Multivariate Testing - Big Results from Small Changes

Multivariate testing can drive big results from small changes.  In a recent Google Blog post, examples of Google’s search experiments are shown and some are surprisingly small.  For example, take a look at the two pages below:

Multivariate Testing - Big Results from Small Changes

Multivariate Testing - Big Results from Small Changes

Can you tell the difference between the two?  I couldn’t.  Turns out, the difference lies in the amount of white space in the first search result, making the first result in the Picture 2 a bit more visible.  This test is designed to determine if these types of changes help you search faster and if it makes results more prominent.

Here’s another example of a recent Google experiment:

Multivariate Testing - Big Results from Small Changes

Multivariate Testing - Big Results from Small Changes

If you can spot the difference in these two pictures, you’re visually astute.  The difference is the the thickness of the plus box next to the stock quote.  It’s difficult to determine if one is “better”, because results of the tests have both positive and negative effects.  Searchers might click on the the thicker box, causing them to possible ignore other helpful results.

So, what?  Subtle differences can make a big impact on results.  Multivariate testing helps marketers to test even the most minute details to determine which has the greater impact.  Without the use of multivariate testing, the findings could take lots of time, man hours or just never be determined.

At WebsiteBiz, we use Omniture Test & Target as well as Google Optimizer.  We have expertise using these tools and are leveraging technology in our partnerships.

Multivariate Testing - Big Results from Small Changes

Google Adwords Seminars Coming to Charlotte, NC

Google Adwords Seminars Coming to Charlotte, NCSeminars are currently offered in two areas: AdWords and Analytics. AdWords Seminars are designed to teach you how to manage your AdWords account more effectively by covering topics from best practices on account setup to bidding and campaign management strategies. Analytics Seminars will help you better leverage the insights found in your Analytics report by covering topics such as setup tips, best practices, and analysis techniques. Seminar leaders will use in-depth course materials, hands-on exercises and valuable tips from their years of managing AdWords and Analytics accounts to teach you how to achieve your goals.

September 8, 2008 (Monday) AdWords 101: Charlotte, NC
September 9, 2008 (Tuesday) AdWords 201: Charlotte, NC
September 8 & September 9 Two Day Seminar Package

Google Adwords / Charlotte NC / Online Marketing Seminar Registration

Google Adwords Seminars Coming to Charlotte, NC

SES - Black Hat, White Hat: Playing Dirty with SEO

Some say that “black hat” search marketers will do anything to gain a top ranking and others argue that even “white hat” marketers who embrace ethical search engine optimization practices are ultimately trying to game the search ranking system. Are white hats being naive? Are black hats failing to see the long-term picture? This session will include an exploration of the latest black and white issues, with lots of time for dialog and discussion.

Moderator:
Speakers:
Notes:
  • No Presentations will be given. It is a packed session here at SES. Quite a rowdy crowd. The discussion will be town hall format.
  • Definition of Black Hat -
    • Bruce Clay: The search engines define what it is. Most white hats play in the middle; gray hats play on the edge; black hats focus out of bounds
    • Jill Whalen: Attempting to deceive the search engines
  • Definition of White Hat -
    • Todd Friesen: He agrees with Jill’s assessment but brings up the verticals differences
    • Greg Boser: White hat is defined as SEO with no game
    • Dave Naylor: Black Hat is like driving a Porsche in competitive industries
  • Is there risk in both white and black hat techniques?
    • Jill says work hard to make a great website
    • Bruce says if you are working on your site and Matt Cutts comes up behind you, you close your notebook- you are at risk. White Hats color within the lines. Since search engines don’t publish the rules some white hat techniques can appear as black hat even though legitimate.
    • Todd agrees.
    • Greg indicates that most of the black hat techniques are server on his personal site not clients.
    • Jill brings up the idea of incompetent SEOs and the fact that they are the worse group. The crowd agrees!
    • The topic has turned to buying links and what that really means to the search engines. Dave says what does the footprint of a specific tactic look like. Black hats are always testing the limits of link building.
    • Bruce agrees if we spend time working on a project we need to know what the rules are. The question becomes do we take the time or do we buy links.
    • Jill chimes in on the rules…don’t read Google’s guidelines she says it’s common sense. We are all adults. Todd disagrees…this is not common sense.
Questions now coming from the floor.
  • Will links in RSS feeds help?
    The panel defines it as really simple spam (LOL) This could be a dangerous technique.
  • Using widgets effectively could be helpful.
  • The owner of SEOChamp.com chimes in about how black hat techniques are coming from overseas. He really is passionate about how there are 70 parts to the Google algorithm. Matt chimes in at over 200.
  • The challenges between a 1 month vs long term plan. Ticket brokers buy links and it is a tough space.
    • Greg says look at the competitive research and then explain what can be dome and what is the risk tolerance.
  • Affiliate marketing does get a bad rap sometimes. Todd is a big fan of making money. Certain verticals have a hard time following the rules to compete.
  • Bruce says you should take time to build credibility over time and that an aggressive link building techniques can be risky.
  • The BMW brand who was penalized for 72 hours shows that the bigger the brand the easier it is to ask for forgiveness. The interesting fact was they got tons of blog links fro the PR. The BMW SEO work was amateur at best.
  • Now Matt Cutts has asked for the mic: His disclaimer is when we remove sites, they are removed based on the facts not the name. Matt indicates that there is a large subsection of a major newspaper that is currently banned and has been banned for some time. Google does not talk about what sites big or small that they take out. Matt says don’t take the risk.
  • If you are banned you can ask for re-inclusion. Greg says look at the entire space when you choose to ban sites.
SES - Black Hat, White Hat: Playing Dirty with SEO

