Web Metrics: A Primer for UX Pros - UPDATED

UX + WEB METRICS
OXFORD TECHNOLOGY
VENTURES
How to Benchmark, Measure
& Evaluate UX Impact
!
!
General Assembly
August 03, 2014
Bev May
Beverly@OxfordTech.us @OxfordTech OxfordTech.us
@UXAwards UXAwards.org

QUESTIONS
OXFORD TECHNOLOGY
VENTURES
1. Have own Project / Sector to Research?
2. Web-Based?
3. Bring Metrics in a Spreadsheet?
4. Have a Computer?
5. Level of Familiarity with Web Metrics?
6. Any Data Scientists / Statisticians / Metrics Pros?
7. # Years Experience in Tech / Digital?

OxfordTech.us | GA: Web Metrics l August 2014
http://rlv.zcache.com/im_right_youre_wrong_next_coffee_mug-rbaffc7f99bad4d23bca7de839ac40bff_x7k28_8byvr_512.jpg

March to Nowhere
NO $$
NO TIME
WE KNOW OUR INDUSTRY
WE KNOW WHAT OUR AUDIENCE WANTS
AS AN EXPERT- YOU SHOULD KNOW WHAT’S BEST
STIFLE CREATIVITY / FUEL MEDIOCRITY

Benefit 1: THE UNKNOWN UNKNOWNS.
Donald Rumsefeld, wikipedia.com

Munch, the Scream- Wikipedia.org/
+ Improve Outcomes

LEAN: Helps Validate
Before Dev & Launch
(not just after!)
http://static.guim.co.uk/sys-images/Guardian/Pix/pictures/2011/8/22/1314031516692/A-boy-jumps-off-a-diving--007.jpg

WOW!
http://static.guim.co.uk/sys-images/Guardian/Pix/pictures/2011/8/22/1314031516692/A-boy-jumps-off-a-diving--007.jpg

Today
PART 1: OVERVIEW & CASE STUDY 1.5 HOURS
▪ CONTEXT & METRICS OVERVIEW
▪ PROCESS & IN-DEPTH CASE STUDY EXAMPLE (11 STEPS)
▪ TOP 10 NEWBIE MISTAKES
▪ SUMMARY & Q+A
!
BREAK
!
PART 2: WORKSHOP!
▪ DATA VIZ CHOICES
▪ DEFINE YOUR KPIS
▪ RESEARCH & GRAPH DATA OF COMPETITORS
POSSIBLE 2nd BREAK
!
▪ ANALYZE RESULTS & SHARE
▪ IF TIME - KPI-FOCUSED COMPETITIVE ANALYSIS
▪ RESOURCES

CONTEXT
OXFORD TECHNOLOGY
VENTURES

What is UX?
http://etc.usf.edu/clipart/70500/70542/70542_264_ra-090_o_lg.gif

User Centered Design
AUDIENCE
!
CUSTOMER PROBLEM
!
RESEARCH
!
PROTOTYPE
!
VALIDATE & EVALUATE
!
ITERATE

Holistic Design Thinking
Assumptions
true? How to
improve?
Who is your
customer?
What’s
their problem?
What’s
your
solution?
How is it unique/
better than
current?

Lean Startup = Good UCD / UX
MEASURE
LEARN
IDEA/ BUILD

Validation & Benchmarks: Scary but Critical

▪ Data-driven UX / Product / Strategy agency
▪ Founded 2008, NYC & SF
▪ My background: Product, UX. MBA + MS Tech/UX
▪ OxfordTech.us
▪ Effective
▪ Useful
▪ Engaging
▪ Internal & External Benchmarking
!

Clients

Product Development

UX Awards
Premier awards for exceptional digital experience, now in its 4th year
UXAWARDS.ORG

UXies: Data Driven Judging
UXAWARDS.ORG

Be a Competent UX Generalist
http://lawyerkm.com/wp-content/uploads/2008/09/swiss_army_knife1.jpg
Metrics
UX

What are Metrics?
#%<>

UX Research Methods
• HALLWAY USABILITY
• OBSERVATION
• INTERACTIVE TESTING (EYE TRACKING, ETC.)
• HEAT MAPS
• CARD SORTS
• SURVEYS
• HEURISTIC EVALUATIONS
• MARKET RESEARCH
• PUBLIC METRICS RESEARCH
• METRICS ANALYSIS
• MVT & A/B
OxfordTech.us | Web Metrics l July 2014
http://www.dsr.wa.gov.au/assets/images/Diagrams/Darts-playing-area.gif!

