Organizations today use web analytics tools to track hits and page views but a hit is only a line in a web log. What do you do with that data?
Web analytics is no longer just about hits and page view traffic. Mechanisms are available that allow a practitioner to get a fully-rounded view of the user via their navigation behavior. In the realm of usability, we tend to concentrate on qualitative information, which serves us well. However, quantitative data can be more salient to clients. Web analytics provides this quantitative data and offers a perfect complement to usability studies. Demographics are only part of the equation. As Joseph Carrabis, Principal of Next Stage Evolution, states, “I know who you are within two seconds, but those are the cheap seats. That doesn't interest me anymore.” Companies want to determine how the online behavior of their website visitors aligns with their business goals and strategic direction. Or as Judah Phillips, Director of Web Analytics at Reed Business, says, “Web analytics refers to the collection, measurement, reporting, and analysis around qualitative and quantitative information related to the behavior of an online audience.”
This presentation will inform usability practitioners about the pros and cons of incorporating web analytics into their work. It will discuss different approaches to web analytics, including:
“Homepage hits”
“Integration of website activity… with the offline business”
“Navigation behavior of the visitor”
“Psychological engagement of the visitor”
“Physiological measurements of the visitor”
“Return on investment (ROI), as measured by key performance indicators (KPI)”
- Paul Legutko, Vice President of Analytics, Semphonics
Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...
Convergence of Usability and Web Analytics
1. The Convergence of Usability
and Web Analytics
Dan Berlin
Mike Ryan
Bob Thomas
May 28, 2008
Source: http://crazyegg.com/. Reprinted with permission of Crazy Egg. Inc.
2. The Convergence of Usability and Web Analytics
Agenda
• What is Web Analytics
• How Organizations Use the Data
• Pros & Cons of Using Web Analytics for Usability
• Key Players
• Conclusions
• References
3. The Convergence of Usability and Web Analytics
What Is Web Analytics
“Web analytics is the study of the
behavior of website visitors [and]
refers to the use of data collected…
to determine which aspects… work
towards the business objectives.”
- Wikipedia
4. The Convergence of Usability and Web Analytics
What Is Web Analytics
• “Web analytics brings in the data to provide a 360
degree view of the customer.”
• Track demographics and behaviors through data modeling
• Inspect web log data and compare it to baseline
historical data to draw out behaviors and demographics
An Exit point report may show you what visitors were looking for.
(Source: http://www.summary.net/manual/tutorial/lesson7-1.html)
Before After
5. The Convergence of Usability and Web Analytics
Demographics
“I know who you are within 2 seconds, but those are the cheap seats.
That doesn’t interest me anymore.”
-Joseph Carrabis, Principal, Next Stage Evolution
6. The Convergence of Usability and Web Analytics
Behaviors
“Web analytics refers to the collection, measurement, reporting, and
analysis around qualitative and quantitative information related to
the behavior of an online audience.”
-Judah Phillips, Director of Web Analytics, Reed Business
7. The Convergence of Usability and Web Analytics
Statistical Analysis
“The best web analytics… [provide] actionable insights into the
online channel from all different perspectives.”
-Paul Legutko, VP of Senior Research, Semphonic
8. The Convergence of Usability and Web Analytics
How Organizations Use the Data
• Website optimization
• Actionable items and decision support
• Alignment with business goals
9. The Convergence of Usability and Web Analytics
How Organizations Use the Data
• Website optimization
• Compare data to databases
• Where are users going?
• What is their behavior?
• Target data to those who can use it
• Statistical analysis
• Sourcing, channel return on investment (ROI)
10. The Convergence of Usability and Web Analytics
How Organizations Use the Data
“Optimization concerns not only… how the page might be designed.
It also has to do with things like… ROI.”
