easier
ags are not cookies and cookies are not tags.
True, A tag can be used to set a cookie. Cookies are text-only strings of code placed on a computer or device
Cookie
Cookies
Tracking mechanism-text file put in browser that stores information about user preferences such as location, to personalize & collect analytics data
Cookie not shared if different browsers
True
Basic type for cookie
1st party-matches domain user is on Session or Persistent (until you delete or expiration date) Security 3rd party (ad networks)-weakest Zombie (recreates after deletion)
Cookie challenges
- Problem with cookie deletions- ~30% deleted on monthly basis
- Europe has a policy were websites must disclose use of cookies when user lands on the page first time
- Some browsers block 3rd party cookies by default– latest versions of Firefox, Safari very restrictive
- Advertisers should allow Opt out of behavioral tracking-icon from AdChoices next to retargeting ads
How mobile cookie affected?
device type, if website or app, 1st vs. 3rd party cookies, browser, version and browser settings
Walled Garden
cookies from one app are not shared with another app
What factor of GA real time analysis you need to pay attention to
due to low volume, it might not be statistically significant
What is cohort analysis in Audience’s category?
grouping or subset of visitors based on a certain metric. Data broke down by cohort (new users spent every month example)
If don’t sell anything, can I use e-commerce function under conversion metrics?
Can still be used if website doesn’t sell anything, your conversion can be completing a form, registration etc.
