Actionable PPC Insights From Analytics Data #smx

Rob Jackson, Head of Conversion Analytics, Latitude Group

Conversion Analytics

  • Web Analytics
  • Conversion Attribution
  • Website Usability
  • Landing Page Optimization

Web Analytics

  • $1 billion industry by 2014 – Forrester
  • Growth area in digital marketing
  • Some companies still aren’t getting the basic right

Top Tips – Web Analytics

  1. Get your setup right (e-commerce, goal funnels, campaigns by channel, test conversions, have the right people in place)
  2. Look beyond conversions (some keywords are at the beginning of the purchase funnel, If you are using last click CPA model, look at bounce rates and time on site)
  3. Monitor site performance (scenario analysis and content reports show when pages are under performing; again, look at bounce rate and time on site)

Conversion Attribution

  • Go beyond the last click metric (ways to create rudimentary cookies to track first click)
  • Stop the keyword rot
  • Remove keywords that never convert

Top Tips – Attribution

  1. Report with last click (but analyze other weightings and share these with your clients)
  2. Pull what doesn’t work (keep it simple and avoid over analysis)

Usability Software

  • “Real-time customer research”
  • You know where – now see why
  • Link & form analytics
  • Heat mapping
  • Site issue resolution

Top Tips – Usability Software

  1. Be proactive (the remit of a PPC manager is now beyond the click, look at what your users are telling you)
  2. Segment if possible (use tools like Click Tale, Tea leaf)

Top Tips – Landing Page Optimization

  1. Use Data not Intuition (Take opinion out of optimization by using analytics and usability)
  2. Star over and repeat… (Continuous improvement yields better results than one off tests)

Allisa Ruehl, Manager of Website Effectiveness Consulting, Apogee Search

Favorite Things

  • Advanced Segments
    – Default segments
    – Custom segments
  • Looking deeper into GWO tests
  • Custom reporting

Advanced Segments

  • Like filters, but better!
  • Easy to create
  • Apply to past data
  • Some are conveniently created for you
    – Paid traffic
    – Search traffic
    – New visits
    – Returning visits

Custom Advanced Segments

  • Drag and drop functions
  • Filter and segment by things like:
    – Hour of the day
    – Language
    – City
    – E-Commerce
    – Traffic Sources

Analyzing time periods – making dayparting decisions!

  • Create a segment for traffic that arrives during a time you’re worried about
    – Time of day is less than 6 OR time of day is great than 22
    – If you currently have enough PPC traffic during that time, limit segment to PPC (pay attention to sample size)
  • Compare conversion rate of that segment, to your overall conversion rate for PPC or search

A few other uses

  • Turning any report into a PPC-only report
  • Running specialized reports in seconds
  • Aggregating engagement and conversion metrics
  • Studying geotargeting decisions
  • analyzing GWO tests

Looking deeper into a GWO Test

  • Care about more than just conversions?
    – Order value by test page?
    – Bounce rate for landing pages?
    – Other engagement metrics?
  • For A/B test only
    – Create segment for each page version
    – Run the desired report with each segment (e.g. ecommmerce reporting)

Segment Tips

  • Mind your “or”s and “and”s
  • Look at segment size (test segment size)
  • You can compare up to 3 segments at once
  • Don’t tell you’re boss/CEO/client how easy it was 😉

Tracking Multiple Attributions

  • By default, GA gives you last click attribution
  • You can see first click attribution (with a little extra java coding)
  • Stay away from “no override”
  • Set user defined value
    – Decide what to track (search engine_keyword_PPC/SEO
    – Add setVar java below GA code
    – Create script that checks for cookies

Is this the best way?

  • No
    – you only have 1 free value
    – it’s difficult to see entire history
    – you have to do manual work
  • Yes
    – maybe if Google hears enough from us, they’ll add it

Andrew Goodman, Founder & Principal, Page Zero Media

Can we really simplify this incredibly complex universe?

  • Yes, and we must

Google AdWords Conversion Optimizer

Non-obvious takeaways

  • If you don’t love marketing, you’ll “manage” to mediocrity
  • How patient are you? Check the right date range.
  • Is granularity a great way to go crazy?

Other takeaways

  • Use Google Advanced Segments to make better choices, especially when you don’t have revenue information

Wister Walcott, Co-Founder & VP of Products, Marin Software

Paid Search – Analytics Myths

  1. Google will pick the best creative for you.
  2. Look at all prior clicks for accurate bidding.
  3. “Portfolio Management” is really really hard.

Myth #1: Google will pick the best creative for you.

  • Periodically evaluate conversions/revenue per 1000 impressions
  • Compare each creative in an ad group vs. the average (below average = loser, near average = draw)

Myth #2: Need to look all prior clicks for accuracy.

  • Evaluated “path to purchase” for well-known brand
  • Goal: ensure proper credit to product keywords
  • Only 26% of paid search conversions had more than one click

Myth #3: “Portfolio Management” is really really hard.

  • Use something like Google Conversion Optimizer and this is not as hard as you think

Questions

  • Omniture? Used for large campaigns, and worked well
  • Pay-Per-Response is much easier.

    To figure a PPR Response Rate (RR):

    Responses (outbound) / Inquiries (inbound) = Response Rate (RR)

    R / I = RR

    25 / 100 = 25%

  • Interesting comments, thanks. In addition to clickstream and SEO analysis, I’d like to see better emphasis on analysis of visitor behavior *after* they arrive at a site – how and why they do what they do. In my experience, it’s understanding user behavior on a website (and so improving the number of successful conversions) which is the key driver to measurable success for that site – in particular incremental revenue.

    SEM Sensei.