Search Marketing Conversation with Google Engineer Matt Cutts

Tom Dressler and Eric Dudley with Matt Cutts from Google

Tom Dressler and Eric Dudley with Matt Cutts from Google

Eric Dudley and Tom Dressler spend time with Google Engineer Matt Cutts at SES San Jose.

Search Marketing Conversation with Google Engineer Matt Cutts

SES - Social Media Analysis and Tracking

Social Search can be used to drive traffic, conversions, and increase ROI by monitoring conversations happening in cyberspace, often in real time. By using buzz-monitoring tools such as Buzzlogic, search professionals can be very effective in finding influencers within online conversations and reaching out to them. As one example of a few that will be presented, learn how Military.com leverages social media and buzz-monitoring tools, together with a solid understanding of business goals, to increase profitability.

Moderator:
Speakers:
Two top level ideas on how to measure Social Media.
  1. Look at your referrer logs and breakout Conversational Media
  2. Look into data or let Google do it via site trends
Idea: Have CMO’s work closely with Web Analyst
Notes from Rob Key, CEO, Converseon:
  • One of the challenges is what do we do with the data once we have it. Rob indicates that 5 of the top 10 sites online today are Social Media sites. The question is, what are the people talking about and where are the talking about it?
  • He is outlining “How to Design a Social Media Strategy”
    1. Step 1 is to Listen - What is the external tone
    2. Mining the Universe - First Know What You Want To Know
      1. What is the Volume of Conversation?
      2. Who are the Influentials? - You should only be contributing 10% of the time overall
      3. Report on Voices
      4. Report on Relationship Topics
      5. Report on Competitors/Features
      6. Report by Cloud Tags
    3. How to use the data?
      1. An Extension of Customer Service ( Rob sites a BusinessWeek article) & Reputation Mgmt.
      2. Search Results - Engagment Strategy
      3. Listen, Plan, Engage (Graco Case Study is being sited)
    4. Where does this all go?
      1. Trending - We have ___% of positive sentiment and are mapping this conversation trend over time
      2. Social Media Monitoring and Analysis will Flatten
      3. Have a Grand Unified Vision - Sales, Brand Tracking, Conversion Analytics, Traffic & Conversation Monitoring.

Notes from Todd & Breanna who have prepared a case study on Military.com:

  • Site Focus: Connecting Military Population with appropriate communities
  • Goals: Increase awareness and drive traffic
  • Fact: Social Media visitors do convert higher than baseline visitors
  • DoD Buzz
    • Strategy: Isolate the influencers and reach passionate readers of military defense news and information
    • Goal: Increase product awareness to the (new) influencers and their audiences
    • Engaged BuzzLogic
      1. Uncover Conversations utilizing keyword data and break into categories - then cross reference to social media outlets
      2. There was a set of blogs that were very rigorus on topical point
      3. Rank the influencal site based on content
      4. Identify the influencer networks of ad placement - influencer to influencer connection
    • Campaign Overview
      • Create compelling, informative and clear call to action
      • Ran on a cross section of 250 blogs including influencers
      • Results:
        • 86% Higher CTR than historical average
        • Social Media Users Took Action - 5.3% lift in conversion rate overall (RSS feeds and Email Subscriptions)
        • 60% higher new visitors than the site average
        • the SM ads performed an avg. of 6.25% better than the site historical average.
    • Key Observations/Learnings
      • Active conversations about specific topics attract passionate audiences.
      • Social search is different than web search it’s gets you closer to keyword search intent
      • Influencer’s and their network increase ad performance
      • Conversation’s offer a new window on consumer psychology
    • Social Media ia about selecting trust sources and trusting in what they say.
Notes from Edmund:
  • What to track
    • Analyze page views over time - since many forums are optimized for search, engagement is highly visible…forever: SEO continues to offer new visitors over time
    • Track links posted and clicks received by the site (Using match back through Omniture data)
    • Calculation of a Social Media CPM is possible
  • Categorize and analyze discussion topics
SES - Social Media Analysis and Tracking