Get Free Quote
« Redefining the... | Blog home | Search Around The... »
March 17, 2008

Analytics: Data Into Action

  • Print Friendly

What do you know? Someone actually listened to me and decided NOT to place the token Analytics session on the last day of the conference. Sweet! Perhaps my mind will actually be able to comprehend this one. Maybe.

[Or maybe not. My mind is clearly not working well as I just ripped half my finger off on a pin sticking out of my laptop bag. The skin is actually still on the pin. Gross. And burning.]

This morning we have the famed Kevin Ryan moderating with Matt Bailey (SiteLogic) and Steve Keller (Assurant Health) presenting. It’s good to see Matt Bailey again. I feel like I haven’t seen him in forever, though it’s probably only been a few months. He’s looking very festive with this St. Patrick’s Day-themed tie. Or maybe it’s just green. I should stop staring at him before they ask me to leave.

Kevin Ryan gets things started and says next time he’ll get a bigger room. Hee, seriously. And it’s not that the room is even small, it’s just there are about a gazillion people trying to get in. The SES conference series is alive and well, people.

Kevin also lets us know that the flu shot this year is worthless and being high on NyQuil isn’t all it’s cracked up to be. Also from Kevin: Stay away from the Irish pub places tonight. Korean BBQs rock, so do Thai places. There are 364 other days to go to an Irish pub, St. Patrick’s Day in NYC is not the day to do it. Hee, drugged Kevin is pretty funny.

Matt Bailey is up first. He asks how many people are baffled by analytics. I raise my hand.

Analysis or Reporting? When you’re looking at the same numbers every time and reporting up the chain, that’s not analysis, it’s reporting. There’s no emphasis on what numbers are important. You’re giving grade cards and nothing more. The rest of your existence is now based upon why that number is higher or lower. Analysis leads you to conclusions that you can implement on your Web site to see positive or negative results. Analytics is about going beyond reporting

The first step to do this is to have clearly defined goals. You should not bring in an analyst and ask them what you should be tracking. You have to set the goals for your site. What do you want people to do? Why do you even have a site? If you don’t have a goal established, you don’t know if you’re reaching it.

Focus on KPIs like sales, leads, contact forms, downloads, page views, etc. They’ll help you establish your goals. Things like page views, hits, monthly visitors, etc, tell you NOTHING about your Web site. He shows a picture of a rat’s ass and says that’s how much they matter. Heh, the panelists are snarky today.

When it comes to analytics, there is no accuracy. It’s just not a principle of analytics. It’s so hard to target. Every analytics program will measure things differently. Trends are much more important.

The best thing you can do to establish trends is segmentation: Break customers into buckets by goals and motivations. Segmenting can help you get intelligence on what they were looking for. You can’t treat all of your customers the same. Visitors come to your site looking for vastly different things and you need to classify them based on what they were looking for.

What’s segmentation? He uses Star Trek as an example. Oh Jesus…

The Starship enterprise had a crew of 430 people. He knows this because startrek.org told him. There were 59 total deaths in the 5 year mission. That’s a 13.7 percent mortality rate (aka conversion rate). Of the 54 deaths, yellow shirts made up 10 percent, blue shirts made up 7.2 percent and red shirts made up 72.8 percent. That’s our data but we still have no action. We have knowledge, but no indicator on how we can improve it or make it worse.

Factors that lead to a red shirt death: If you beamed down with Captain Kirk and wore a red shirt, you died 57.5 percent of the time. This is the number one factor that leads to the death of a red shirt. OMG the giggles.

To increase the survival rate: You’ll see that if Captain Kirk meets an alien woman, the red shirt survival rate increases to 84 percent.

How often do these factors occur? Capt Kirk has a conquest rate of 30 percent. If you go to a land of [insert name of things that fight Captain Kirk. I’m not geeky enough for this.] and you’re a red shirt, you’ll probably die. If you go to a land of peaceful women, you’ll live 30 percent more of the time. That’s segmentation!

One you have segmentation complete, you have to build a context. Build a story. Where did users come from? What did the link say that they click on to get to you? Where was it located on the page? Then compare and contrast each segment. You’ll find out where you’re doing well and where you’re not doing well.

