1. Landing Page Testing to Attract Super Affiliates Tim Ash, President, SiteTuners.com

2. Overview

The mind of the super affiliate

Landing page testing basics

Case studies

Picking a testing method

Pitfalls to avoid

Q&A

3. The Mind of theSuper Affiliate 4. Super Affiliates are

Mercenary

• Work for themselves, always looking for best opportunity

Busy

• Always testing new programs, will quickly cut their losses

Greedy

• Bombarded with offers, picky about testing new programs, expect better than rack rate deals

Lazy

• Want to minimize up-front and ongoing work, do not want changes or surprises

5. Top-10 Ways to Attract Super Affiliates

#10 Join an affiliate network

#9 Offer better payouts

#8 Pay up

#7 Allow bidding on your brand

#6 Dont be high maintenance

#5 Dont change the rules of the game

#4 Share sensitive information

#3 Listen to suggestions

#2 Go the extra mile

#1 Fix your conversion rate

6. Impact of Improved Conversion Rate

BEFORE

Payout - $1.25 Cost - $1.00 Profit - $0.25

A 20% increase in conversion rate

AFTER

Payout - $1.50 ($1.25 x 120%) Cost - $1.00

Profit - $0.50

will DOUBLE the super affiliates profits!

7. Landing Page Testing Basics 8. What is Landing Page Testing?

Online Marketing Activities

• Acquisition Get people to your site

• Conversion Persuade them to take desired action

• Retention Increase lifetime value of relationship

Conversion Rate

• Percentage of visitors who take desired action

Landing Page Testing

• Improving conversion rate by testing website changes

9. Why should you care?

You have neglected your landing pages

Your conversion rate is too low

This is driving away super affiliates

10. Question: What is a Camel ? Answer: A Horse Designed by Committee 11. Who should design your site? I.T. Marketing Ad Agency YourBoss Webmaster 12. None of the above Your website serves too many masters. It should be focused only on conversion. I.T. Marketing Ad Agency YourBoss Webmaster YOUR WEBSITE VISITORS 13. Case Studies 14. Case Study - RealAge.com

Conversion action: Completion of free RealAge Test registration

Test Size: 552,960 unique recipes

15. Before After 16. Results - RealAge

40% lift in conversion rate

$3,285,000 annual profit improvement

17. Case Study Uniblue Registry Booster

Conversion Action: Free software download

Extensive previous page testing by client

Test Size: 3,110,400 unique recipes

18. Before After 19. Results - Uniblue Registry Booster

24% increase in conversion

CR improved from 30.3% to 37.7%

$182,000 annual profit improvement

20. Case Study - PowerOptions

Conversion action: Free 2-week trial sign-up

Tuning Method: A-B Split Testing

Test Size: 3 unique recipes

21. Before After 22. Results - PowerOptions

75% increase in post-trial sales

$200,000 annual profit improvement

23. Case Study EngineReady

Industry: SEM Agencys Debt Negotiation client

Conversion Action: Online form or phone lead

Large-scale PPC campaign

Test Size: 192 unique recipes

24. Before After 25. Results - EngineReady

51% increase in revenue per visitor

$48,000,000 annual revenue increase

26. Picking A Tuning Method 27. How do you tune?

Key Considerations:

Size of Test(total number of unique recipes)

Need to consider variable interactions

28. Example Size of Test Calculation

Page Headerx 3

Navigation Bar x 2

Headline x 5

Call to Action x 4

Photo x 4

Sales Copy x 4

Endorsements x 2

Question Labels x 2

Question Delimiter x 3

Question Explanations x 2

Button Text x 3

Button Format x 4

12 Variables , 38 different values

552,960 unique recipes

29. Ferraris are Really Fast Positive interaction between headline and picture Interaction Example(1of 4) 30. Ferraris are Really Fast Negative interaction between headline and picture Interaction Example(2 of 4) 31. Volvos Are Really Safe No interaction between headline and picture Interaction Example(3 of 4) 32. Volvos Are Really Safe Positive interaction between headline and picture Interaction Example(4 of 4) 33. Interactions are Very Important