-Paul Legutko, VP of Senior Research, Semphonic
11. The Convergence of Usability and Web Analytics
How Organizations Use the Data
• Actionable items and decision support
• Modeling is the key tool
• Can give logic/decision trees
• Can supply quantitative reasons for changes
• Always the client’s decision
• Websites can customize sites to users
12. The Convergence of Usability and Web Analytics
How Organizations Use the Data
“Give them action and reasons for the action, and best possible
outcomes of the actions, so that they can make their own decisions.”
-Joseph Carrabis, Principal, Next Stage Evolution
13. The Convergence of Usability and Web Analytics
How Organizations Use the Data
• Alignment with business goals
• Can’t collect the correct data without knowing goals
• Business goals and rules are paramount
• Is the site achieving its business goal?
• How much are certain types of customers spending?
• What is the conversion rate?
• Give a fully rounded view of the customer
14. The Convergence of Usability and Web Analytics
How Organizations Use the Data
“What’s the reason the website exists? Without knowing the
goals, it’s really hard to apply data to solve the problem.”
-Judah Phillips, Director of Web Analytics, Reed Business
15. The Convergence of Usability and Web Analytics
Using Web Analytics Data to
Understand Usability Issues
• Pros
• Tracks JavaScript events
• Behavior analytics
• Lots of data
• Confirm or deny data found in a usability test
• Form hypothesis to do usability testing
• Information architecture
• “Web analytics can tell you what’s working and
not working”
16. The Convergence of Usability and Web Analytics
Using Web Analytics Data to
Understand Usability Issues
• Cons
• There’s no human element
• Intimidating amount of data
• Isolated from users
• Looking at data, not real behavior
• Just current users
17. The Convergence of Usability and Web Analytics
Cons
“…usability studies that companies do tend to be divorced… from
web analytics.”
-Paul Legutko, VP of Senior Research, Semphonic
18. The Convergence of Usability and Web Analytics
Key Players
Free Google
Analytics
Web-Based
Applications
compete.com crazyegg.com hitwise.com indexTools quantcast.com
Software
Applications
ClickTracks Coremetrics HBX Analytics
Visual Sciences
WebSideStory
Omniture
Discover
Genesis
NetTracker
SiteCatalyst
TouchClarity
SAS
WebHound
Unica
Insight
NetTracker
WebTrends
19. The Convergence of Usability and Web Analytics
Conclusions
• Overlap of web analytics and usability
• Align with business objectives
• Incorporate both into iterative design
best practices
• Use both: one informs the other
• “Web analytics is not only a
technology but also a way of mind”
20. The Convergence of Usability and Web Analytics
References
• Web Analytics Companies
• www.clicktracks.com
• www.compete.com
• www.coremetrics.com
• www.crazyegg.com
• www.google.com/analytics
• www.hitwise.com
• www.omniture.com
• www.quantcast.com
• www.sas.com
• www.unica.com
• www.websidestory.com
• www.webtrends.com
• Web Standards Committee
• www.webanalyticsassociation.org
21. The Convergence of Usability and Web Analytics
References
• Participants’ Websites
• www.nextstagevolution.com
• www.reed-elsevier.com
• www.semphonic.com
• Blogs on Roundtable Discussion
• Joseph Carrabis
www.bizmediascience.com/2007/10/visitors_truth_false_informati_1.html
• Paul Legutko
legutko.typepad.com/waa/2007/10/web-analytics-t.html
Hinweis der Redaktion
How web analytics and usability can inform each other And help you achieve your business objectives by looking at the demographics and behaviors of your website visitors
What is web analytics: define it How organizations use the information: to improve usability of website and to align their findings with business goals: USE WEB ANALYTICS to define what’s happening, and rely on USABILITY to figure out why it’s happening Pros and cons of web analytics Key players: who’s making these software and web-based applications for web analytics (e.g., Google, Crazy Egg, WebTrends) We’ll then conclude and provide references
Web analytics is the study of the behavior of website visitors and refers to the use of data collected from a web site to determine which aspects of the website work towards the business objectives. What are you looking at: BEHAVIORS of web site visitors: WHERE they’re coming from (Google AdWords). WHERE they LAND . HOW THEY TRAVEL (paths they take). And WHEN THEY’RE LEAVING or HOW LONG they stay on the site or page. The most important part: seeing how online behavior of your website visitors aligns with your business goals For example, which landing pages encourage people to make a purchase Or the time visitors spend on your website and your “bounce rate”: % of your web traffic that stayed on your website for a period of time (e.g. less than 10 seconds) or who only viewed one page of your site before leaving Or you might find that visitors spend a really long time on your website, which might tell you EITHER that they’re engaged by your site OR they have problems with your IA, or finding the right content