Metrics Data Sources
▪ Web- public & competitor
▪ Web- internal / private
▪ Social
▪ Mobile Apps
▪ User Testing & Analysis
▪ MVT & A/B
▪ Surveys
▪ Ecommerce

Metrics Categories
TRAFFIC
!
ENGAGEMENT
!
AUDIENCE
!
PLATFORM
!
(REVENUE)

Key Traffic Web Metrics
MARKETING FOCUS
▪ Uniques
▪ Visits
▪ % from Search -Paid vs. Organic,
Top Referring Terms
▪ % from Social
▪ % direct-load
▪ Top Referring Domains

Key Engagement Web Metrics
UX FOCUS
▪ Visits/ Unique
▪ Page Views (PV)
▪ PVs/Visit, PVs/ Unique
▪ Time Spent
▪ Bounce Rate (1 page/ visit)
▪ Top Entry / Exit Pages
▪ Top Sub-Sites/ Sections
http://blog.hugeaim.com/static/wp-content/uploads/2011/07/ballbounce.jpg

Key Audience Web Metrics
▪ Demographics- Age, Income,
Gender, Education, Location,
Ethnicity, Marital Status, Kids
http://clipartist.info/openclipart.org/SVG/rejon/person_outline_4-800px.png

Key Platform Web Metrics
DESIGN FOCUS
▪ % Mobile
▪ Display size & resolution
- desktop & mobile
▪ OS, Device, Web Speed
http://www.gizmoville.com/wp-content/uploads/2012/02/omgitsfullofpixels.png

Summary - Public Web Metrics
TRAFFIC- MARKETING
▪ Uniques
▪ Visits
Top Referring Terms
▪ % from Social
▪ % direct-load
!
AUDIENCE - EVERYONE
ENGAGEMENT- UX
▪ Visits/ Unique
▪ Page Views (PV)
▪ Time Spent
!
PLATFORM- DESIGN
▪ % Mobile
▪ Display size & resolution - desktop
& mobile
!
!

Questions?
TRAFFIC- MARKETING
▪ Uniques
▪ Visits
Top Referring Terms
▪ % from Social
▪ % direct-load
!
AUDIENCE - EVERYONE
ENGAGEMENT- UX
▪ Visits/ Unique
▪ Page Views (PV)
▪ Time Spent
!
PLATFORM- DESIGN
▪ % Mobile
▪ Display size & resolution - desktop
& mobile
!
!

How Public Web Metrics Work
!
Statistical Samples
▪ JavaScript
▪ Cookies
▪ Pixels
▪ Server-side tracking
▪ Web Traffic
Public = Inaccurate
▪ Won’t be listed on public metrics sites if too small/ new
▪ Heed the warnings
http://www.wsgsystems.com/uploads/images/cookies_large.jpg

Internal Metrics
!
▪ Usually more accurate
▪ Requires at least some development
▪ Little competitor visibility (unless high-cost)
▪ Click path Analysis
▪ Heat map Analysis
▪ % Logged In/ Out
▪ Ecommerce: ARPU/RPC, R/T, R/V
▪ Data by Sections/ Categories
▪ Top & Bottom Performing Pages, Sections
▪ Top Entry/ Exit Pages, Sections- More Detailed
▪ Top On-Site Search Terms, 404 pages
▪ % Mobile by Page/ Section/ category
▪ Demographics- Politics, Interests, Credit, Job,
Title
http: //www.damenationblog.com/wp-content/uploads/2012/06/iStock_000019717637Smal l . jpg

Client Case Study, Jan. 2014
TV/ DIGITAL MEDIA BRAND ANALYSIS- OUTPUTS
A. Competitive Metrics Analysis (PPT, 80 slides)
+ Excel Spreadsheet - 24 Competitors, 5 Categories
B. UX Competitive Heuristic Evaluation (PPT, 134 slides)
C. Quantified Summary of Competitor Evaluation
D. Internal Site Metrics Analysis (PPT, 82 slides)
E. Internal UX Site Evaluation (PPT, 114 slides)
F. New UX Concepting

11 Steps
1. Identify key sites, goals & KPIs
2. Get competitor metrics data from multiple sources
3. Graph data after standardizing in Excel
4. Check for oddities
5. Don’t be a Robot. Review, Analyze & THINK
6. Examine High Performers (Heuristic, Quantified)
7. Analyze Internal Site Metrics
8. Review Internal Site (Heuristic)
9. Generate Actionable Findings
10. Concept New Directions
11. Test & Iterate