Analytics programs are software that interprets data. It’s not going to tell you what to do. That’s your job.

Analytics give you data. You have to add information and context to it. Once you start adding more contexts, you understand a little bit more of what’s happening on the Web site. The more data points you add, the clearer picture you are painting of what someone is dong on your Web site. Find out what visitors search for, who entered at what page, who stayed on the site for a certain period of time, who converted. This gives you knowledge, but still doesn’t tell you how to implement it on the site. Ultimately, this comes up to you. Analytics is more about you as the analyst than the package. You should spent 10 percent of your time on the product and 90 percent of the time interpreting the data. That’s what gives you the wisdom, not the tool.

The more you optimize your Web site, the less likely it is people are going to enter your site on your home page. This is good. You don’t want people to enter in on your home page. Graphic artists, no one wants to see your Flash. Flash has its place – on the Coke site. Hee!

Key Performance Indictors

  • Time on site
  • Pages Viewed
  • Conversions
  • Goals

By Segment

  • Blogs
  • Web sites
  • In-market links
  • Social News
  • Search

Don’t compare yourself to your competition. Compare your site to yourself. Compare your visitors to your visitors. Comparing yourself to your competition is only going to help you run your site like theirs – and they might be failing.

You’ve got to build context. You do that by getting out of the geek speak (says the man who just used a 3 min Star Trek example…). People don’t know what "site engagement" is. You have to give your CEO a story. Don’t change the data, change the labels. A visitor who looks at less than one page is an "abandoner", three pages or less is a "flirter", etc. Give them information they can use to change the trend. People respond when you build enough context to tell the story.

One you do your analysis, do something with it. Otherwise you’re just burning your money. Forrester research tells us companies who brought in a full-time analyst had an ROI of 900 and 1,200 percent. That’s the power of Web analytics. If you’re sitting on it, you’re not going to get anything out of it.

Matt Bailey is awesome.

Next up is Steve Keller.

He lets us watch a movie. He says he’s testing the sound systems. I’m confused, but yey movies! Now he’s quizzing the audience and offering up hats as a reward. Luckily, Tamar Weinberg isn’t here or she would have stormed the stage to get one.

[Steve makes fun of Google and then the computer goes haywire. Heh, don’t make fun of The Goog. All your PowerPoint belong to us.]

He’s going to take about his company a bit and use them as a case study.

SEM Objective: Dominate the Page. Steve shows that his company shows up twice in the organic space and once in the paid. He’s going to be focusing on the paid.

Campaign History

  • Paid Search Campaign
    • 2002 launched paid search for Short Term Medical Insurance (STM)
    • 2005 launched paid search for Individual Medical Insurance (IM)
  • Media Buying initially based upon CPA methodology – "allowables"
    • CPA allowable for STM = $100
    • CPA allowed for IM=$200

They used to work with iProspect. They set a goal where iSEBA (an iProspect tool) would manage their bids against the target allowable. The engines would drive customers to their Web site. The problem was iSEBA didn’t have enough information. They weren’t feeding it enough information.

Problems:

  • PPC optimization strategy optimized only on "primary" product line.
  • PPC methodology considered all sales within a product line to be of equal value

Solutions:

  • Implement the ROAS model
  • Two elements: Determine and capture an estimated customer value. (This is the amount you’re willing to page for that sale.) Set an optimization target.

Implementing the ROAS model allowed them to provide iSEBA with real-time, safe-level feedback, which improved big management.

Reporting change: Formerly, their primary metric for effective was CPA. They’ve migrated to ROAS as the primary effectiveness metric.

Conversion Management

Keyword Selection: Seasonally Active Keywords. In summer, an increase in insurance coverage for graduating college students leads to conversions in impressions, clicks and conversion rates.

Ad Copy: Mine data across search campaigns. Understand differences in CTR and Conversions Associated with different ad copy.

Conclusions:

  • Creatively applying existing data may improve your results
  • Avoid the silo – integrate search with your marketing plan
  • Paid search marketing – though data driven – remains a combination of Art and Science.
  • Print Friendly




Comments are closed.



Learn SEO
Content Marketing Book
Free Executives Guide To SEO
By continuing to use the site, you agree to the use of cookies. AcceptDo Not Accept
css.php