MNteractive.com

Thanks to Evantage Consulting for sponsoring last night’s UX Meetup.  Toby won the $50 drawing for Wilde Roast!  Andrew Janis drew in a big crowd last night, we packed the room, and there was a lively and engaging 2-hour discussion of web analytics, reviewing the Animal Humane Society website as an example.

Below are a free form list of my notes from the evening:

AHS site redesign background: Online adoptions were the most frequently visited part of the site

Analytics on the site:

what the move viewed content on the site is (how many times each page is called)

adoption page is more highly viewed than the home page

URL’s that make sense are good for analytics team

Unique pageviews – suggests that a number of people are hitting each page

Bounce rate – a number of people that view that page when they leave the site

Bounce rate of 18% is low, under 30% is good

Bounce rates are really good indicators of whether or not your site is providing people what they’re looking for

Look at how long people are spending on your site

5-second test also helps understand

BOUNCE RATE IS A REALLY INSIGHTFUL METRIC

You can exclude traffic from an IP address so you’re only looking at external visitors, not tracking from within your company

Google Analytics has trouble with flash – it can be done, but it’s a development effort, it can record them as page views if programmed that way

If it’s not a page load, it needs to be programmed separately

Interesting to look at pre-redesign and redesign

Bounce rate went down, which could be OK because people might now be finding information faster

Search analytics

Google Analytics has been free for 3-4 years

Analytics is much less about absolute numbers and much more about TRENDING, all tools will have slightly different numbers, but the trends will look the same

Segment traffic based on new or returning visitor, where they came from (search engine, paid ad, referring website), segment by search queries and track what they did across the website – can create user profiles with that

Pull data into analytics and segment based on data that’s been tagged by audience

Large number of users are not entering via home page

What is a representational timespan?  It depends on the business.  Ecommerce business YOY, if trying to build traffic, might look quarter over quarter

WHEN DOING A REDESIGN, MIGHT LOOK OVER THE PAST FEW MONTHS

LOOK AT SEARCH RESULTS THAT RETURN NO SEARCHES IF YOU’RE GROWING CONTENT

KPI’s: Must be an actionable metric

Building dashboard for clients that show KPI’s, and honing in on most important metrics to their business

4 metrics for ecommerce, traffic, conversion rate, average order value – if any is moving, you want to know that

It’s very important to stay focused on business objectives

Google Analytics – can create a custom segment and pull data in that way

2/3 of traffic is returning visitors – a lot of loyalty with AHS

Can compare new users with returning visitors

Can also see the window between user visits

60-70% of your traffic coming from search engines is common, assuming the site is optimized

80% of all site visits are from paid or organic search, so if  yours is much lower, that’s a good sign to invest in SEO

Direct traffic is a good measure of brand awareness (also means someone typed the URL, bookmarked it)

Time on site is really tricky

Session associated with each web visit – sessions usually set up 30-60 minutes, if you’re inactive it may stop your session

If you leave open a tab and go back to it a while later, doesn’t measure it if the session has been closed

What browsers people are using, what OS, connection speed, device, screen resolutions

Can tweak the results to show up in google how you want to – it’s a google tool (Google Webmaster Tools)

Google is not as good at clickpath analysis – doesn’t give you in-depth tracking

Exit rates are good to see, especially if you have a long form or an order checkout process across multiple pages – you specify which path they should take

WHERE IS THE HIGHEST FALLOUT RATE?

Industry average is 40% for getting through a shopping cart order pages

Hotels.com – analytics and epinions labs teamed up and used Tea Leaf to learn little tweak to the site

Google Analytics usually doesn’t slow down a site because the tag is at the end of the page and is highly cached

Other services can slow down your site

Goals – Google lets you set up 4 goals – typically ecommerce site is buy something, sign up for emails – it’s all about conversion activity – you want to get as close to conversion as possible

GOALS ARE AS CLOSE TO THE END CONVERSION AS YOU CAN GET

With goals, can see what pages are contributing most to people converting – what content is generating the actions you want people to take

BUILD DATA IN PROCESS

IT’S A GOOD IDEA FOR PRACTITIONERS TO GET A LOGIN TO THEIR WEB ANALYTICS PROGRAMS, not just reports

“Web Analytics Demystified” by Eric Peterson