Data Sources Used: MANUAL Process
INTERNAL
1. Comscore
2. Nielsen
3. Adobe Site Catalyst
!
EXTERNAL
1. Quantcast
2. Alexa
3. Compete
4. SimilarWeb
!
SOCIAL
1. Facebook
2. Twitter
3. YouTube
4. Klout
!
MOBILE
1. iTunes
2. Google Play

BUSINESS KPI (RED) & UX (BLACK) Goals
< VIDEO
DISCOVERY
< SOCIAL
VISITS
SEE < FULL
VIDEOS
< BRAND
AWARENESS

Define Your Metrics KPIs
Benchmarks should be similar by industry, sector & objective, BUT each
project will differ
EXAMPLES
ECOMMERCE ARPU or RPV, R/T, R/V
MARKETING CTR, VIEWS, CONVERSION RATE
VIDEO TIME SPENT, # VIDEO VIEWED/VISIT,
# REPEAT VISITS, UNIQUES
PUBLISHING PV/V, V/UNIQUE, (TIME SPENT)
AD-DRIVEN IMPRESSIONS, MONTHLY PVs, PV/V
SOCIAL SHARES, LIKES, FAVES, FOLLOWS

Quantify Your Goals
3+ SHORT VIDEOS/
SESSION ACROSS
MULTIPLE SHOWS
(VIDEO DISCOVERY)
GET 30%
TRAFFIC
FROM
SOCIAL

Translate into UX Characteristics
< CONTENT DISCOVERY
< BROWSE
< SEARCH
< AUTO-SUGGEST
< NAVIGATION
< SOCIAL
ENGAGEMENT &
SHARING TOOLS
< SOCIAL WIDGETS

http://www.aguntherphotography.com/files/images/1833_large.jpg
You Must Choose.

Consensus - Focus & Prioritize

2-3.
CASVENTURES
COMPETITOR
METRICS ANALYSIS-DATA
+ CHARTS

Don’t Be Square.
29

Competitive Metrics Analysis

Time Spent (UX)

Bounce Rate

Pageviews vs. Time Spent

Traffic (Visits) Over Time - All

Traffic (Visits) Over Time - Subset
Visits/Time

Uniques/Time – 1 year (Subset)
30,000,000
22,500,000
15,000,000
7,500,000
0
Dec-‐16 Jan-‐17 Feb-‐17 Mar-‐17 Apr-‐17 May-‐17 Jun-‐17 Jul-‐17 Aug-‐17 Sept-‐17 Oct-‐17 Nov-‐17
Bleacher Report
Vimeo
TMZ
Gawker
The Onion
History
Comedy Central
Discovery
A&E
Funny or Die
Mashable
Boing Boing
Engadget
TruTV
DailyMoRon

Spreadsheet: Traffic Over Time
35 Competitors. Documented Source, Metric, Date Ranges, Date, Average
Below: Uniques/mo over time

Traffic (Visits) - Daily

Linear vs. Log Scales
http://blogs-‐images.forbes.com/naomirobbins/files/2012/01/linear_log.jpg

Demographics- Age (averaged)

Demographics- Age (with analysis)

Demographics- Average Income

Demographics- Income

Demographics- Education (with analysis)

Demographics- Education (computed ave)

Spreadsheet: Demographics Averages

App Ratings: Sentiment - Same Tactic
Google Play, iTunes # Downloads, Ave Reviews (Below)

Social- Same Tactic
# Social Followers (manual tally)

% Social Traffic

All Traffic Sources

4. REVIEW FOR
WEIRDNESS &
CASVENTURES
5.
ANALYZE FOR
MEANING

Review the Data for Anomalies, Exceptions
▪ THINK HARD & CAREFULLY
▪ Cross-Check & Correlate
▪ Be able to defend
▪ Take special note of any warnings

The Case of Netflix
▪ 700 Million monthly visits
▪ 126.5 Million weekly visits
▪ 80 Million weekly visits
▪ 10 Million daily visits
▪ 8 Million monthly uniques?

How many trees?
?

Correlations: Scatterplots

Finding Meaning….but Caution!

