Advanced Web Analytics - Penn State Web Conference 2010

Advanced Web Analytics

Shelby Thayer

Joshua Ellis

Outreach Marketing and Communications

Before we even mention measuring a website, we need to truly understand

the purpose of the website.

Increase:ProspectsApplicantsEnrollmentsDonationsAlumni memberships

Improve:Customer serviceUser engagement

Key Performance Indicators

Measures to use to see if your website is meeting its objective(s).

Increasing Applicants

• % visits that view “apply” page

• % visits that complete the goal (convert)

• % visits that view “apply” page but don’t complete the goal (fallout)

Improve Customer Service

• % visits using chat

• % visits using knowledge base

• % visits subscribing to alerts

Increase User Engagement

• Average time on site

• Average page depth

• % visits that comment on the blog

• % visits from RSS feeds

• % visits that download a PDF

Competitive Intelligence(Alexa, Hitwise, Compete, etc.)

Voice of Customer(Feedback forms, surveys, usability testing, etc.)

Web site BehaviorOnsite Analytics

(Google Analytics, Omniture, Webtrends, etc.)

Web site BehaviorOnsite Analytics

(Google Analytics, Omniture, Webtrends, etc.)

Data Limitations

Data Limitations

Accurate Precise

Accurate vs. Precise• Being precise is predictable and reproducible.• Trends are the key.

Source: Web Analytics 2.0, Avinash Kaushik (images from Wikimedia Commons)

Trends vs. Hard Numbers

Always use trends – what’s changed?• Our visits increased by 5% over last month.

• Our bounce rate is down 4% over last week.

• 35% of visits are from search engines – this is up 7% from last month (great job PPC team!)

Hard Numbers – Essential for Context

Trends are important, but hard numbers give context.

Hard Numbers – Essential for Context

Trends are important, but hard numbers give context.

Analytics and Marketing Efforts

Optimize marketing efforts across all marketing channels.• Email• Paid Keywords• Banner Ads• Print Collateral• Social Media

Track Everything

Use Google’s URL Builder to track links from campaigns from any channel.

Even if you’re using a URL shortener – use a trackable URL as well.

What’s the point when we can use URL shorteners?

URL shorteners only track to your site. They don’t track on your site.

You need to track beyond the click-through.

Use Vanity URLs

www.mysite.psu.edu/vanity

Beyond the Click-through

If you bring a million people to your website, does it matter if they don’t do anything on your website?

But all my “actions” happen off-site.

If you can put code on the third-party site, you can use the Google Analytics Tracking Code (slightly modified) to capture information on the third-party site.

No access to third-party site?

Measure to the point of going off-site to complete action.

• Use GaAddons.js• Use onClick or onSubmit event handler

If possible, put the “thank you” page on your website.

Micro and Macro Conversions

Macro conversions – the big goal (apply, donate, enroll, download, comment, view details page)

Micro and Macro Conversions

Micro conversions – smaller, but important actions (download, click link to application, view video, view the “donate” page)

Credit Allocation

Which campaign gets credit for a conversion?

•Visit #1 – click on a paid keyword (doesn’t convert)•Visit #2 – click on a banner (convert!)

Use “utm_nooverride=1” within the trackable URL

Be consistent if you use this!

Segmentation

Absolutely nothing is more important in analytics than segmentation. Why? Because monoliths don't come to our websites. Yet most of our reporting and analysis happens at the aggregate level.

Avinash Kaushik, Web Analytics 2.0

Segmentation

NY Yankees team batting average = .273

• 1, 2, 3 batters average .354• 4, 5, 6 batters average .289• 7, 8, 9 batters average .139

Segmentation

Segmentation

Total website pages per visit = 4

Returning visitor pages per visit = 8• Coming from campaigns = 10• Organic = 6

Mobile device visitor pages per visit = 1

Segmentation

Now that you have your segments, go back to your website goals and KPIs. Split them out by segment.

Goal: Increase online applicantsKPI: % visits that view apply page• new vs. returning• campaign vs. organic traffic• campaign A vs. campaign B

SEO and Brand

Measuring SEO

Webmaster Tools• Links to your site• SERP impressions vs. clicks• Duplicate title and meta descriptions• Crawl errors• Crawl stats

Analytics Tool• Traffic from search engines• Traffic organic vs. paid search • Branded vs. generic terms (compare performance)• Internal site search keywords

SEO and Landing/Entry Pages

On page optimization is crucial to SEO

• Bounce rates• Conversion rates• Optimize the right pages for the right terms

• “keyword X” – all organic should be going to the same landing/entry page.

Measuring Brand

• Direct traffic (typing in URL, bookmarking)• Traffic from branded keywords• Conversions from branded vs. generic

keywords

total direct traffic + total traffic from branded keywords

total visits from search engines + total direct traffic

Monetize, Monetize, Monetize

• Page Value ($Index)

• How much does poor design cost?

Page Value ($Index)

Give your goal a value. Then you can see what pages are contributing most to your goal.

How much does poor design cost?

Resources

http://delicious.com/shelbythayer/psuweb10-wa

http://webstandards.psu.edu/analytics

Thank You

Shelby Thayer (@shelbythayer)Joshua Ellis (@apostate)

Questions?