You want to look at the entire customer experience: measuring demographics and behaviors Using web analytics software to create data models so that you can analyze those demographics and behaviors, for example, the navigation path that 18-34 year olds living in the Northeast who come to your website spend looking at your calling plan HISTORICAL DATA. And then seeing how past behaviors differ from current behavior. What have you done differently to encourage visitors to travel through your website, and what current and unique visitors are you attracting? WEB ANALYTICS can often tell you the WHAT but not the WHY. You may need usability for that. EXAMPLE: EXIT POINTS It could be that LAST YEAR you designed a long and complicated path for visitors to sign up for a calling plan with a cell phone carrier, and that users were dropping off and giving up. And now you’ve designed a better method through persona development targeting different demographics so that visitors can easily figure out what kind of calling plan they want based on what kind of person they are: a single parent with two teenagers who wants a family calling plan with local rates. analyzing exit pages to look for sources of lost revenue: Did visitors leave because they were lost or confused rather than moving on to a purchase page? An Exit Point report can also be used to understand the visitors’ goals. If you can identify certain exit pages that are obvious end-points to a search, then you might want to make navigation to those pages more prominent to help visitors find them
Roundtable discussion with 4 people who have experience in the web analytics field: Judah Phillips, Director of Web Analytics, Reed Business Information Paul Legutko, Vice President of Analytics, Semphonics, Web Consultancy Jack Carroll, Principal, VOC Partners Joseph Carrabis, Principal, NextStage Evolution While Judah and Paul are concerned with how web analytics align with the business goals of an organization Joseph Carrabis is interested in the psycho-behavioral view of web analytics, based on the predatory behavior of animals: what information do they detect and how do they detect it?) Relies on preattentive processing : to determine what people are looking at and what they’re engaged by Examples: heat maps from clickstream data or from eye tracking ONE THING: Demographics: age, gender, income, education, location ANOTHER THING: Psychological engagement of visitors (how long do they remain on a page and what grabs their attention?)
You want to look at the paths that visitors take to: get to your site (e.g., from Google search or from a link from another site) the trail visitors follow once they’re in your site where and when visitors leave your site (e.g., leave the purchase area of your site after 5 minutes, which could indicate they had usability problems with the purchase area: couldn’t see what was in their cart, they were concerned about the security of their transaction, they wanted to pay by another method than those presented, e.g., PayPal) The Data Model The data model is divided into hits. [A hit is] anything that gets recorded into a log file.” Conceptual model You use web analytics tools to track hits and page views – but all a hit is, is a line in a log QUANTITATIVE: effectiveness, efficiency, & satisfaction (through utests) QUALITATITVE: think aloud methods (through utests)
Takes into account: What you’re measuring: page views, hits, impressions, visitors (returning and unique) Variables you’re using: page name, site section, something you want to measure Taking into account The 360 degree view of the customer To align with your goals for the online channel: They want FOR EXAMPLE, a 25% increase in revenue and a 15% conversion from trial to full versions of a product this much money and conversions (people coming to the site and purchasing something) You want to find patterns in the data to make recommendations: such as trying to attract 18 to 34 year olds who reach your golfing website from Google AdWords and then go down a path where they purchase a new line of golf clubs recommended by Tiger Woods
Collect data and compare it to historical DBs Can determine where users are spending the most time - for one reason or another; are they confused or intrigued? Are users going back to a page repeatedly? How much of the website are they actually using? Only certain people in a company can actually use this data, so the data collected is gear towards those departments - marketing, customer service, etc Are users going to pages by chance? Optimization doesn’t have to do with website design
Through data collection and the historical DB, various models can be developed and tested The key is giving many ‘what if’ scenarios Quantitative measures always impress people (particular challenge in usability) WA practitioners give the models and suggestions to clients, but always leave the decision in their hands Try to guide them to the most practical solution If a company knows the user’s demographic, either at the onset of the visit, by learning from his behavior, the website can be quickly customized to that user. Button in the car for different people example.