Tell AND Show

CASVENTUR6E.S
COMPETITOR
HEURISTIC ANALYSIS

Borrow from the Best
h t tp: //i m a g e s 4 . fanpop.com/image/p h o to s / 14700000/S e e -n o - ev i l -h e a r-n o - ev i l - sp e a k-n o - ev i l -monkey s-14750406 -1600-1200. jpg

Competitor Evaluation

Competitor Heuristic Evaluation
▪Deep-dive into high-performers- how & why
▪Correlated to the high-performing sites from metrics

Competitor Heuristic Evaluation

Try to find out the WHY behind the #s

Quantify Your Heuristic Analysis
Social Features Considered:
• Easy to find content based on tagging?
• Easy to follow/receive notification of upcoming shows?
• User ratings or comments?
• Save a show with login?
• Easy to follow/watch/share on social media?
!
Social-Top Features:
• Clips organized by
content tag
• “Follow” content
• Account profile with
saved shows

Quantify Your Heuristic Analysis
Video Player-Best Experience:
Hulu, CBS, MTV, Amazon
Instant Video
Video Player- Top characteristics include:
• Episode/series description obvious to user
• Previews, clips, and recommended content available
under or next to player
• Social Media sharing in player

Topline: Overall Rank

Actionable Takeaways (KPI focused)

CASVENTUR7E.S
INTERNAL METRICS
ANALYSIS

Internal Metrics: A Deep Dive
http://346fae7859434bda978f-‐1902f231618c5697bb2c852c565827b8.r12.cf5.rackcdn.com/wp-‐content/uploads/2014/02/diving1.jpg

Internal Metrics Analysis

Internal- Social Traffic Breakdown

Traffic Sources - Internal vs. Public

Key Internal Site Metrics- Top Sections

Internal Site Metrics- Entry, Exit Points

Internal Site Metrics- Paths
Top Sections (Comscore)
!
Top paths (Comscore)

Internal Clickpaths
Click path Analysis
from Homepage

Key Internal Site Metrics Over Time

Internal Site Metrics – Cohort Analysis
▪ Lean Startup / Agile technique
▪ Tracks KPIs per sprint/ at fixed time blocks to evaluate if there are
improvements from new product releases
http://idea-stack.blogspot.com/2013/04/quick-hack-setting-up-cohort-analysis.html

Internal Site Metrics- Mobile vs. Desktop
Web vs. mobile activity by Site Section

Demographics- More Detailed + Accurate

Demographics- New Information

Demographics: Audience Definition

Internal Site Metrics- Design Data

Internal Site Metrics- Device Data

CASVENTUR8E.S
INTERNAL HEURISTIC
PRODUCT ANALYSIS

Internal Heuristic Evaluation

Internal Heuristic, Focused on Key KPIs

Internal Heuristic- Deep Dive on Key Pages

Heuristic- Navigation vs. Metrics Paths

9.
CASVENTURES
THE SEXY PART:
THE FINDINGS

SEXY IT UP!
http://tng.trekcore.com/hd/albums/1x13/datalore_hd_027.jpg

Snoozeville
http://m.cdn.blog.hu/in/investo/image/Robot.jpg
http://cailincreature.blogspot.com/2010_06_01_archive.html

Don’t Make Me Think!

Risk of Wrong Conclusions

The path forward - based on data

KPI Impact is SEXY
$
:-)

Analogy: UX Awards
Youtube.com/UXAwards

CASVENTU1R0E.S
FOCUSED IDEATION

New UX Ideation Guided By KPIs, Metrics
UX GOALS BASED ON PRODUCT KPIS
- BUSINESS: Increase # full video views/session (with prerolls)
- USER: Find & watch videos, fast (in under N clicks/ X seconds)
http://www.sonicftp.com/news/images/guitarstorage_carousel.jpg

Example UX Ideation
ENGAGEMENT KPIS
- # Video Views
!
IDEATION: CONTENT DISCOVERY
- What are all the ways to browse or find similar content?
- What are all the ways to discover new content?
- “What if”….?
- What are successful competitors / “admirables” doing?
- Offline analogies?
http://www.sonicftp.com/news/images/guitarstorage_carousel.jpg

Seeking Inspiration Offline
OFFLINE SHOPPERS: BROWSE BY BRAND OR FUNCTIONALITY
ONLINE: BY SHOW, ACTOR, CHARACTER OR GENRE, LENGTH
thefashionspot.OxfordTech.us com/| GA: shop/Web 364135-Metrics boutique-l August of-the-2014
week-the-chanel-shop-at-bergdorf-goodman/

