Top Banner
20

eMetrics Stockholm 2009

Oct 28, 2014

Download

Business

Lars J

Presentation.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: eMetrics Stockholm 2009
Page 2: eMetrics Stockholm 2009

eMetricsOctober 12

2009

Page 3: eMetrics Stockholm 2009

www.markred.se excellentanalytics.com 

twitter.com/analyticsplugin twitter.com/markred

Who Is Mark Red?

Page 4: eMetrics Stockholm 2009

Who is Lars Johansson?

www.WebAnalysts.Info

Page 5: eMetrics Stockholm 2009

What Is a Landing Page?

• First page upon website entry (from link, ad, etc.)• Page part of main website• Separate campaign page

Page 6: eMetrics Stockholm 2009

Why On-Site Optimization Matters

Doubling the conversion rate will have the same effect as doubling the ad spend.

Page 7: eMetrics Stockholm 2009

What Is Continuous Improvement?

• Always evolve and evaluate changes to the website• Always strive to increase revenue, decrease costs, and

improve customer satisfaction. You can also have an impact on brand perception.

• Use web analytics, web surveys, usability testing, and heuristic evaluations to discover flaws and prioritize testing

• Describe the hypothesis • Test variations on a limited share of visitors

Page 8: eMetrics Stockholm 2009

(cont.)• Show variations during the same circumstances• Only one version per visitor/browser• Analyze results once sufficient data has been collected• Keep/implement the winner• Start a new round of testing – you can always improve!

Page 9: eMetrics Stockholm 2009

Why Continuous Improvement?Turquoise = improvement per changePurple = accumulated improvement

Page 10: eMetrics Stockholm 2009

Types of Testing

Page 11: eMetrics Stockholm 2009

How Many Things to Test at Once

The time it will take to complete a test is determined by:• Traffic• % of traffic included in test• # of combinations to test• Conversion rate• Size of improvement

Page 12: eMetrics Stockholm 2009

How Many Things to Test at Once

Page 13: eMetrics Stockholm 2009

Two Types of Multivariate Tests

• Full factorial – every possible combination is tested (slow – thorough)

• Fractional factorial – fewer factors are tested (fast – you may miss something, but likely it'll be unimportant correlations)

Page 14: eMetrics Stockholm 2009

Finding What to Test• Web analytics (bounce rate, a page’s contribution to

revenue, cost, etc.)• Usability• Eye-tracking• Surveys/feedback• Demand (external/on-site search trend analysis)• Heat maps• Customer support input

Page 15: eMetrics Stockholm 2009

Landing PageEvaluation

Page 16: eMetrics Stockholm 2009

CheckpointsArea - Question – Idea• Relevant -- have I come to the right place? (echo keywords from the sponsored

link or copy/design from the banner)• Personal -- is the page speaking to me? (use the four temperaments)• Descriptive -- what is it that I'm getting? (list features)• Beneficial -- what's in it for me? (list benefits)• Credible -- can I trust this website? (add company facts, provide contact

information, and display reviews)• Intuitive -- how do I proceed? (check how you can help the visitor easily carry

out the desired action, check call to action)• Focused -- are there any hurdles/distractions on my journey? (remove

distractions such as unnecessary links)• Time-critical -- do I have to buy now or can I wait? (add a sense of

urgency -- limit the time a deal is valid)

Page 17: eMetrics Stockholm 2009

Landing page for paid search term “buy playstation 3”

Relevant: No, I seem to get to a landing page for video games, not for gaming consoles. And the games listed aren’t even for Playstation 3!

Personal: I am humanistic according to the four temperaments. I see no reviews or testimonials. Irrelevant for spontaneous, competitive and methodical too.

Descriptive: No list of features. Well, duh! The page is irrelevant.

Beneficial: Irrelevant page.

Credible: Can’t see anything about how secure payments are, reviews etc. More crucial for small brands, but still.

Intuitive: No. I resorted to on-site search + scrolling 5 rows.

Focused: No. Plenty of links and content to distract me.

Time-critical: No sense of urgency.

Page 18: eMetrics Stockholm 2009

The page they should be using for paid search term “buy playstation 3”

Relevant: The heading mentions the Playstation 3.

Personal: No information to make competitive feel they’re making the best deal possible, not enough details for methodical, review area for humanistic (but no reviews), no quick checkout for spontaneous.

Descriptive: Pretty descriptive. Possibility to view enlarged image. Problem is that there is a text saying “not exactly as shown.”

Beneficial: The benefits are hidden – no benefits described in copy.

Credible: Can’t see anything about how secure payments are.

Intuitive: Not too bad, the button for adding to cart is above the fold even on a laptop.

Focused: More focused than the other page, but still showing information/links that are not necessary to buy a PS3.

Time-critical: It just says “current price”.No idea if it’s higher or lower than usual, orwhen it will change. Doesn’t feel urgent.

Page 19: eMetrics Stockholm 2009

Hypothesis Development

Examples:• Benefits are not communicated > Describing what’s in it for

the buyer will increase conversions• Doesn’t speak to the four temperaments > Writing copy

around the four temperaments will increase conversions

Page 20: eMetrics Stockholm 2009

Questions?Lars Johansson+46 70 244 89 [email protected]/in/larshjohansson