Christi olson-huzzah-searchand social

Advanced Paid Search Marketing

Christi OlsonWindows Consumer Mktg

Introduction

Today is all about YOU!• Who is here?• What was your first online

marketing campaign?• What do you want to learn

today?

AGENCY OR IN-HOUSE SEM MGMTPaid Search Marketing Track:

Agency or In-House?

• Do you need an agency?• Choosing an agency• Fee structures

Do you need an agency?

• What do you want/need?

• Are you looking for an agency or a tool set?

• Will an agency be able to meet your expectations?

How to Choose an Agency

• Define and prioritize your needs before you review solutions

• Identify a short list of vendors that meet your requirements

• Check references– Don’t just check the names that they give, but use

your network to find others that you know who are also using their services

• Ask for a low priced trial or a test period• Negotiate on price

Key RequirementsCreate a matrix of your key requirements for an agency and evaluate the candidates:

PricingSimilarly, look at the pricing of the agencies you are evaluating and determine if the costs are inline with your expectations.

Agency Pricing – Example 2Understanding Tiered Fee Structure… and determining the break even points

Some agencies negotiate a tiered fee structure. Make sure you get in writing how the agency calculates the fee’s based off of the tiered structure.

Make sure you evaluate the pricing structure to find the breakeven point for minimum monthly fee’s.

Agency Pricing – Example 2Understanding Tiered Fee Structure… and determining the break even points

Some agencies negotiate a tiered fee structure. Make sure you get in writing how the agency calculates the fee’s based off of the tiered structure.

Agency or In-House Key Takeaways• Use an agency if you do not have anyone in

house with online marketing expertise, or you do not want to spend $ and time training someone.

• There are good agencies at all price levels– However, you usually give up 1:1 time account

management w/lower prices• Identify and prioritize your needs first• Meet with 2-5 agencies that meet those needs• Choose based on functionality & price– If you are unsure about an agency ask for a 3 month

trial period.

TESTING METHODOLOGIESPaid Search Marketing Track:

Testing Methodologies

• A/B vs. Multivariate Testing

• Planning• Getting Started• Monitoring & Reporting• What’s next?

Difference between A/B & Multivariate

• A/B Testing: – Simply comparing one object against another

• Multivariate Testing:– Testing multiple variables simultaneously

Difference between A/B & Multivariate

• A/B Testing: – Simply comparing one object against another

• Multivariate Testing:– Testing multiple variables simultaneously

A/B Tests

Best for:• Low to High volume traffic• Test the layout and design of a

webpage

Multivariate Tests

Best for:• High volume traffic (1,000+ clicks/page

views per week)• More complex to implement• Layout and design stays the same,

testing content• Change specific parts of the site

(headlines, images, copy, logos, footers, etc)

A/B & Multivariate Tests

6 Simple Steps for Planning

• Choose your test page• Choose the conversion page• Choose the type of test (Multivariate or A/B)• Select & Create Content for Experiment Pages• Add tracking tags to Experiment Pages• Q/A your Experiment PageGO!

What should you test?

• Determine what is the goal of the page?– What end action do you want users to take?

– Is your end action clearly represented on your page?

– What sections of the page significantly impact a users ability to take that action?

What should you test?

• Determine what is the goal of the page?– What end action do you want users to take?

– Is your end action clearly represented on your page?

– What sections of the page significantly impact a users ability to take that action?

What should you test?

• Determine what is the goal of the page?– What end action do you want users to take?

– Is your end action clearly represented on your page?

– What sections of the page significantly impact a users ability to take that action?

Windows Landing Page Optimization Test

Control Layout

Case Study:

• Goals:– Understand if the layout and

design of page could be altered to send more traffic down the conversion funnel.

– Increase on-page click-thru to 50%

Created a custom Search Landing Page to drive users down one of three possible conversion paths: Purchase, Research, & Upgrade.

Windows Landing Page Optimization TestControl Layout Experimental Layout:

Results:

Case Study:

New Layout showed 89% lift in on-page click-thru

Windows Landing Page Optimization Test

Control Layout Experimental Layout:

Results:

Next we tested if the location of the pictures influced on page click-thru:

YES! Click-thru saw a 49% lift with new design

Measurement & Reporting

• Don’t get lost in the numbers!– Know what you need to look at, ignore the rest

• Build your own conversion funnel analysis to understand where you are losing people – and how to keep them on the path to conversion

• Create a forecast for your campaigns– Defining quantitative goals in advance give you a

measure for success.– Change and redefine your goals as your tests

become more sophisticated

Testing Take-aways

• Develop a testing schedule– Look at least 3 months out– Know what to look at, ignore the rest

• The closer to conversion that you can improve the landing page, the more impact it will have on bottom line– Landing page tests are critical

• Be sure your % breakdowns between the control & experiment pages provide a large enough sample size and confirm your results are statistically significant.

Huzzah! Search & Social Workshop

Thanks for Attending!Any Final Questions?

Christi OlsonHuzzah & WOOT Trainer

Senior Search Manager, Microsoft509.869.7545 | ChristiOlson@live.com

APPENDIX

A/B Ad Copy Tests

Titles

C. Get a PC designed for Work & PlayD. Computers for Work & Play

Desc. Line 1

A. Choosing a Computer?B. Get your Perfect PC

E. Have a good time, anytime, anywhereF. PCs with Windows 7 do it all!

Desc. Line 2

G. Microsoft.com/Why-Choose-a-PCH. Microsoft.com/PC-vs-Mac

Display URL

Get a PC designed for Work & PlayChoosing a Computer?

Have a good time, anytime, anywhereMicrosoft.com/Why-Choose-a-PC

Get a PC designed for Work & PlayChoosing a Computer?

Have a good time, anytime, anywhereMicrosoft.com/PC-vs-Mac

Get a PC designed for Work & PlayGet your Perfect PC

Have a good time, anytime, anywhereMicrosoft.com/Why-Choose-a-PC

Get a PC designed for Work & PlayGet your Perfect PC

Have a good time, anytime, anywhereMicrosoft.com/PC-vs-Mac

ACEG

ACEH

BCEG

BCEH

A/B Ad Copy Tests

Display URL

Description Line 1Description Line 2

Title

Titles

C. Get a PC designed for Work & Play

D. Computers for Work & PlayDesc. Line 1

A. Choosing a Computer?

B. Get your Perfect PC

E. Have a good time, anytime, anywhere

F. PCs with Windows 7 do it all!Desc. Line 2

G. Microsoft.com/Why-Choose-a-PC

H. Microsoft.com/PC-vs-MacDisplay URL

ACEG

ACEH

ADEG

ADEH

ACFG

ACFH

ADFG

ADFH

BDFG

BCEG

BCEH

BDEG

BDEH

BCFG

BCFH

BDFH

AECG

AECH

AEDG

AEDH

AFCG

AFCH

AFDG

AFDH

BECG

BECH

BEDG

BEDH

BFCG

BCEH

BFDG

BFDH

How many possible combinations can you be testing?

8

1 6 3 2

2 4

If you add in Destination URL: 64 combinations