Brainstorming to Solve KPI Problems

Getting Creative- Summary
▪ ASK “WHAT IF” / “WHAT ARE” - FROM KPIS & TARGET ACTIONS
▪ LIST ALL THE POSSIBLE OPTIONS
▪ CONSIDER RELEVANCE LAST
http://openclipart.org/image/2400px/svg_to_png/185271/ico_light_bulb_2.png

KPIs, Goals Can Conflict (UX must Solve)
Final direction may need to balance competing needs
!
Ecommerce
▪ Business: maximum revenue/ visitor or /transaction
▪ Customer: cheapest, best option that’s found the most
easily (least clicks & time on site before transaction)
Publishing (ad-driven)
▪ Business: maximum ad impressions & PVs/visit
▪ Visitor: least ads, clicks & PVs to find & consume desired
content

CASVENTU1R1E.S
PROTOTYPE, TEST,
ITERATE (REPEAT)

Testing & Iterating Your Concepts
GET FEEDBACK BY TESTING BEFORE DEV.
▪Clickable Prototypes
▪MVT - Landing Pages
USE METRICS TO EVALUATE LAUNCHED FEATURES
http://www.matraxis.co.uk/services/ab-multivariate-testing/

Testing & Iterating Your Concepts
ANDREW MCKINNEY!
http://andrewmckinney.com/projects/weight-watchers-iphone-app/

Review of the 11 Steps
!
2. Get competitor metrics data from multiple sources
!
!
8. Review Internal Site (Heuristic, also Quantified)
!
9. Generate Actionable Findings inside a narrative
11. Test & Iterate

TOP 10
NEWBIE ANALYSIS
CASVENTURES
MISTAKES

Top 10 Newbie Metrics Mistakes
1. Not knowing what to look for, based on business & user goals
2. Using the wrong metrics & KPIs for the job
3. Believing any metrics are accurate – esp. public web
4. Being a robot. Think and be skeptical
5. Comparing across different time ranges
6. Using only 1 short time period; forgetting seasonality/ trends
7. Comparing across different sources with different names
8. Comparing internal to public metrics (correlation, not exact)
9. Not understanding why bad sites have good metrics.
10. Not asking WHY or HOW
DON’T BE A ROBOT.

Using the Wrong Metrics
http://www.clker.com/clipart-10628.html

Consider Seasonality
http://linkoroo.com/fun.php?funid=518&funname=Winter-Trees

When Good #s Are Bad.
HIGH ENGAGEMENT CAN BE FROM TERRIBLE UX.
Poor usability can lead to high PV/u, Time Spent, Deep & wide clickpaths
http://www.1800attorney.com/

Humans > Data

Steps
1. Identify key sites & KPIs
2. Get data from multiple sources
3. Graph it
11. Test & Iterate

Web Metrics Are NOT…
▪Accurate
!
▪Absolutes
!
▪Useful when quoted in isolation or without analysis
!
▪Useful without understanding the project & context

Theory of Relativity
chican-izmo.blogspot.com/2010/06/if-tree-falls-in-forest.html

Web Metrics Are…
▪Useful approximate data points
!
▪Fantastic for ALL STAGES OF UX
!
▪A means to get UX out of the gallery of opinions by quantifying
the value of UX based on performance
!
▪ A way to pinpoint top competitors for further review, ideas
!
▪Great to narrow product goals & engagement KPIs
!
▪Helpful with ideation through goal-focused brainstorming
!
▪ A means to improve performance through multivariate testing

SEXY

Show AND Tell the Story
lorenweisman.com/2013/06/21/music-marketing-plan/music-marketing-plan-storytime-artists-guide-show-and-tell/

Metrics Help Guarantee Success!

BREAK THANKS!
!
!
Bev May

PART II:
YOUR TURN!
METRICS WORKSHOP
!
!
Bev May

Q&A: DATA
VISUALIZATION
CASVENTURES
CHOICES

Which to use?
Education Level (internal) - 4 different levels
?

Which to use? Single Source %s
Education Level (internal)

Which to use?
% MALE vs. FEMALE (External Comparison)
?

Which to use? Amounts & Rank

Trendlines + Simplicity

Which to use?
Trends over time on multiple sites
?

Which to use? Trends over Time
Traffic over time

Which to use?
Gender vs. Traffic (Correlation)
?