Best setting for variable depends on its context

Interactions exist & can be very strong

Ignoring interactions will lead to suboptimal results

A/B Splits & Parametric Multivariate testingassume that there are no interactions

34. A-B Split Testing

Test one variable at a time(with 2 or more values)

Send equal traffic to all versions

Very simple to implement & track

Minimum Data Rate:

10 conversions / day

Typical Test Size:

1-10 recipes

35. Parametric Multivariate Testing

A.K.A. Design of Experiments & Taguchi Method

Tests several variables at the same time

Ignores variable interactions

Minimum Data Rate:

50 conversions / day

Typical Test Size:

10-100 recipes

36. Non-parametric Tuning

Proprietary math for Internet marketing

Designed for large-scale tests

Takes variable interactions into account

Minimum Data Rate:

100 conversions / day

Typical Test Size:

1,000,000+ recipes

37. Avoiding The Pitfalls 38. Awareness Squandered 146 clickable links! 39. Focus on Visitors Goals 40. Pitfall #1 Squandering Attention

Stop screaming at your visitors

Eliminate choices

Unclutter what remains

41. Frustrating User Experience 42. Guide people safely to their goal

Context sensitive search wizard

• Text

• Category

• Brand

• Size

• Color

• On Sale & New

43. Pitfall #2 Frustrating Users

Dont make people feel stupid

Avoid painting them into corners

44. Invisible Risk Reducers 45. Prominent Risk Reducers 46. Pitfall #3 Invisible Risk Reducers

Make risk reducers prominent

Show them on all pages

47. Prominent Client List 48. Prominent Media Coverage 49. Pitfall #4 Lack of Social Proof

Transfer trust from larger brands

• Endorsements

• Affiliations

• Media coverage

Transfer trust from your clients

• Case studies

• Client logos

• Testimonials

50. Transaction Interruption 51. Poor Transaction Design(1 of 2) 52. Poor Transaction Design(2 of 2) 53. Better Transaction Design * *However best solution is toremove registration requirement altogether. 54. Pitfall #5 Surprising or Confusing Visitors

No one likes surprises online

Dont interrupt people unnecessarily

Clearly identify important roles

Test any interruption tactics to determine impact

Consider long-term implications

55. Pop Quiz

You test a change that increases CR

Week 1 -CR = 4.63%

Week 2 - CR = 5.03%

Should you get a promotion?

56. Answer

No.

NewBaseline Lift

Week 1 -4.63% 3.59% +29%

Week 2 - 5.03% 6.18% -19%

Things have actually gotten much worse

relative to the baseline.

57. Pitfall #6 Ignoring Your Baseline

Always devote some bandwidth to your current version (the baseline)

Measurerelativeto the baseline, notabsoluteperformance

58. Pop Quiz

You are performing an A/B Split Test

Option A - 90 conversions

Option B - 100 conversions

Is B really 10% better than A?

59. Answer

Dont know.

1/3 of the time

this difference will bedue to chance

and A is actually better than B.

60. Data Collection Example(1 of 2) 10% difference - 90 vs. 100 conversions Very wide and overlapping error bars 61. Data Collection Example(2 of 2) 10% difference - 1350 vs. 1500 conversions Error bars have gotten narrow and have uncrossed 62. Pitfall #7 Not Collecting Enough Data

Do not make decisions based on too little data

Pick a confidence level(90%, 95%, 99%)

Wait to see which version is better

63. Summary

The mind of the super affiliate

Landing page testing basics

Case studies

Picking a testing method

Pitfalls to avoid

Q&A

64. Q&A & Additional Resources

How to Attract Super Affiliates Whitepaperhttp:// SiteTuners.com/downloads.html

Landing Page Optimization Book http://LandingPageOptimizationBook.com