Can’t put the cart before the horse What does the client want to get out of the website? What does the client want the customers to get out of the website? Are these goals being met? Are they missing out on a market segment? Historical DBs can determine a user’s demographics very quickly
As we have been saying, WA is no longer just about logging hit rates and page view traffic Anything that triggers a javascript event can be tracked and run through a huge behavioral database That includes mouse movements, selections, hovers Big behavioral database with 75 variables Use this data for Behavior Analytics to calculate For example user frustration, through defined mouse movements Tracking all of your users over time will generate a lot of data Can be a supplement to usability testing to confirm findings Discoveries in data can be used to form hypotheses to perform study Especially helpful in uncovering all kinds of IA issues: navigation, organization, discoverability, where people get stuck, what interaction sequences they are using
Obvious cons are that there is no human element involved Only as smart as the behavioral database is A lot of data. Takes a great understanding of stats and usability to decipher Isolated from users. There are a lot of blanks to fill in No context No verbal explanation of actions No facial or body cues You’re looking your current users not potential users
What tool you want depends on what you want to do and how much you want to spend. An expensive tool or a cheap tool only provides data. You still have to analyze that data. Google Analytics: free, often used with Google AdWords to track success of your online Google Search campaigns: TAGS PAGES. USE IF WANT to measure ROI on Google AdWords campaigns and optimizing search. So you can use it to drive traffic to your website, but then what happens after your visitors get there? Crazy Egg: popular web-based app. Creates heat maps based on clickstream data (similar to eye tracking heat maps). indexTools: just acquired by Yahoo 4 BIGGEST VENDORS: Coremetrics, HBX Analytics (WebSideStory), Omniture, WebTrends. Clients who chose the ASP model pay for basic analytics using a cost per million page views or server calls. Those who chose software pay either traditional license fees or the same ASP-style volume pricing, depending on the vendor. AND THEN THERE’S A USABILITY ISSUE. FORRESTER RESEARCH: THESE tools have usability problems — some serious. Common problems included unfamiliar terminology, task flows that don’t match how a user would logically approach the problem, and interfaces that behave differently in different situations, causing user confusion. So many companies buy the web analytics software and then don’t know how to analyze the results – because they haven’t determined ahead of time what their objectives are – what they want to analyze and why. In Web Analytics: An Hour a Day, Avinash Kaushik recommends following the 1/9 rule: you’ll devote $9 out of every $10 on analyzing the data. SPEND 10% of your budget on tools and 90% on people who can give you insights into the data
Overlap of web analytics and usability Align with business objectives Incorporate both into iterative design best practices Use both: one informs the other “ Web analytics is not only a technology but also a way of mind” Drink the kool-aid
Here are sites for the major players in web analytics And the committee that is establishing standards for web analytics ** “It’s a field with no real standards.” challenge: reports are similar, overlapping metrics. But when look closer see different collection methods: sampling methods, data collection methods, algorithms, they all differ The true goal of standardization is not the end result but standardizing on the process involved to analyze your results “ at end of the day, it’s still writing a log file”
Here are the company websites for the participants that came in for our summit Two of them wrote about it on their respective blog sites.