Which to use? Correlations
Gender vs. Traffic

DATA DOWNLOAD &
CASVENTURES
ANALYSIS

Process Steps We’ll Do Today
2. Get competitor metrics data from multiple
sources
6. If Time- Examine High Performers (Heuristic,
Quantified)
11. Test & Iterate

Workshop Agenda
1. Form Teams
2. Define Project KPIs
3. Web Metrics- Download & Graph
4. Find Analysis & Meaning
5. Share Findings
6. (If Time) KPI-Focused Competitive Analysis
7. Resources + Q&A

Form Teams
▪ Teams of 2-4
▪ BUT RECOMMEND EVERYONE DOES THE WORK (IN PARALLEL)
▪ Introduce yourself to your teams!
NEEDS:
▪ Computer with Excel / Spreadsheet Program for data work
▪ Choose a Web-Based Sector – with established metrics
▪ 5 Competitors / Comparables (known, large sites only)
▪ Default sectors to research if you don’t have any: Publishing or Video

1. KPIs & Goals (Strategy)- 5-10 Mins
▪ Define Business Strategy (Revenue Model) & Business KPIS
▪ Define User Goals & Priorities
▪ Translate into UX Features, Goals
▪ Define Evaluative Metrics KPIs & Goals
▪ PICK 2-3 KEY WEB METRICS TO RESEARCH
▪ Sector & Sub-Sector
▪ Comparables
▪ “Admirables”
▪ PICK 5 SITES

BUSINESS (RED), USER (GREY) & UX (BLACK)
< VIDEO
DISCOVERY
< SOCIAL VISITS
SEE < FULL
(PAID) VIDEOS
< BRAND
AWARENESS
FIND + WATCH TOP
FREE CLIPS
EASILY SHARE
WITH FRIENDS

Define Your Evaluative Metrics KPIs
Benchmarks should be similar by industry, sector & objective
EXAMPLES
ECOMMERCE ARPU or RPV, R/T, R/V
MARKETING CTR, VIEWS, CONVERSION RATE
VIDEO TIME SPENT, # VIDEOS VIEWED/VISIT
PUBLISHING PV/V, V/UNIQUE, (TIME SPENT)
AD-DRIVEN IMPRESSIONS, MONTHLY PVs, PV/V
SOCIAL SHARES, LIKES, FAVES, FOLLOWS

Translate into Key UX Characteristics
< CONTENT DISCOVERY
< BROWSE
< SEARCH
< AUTO-SUGGEST
< NAVIGATION
< SOCIAL
ENGAGEMENT &
SHARING TOOLS
< SOCIAL WIDGETS

(Quantify Your Goals)
3+ SHORT VIDEOS/ SESSION
ACROSS MULTIPLE SHOWS
(VIDEO DISCOVERY)
GET 30%
TRAFFIC FROM
SOCIAL

KPIs & Goals (Strategy)- Share!

2. Get Competitor Data into a Spreadsheet
▪ Download/ manually transcribe from many sources- never just one
▪ ALEXA: http://www.alexa.com/siteinfo/nameofsite.com
▪ COMPETE: https://siteanalytics.compete.com/nameofsite.com
▪ QUANTCAST: https://www.quantcast.com/nameofsite.com
▪ SIMILARWEB: http://www.similarweb.com/website/nameofsite.com

2. Get Data, Put Into Spreadsheet
TODAY AS A GROUP- WEB METRICS
• Time Spent – Engagement (Bar Chart)
• Bounce Rate
• Demographics-Gender – Audience (Bar Chart, Pie Chart)
• Unique Visitors over Time – Popularity (Line Graphs)
• Comparisons- Bounce Rate
• Your KPIs
• Scatterplot Comparisons

A. Time Spent on Site- Alexa – 5M
http://www.alexa.com/siteinfo/nameofsite.com
▪ Make a new spreadsheet with companies as column A and column B
labeled “time spent”/”minutes on site”
▪ Enter data from Alexa in column B.
▪ Mark date range, source, URL for your data (click ? - trailing 3 mo)

A. Time Spent- Graph
▪ Select all and sort by column B
▪ Select data, then choose Insert > Column Chart while Data is
selected.

So What?

B. Bounce Rate- Alexa
http://www.alexa.com/siteinfo/nameofsite.com
▪ Make a new chart tab with same companies as column A and
column B “Bounce” - under a new tab (copy chart, delete rows 2 on)
▪ Enter data on % Bounce in column B as a whole # (“78” for 78%)
▪ Mark date range, source, URL for your data (click ? - trailing 3 mo)

B. Bounce Rate- Graph & Analyze
▪ Select data, then choose Insert > Column Chart while Data is
selected.
▪ Is this a problem?
▪ ASK WHY??

C. Traffic Sources- SimilarWeb – 5M
http://www.similarweb.com/website/nameofsite.com
▪ Make a new chart tab with same companies as column A and
column B “Social” - under a new tab (copy chart, delete rows 2 on)
▪ Enter data on % Social in column B as a whole # (“78” for 78%)
▪ (Add other columns for other traffic types. Label each column)
!
!

C. Traffic Sources-% Social
▪ Select all data, then choose Insert > Column Chart while Data is
selected.
!
!

D. Demographics- Quantcast – 5M
https://www.quantcast.com/nameofsite.com
▪ Use Quantcast on quantified sites. Mark date range, source, URL
▪ Examples: A&E (Aetv.com - partial), Bleacher Report, TMZ, Gawker,
The Onion, Quantcast.com
▪ Gender tab (not gender “Index”)
▪ Make a new chart with same companies as first tab- under a new tab
(copy chart, delete rows 2 on)
▪ Enter data on % Male in column B as a whole # (“78” for 78%)
▪ Do the same for Female in column C
!
!
Site X

D. Quantcast - when there’s Data
▪ Go to https://www.quantcast.com/nameofsite.com

D. Demographics- Gender- Alexa
▪ Pro: Available free. Con: Relative Rank #s
▪ Can get real #s for $50/mo
▪ If don’t want to pay and just want to get comparisons- can assign
numeric estimates to Alexa bars, assuming linear scale (-5 to +5)
!

D. Demographics- Gender
▪ Select all data, then choose Insert > Column Chart while Data is
selected.

D. Demographics- Gender -Alt View
Always consider the best visualization of data

D. Demographics- Gender (Single Site)
Always consider the best visualization of data

E. Uniques over Time- Compete – 5 M
▪ https://siteanalytics.compete.com/nameofsite.com
▪ 6 Month Tab. Mouseover Dots.
▪ Make a new tab with same companies as column A and columns B-G each a
month - under a new tab (copy chart, delete rows B on)
▪ Manually Enter Uniques Data into columns B-G
▪ Column H: mark a column for Average

E. Uniques over Time Spreadsheet
▪ First Column: Sites. Columns B-G: 6 Months data as #s with labels
▪ Mark the date range, source, URL for your data
▪ Add an AVERAGES column on right. Leave blank for now

E. Uniques over Time- Graph! – 5 M
▪ For 5 Sites & 6 Months- Select cells 1A-6G (include months, names of
sites and numbers, but excluding average column)
▪ Choose Insert > Line Chart while Data is selected.
▪ Right-Click on the new chart. Choose “Select Data”
▪ Click “Switch Row/ Column”, then OK (“Plot Rows as Series” on Mac)

E. Uniques- Outliers
▪ Sometimes Excluding Data in Graph Can Reveal More Detail

E. Bounce Rate alternate- Similarweb -5M
▪ http://www.similarweb.com/website/nameofsite.com
▪ Open up original Bounce Rate Tab
▪ Use to compare against Alexa- find anomalies
▪ Add 1 more Column- Enter Alexa Data (and note source!)
▪ Sort by both columns

E. Bounce- compare – 5 M
▪ Eyeball for major differences in Data
▪ Do a double graph of both values (Insert > Chart - bar chart)
▪ Anything interesting?

F. Your KPIs - Download, Graph in a new tab
▪ Download/ manually transcribe some other interesting data points
▪ Ex: Bounce Rate, Page Views, Visits/Unique
▪ ALEXA: http://www.alexa.com/siteinfo/nameofsite.com
▪ COMPETE: https://siteanalytics.compete.com/nameofsite.com
▪ QUANTCAST: https://www.quantcast.com/nameofsite.com
▪ SIMILARWEB: http://www.similarweb.com/website/nameofsite.com

Review, Analyze & Think!
▪ DON’T BE A ROBOT.
▪ Does the data make sense?
▪ Do the metrics match up across sources?
▪ Double check anything suspicious.
▪ Note anomalies

Unlock the SEXY. Write relevance for UX

Scatterplot- Find Patterns – 10 M
▪ Use your data for % social vs. Age or any other point to form a scatter
▪ Make a new tab, copying over all of the % Social Data
▪ Sort by Site Name (column A)
▪ Go to Age tab, sort by Site Name (col A), then copy Column B data
▪ Go back to new Scatter tab, paste in Age data in Column C
▪ Make sure both tabs were sorted by name first to match the data

Scatterplot- Look for Patterns
▪ After sorting, choose 2 columns of just the data- not labels
▪ Sort data by one column (or both)
▪ Insert Scatterplot
▪ May need to invert to plot X:Y instead of Y:X (edit chart data > plot rows/
columns as series in lower left on a Mac; on a PC, invert plot)
▪ Add labels to graph, cross-check graph values

Scatterplot- Any Patterns? Make More!
▪ Won’t always get interesting results…. make more!

Scatterplot- Correlation vs. Causation
▪ ASK WHY?
▪ Be careful with your conclusions
▪ Regression analysis
▪ Consider alternate reasons

Review & Have Details Documented
▪ Document in Excel & Presentations
▪ Sources
▪ Date ranges- exact
▪ Date data was obtained
▪ Name of metric
▪ Source URLs

SHARE YOUR
CASVENTURES
RESULTS!

EXAMINE TOP
PERFORMERS
CASVENTURES

If Time: Examine top Performers & ask…
HOW?
http://openclipart.org/image/2400px/svg_to_png/185271/ico_light_bulb_2.png

Asking How- Documenting
▪ Work in Teams
▪ Define “success metrics” & KPIs
before examining
!
▪ Take Screenshots
▪ Make notes on pages & note URLs
▪ See what’s similar/ different
across sites
▪ See what stands out
▪ Form hypotheses on how/ why

“Quantifying” your Approach
▪ Define “success metrics” & KPIs and features/approaches that support
them
▪ Rank competitors based on those features/ approaches
▪ Graph your results

Some Public Web Metrics Sources
▪Quantcast
▪Compete
▪Alexa
▪SimilarWeb
▪Comscore (Paid)

Some Private Internal Tools
More accurate, but requires developer implementation &
unlikely to provide competitive insights
!
▪Google Analytics (free)
▪ Yahoo! Analytics (free)
▪ Coremetrics
▪Omniture
▪WebTrends
▪ Adobe Site Catalyst
▪ Piwik, Spring, Woopra,
Clicky, Mint, Unica, HitsLink,
OneStat, GoStats, NextStat,
WebTrekk, Fireclick, AWStats,
GoSquared, Firestats, SiteMeter,
Mixpanel, Foxmetrics, HisStats,
Reinvigorate, Open Web Analytics
inviteads.com/wp-content/uploads/2013/07/Heatmap.jpg

Other On-Site Metrics Tools
▪ Optimize.ly – MVT & A/B Testing
▪ Chartbeat – Real-Time Data
▪ Kiss Metrics – Patterns over Time/ Click path
▪ Unbounce, Launchrock – Landing Page MVT
▪ Google Web Optimizer
▪ Clicktale, Crazy Egg –Heatmaps
▪ Google or Bing Webmaster Tools – SEO
▪ Experian Hitwise: Demographics
▪ Online testing: Loop11, UserTesting, KissInsights
▪ More: InspectLet, GoingUp, eTracker, Grape
Webstats, AT Internet, IBM NetInsight, Webalyzer,
StatCounter, BBClone, Piwik, Stuffed Tracker
en.wikipedia.org/wiki/Web_analytics

Social Tools
▪ Twitalyzer
▪ Facebook Insights- Likes, Talking About, Engagement Rate
▪ Klout (not recommended)
▪ Topsy- what’s shared
▪ Sprout Social
▪Google Analytics- Social
▪Oracle Social
▪ Attensity
mikefrancois.net/wp-content/uploads/2011/01/LinkedIn-Map1.png

Mobile Tools (for Native Apps)
▪ iTunes
▪Google Play
▪Crittercism
▪Mobile App Tracking
▪WebTrekk
▪ Ad-X
▪ Fiksu
▪ LinkShare
▪ Flurry
▪ InMobi
mikefrancois.net/wp-content/uploads/2011/01/LinkedIn-Map1.png

UX + WEB METRICS
How to Benchmark, Measure
& Evaluate UX Impact
General Assembly
August 03, 2014
THANKS!
!
!
Bev May

Web Metrics: A Primer for UX Pros - UPDATED

More Related Content

Viewers also liked (16)

Similar to Web Metrics: A Primer for UX Pros - UPDATED (20)

More from Oxford Tech + UX (13)

Recently uploaded (20)

Web Metrics: A Primer for UX Pros - UPDATED