Avinash Kaushik is an analytics rockstar. As weird as that may sound, his enthusiasm pulses through his analytics blog which makes you want to be a better data analyst – no easy feat.
Shane Johnson, the president of Triangle AMA steps up to get things started. After some intro talk, he introduces Jim Hazen from Capstrat to introduce Avinash.
Avinash starts out talking about the company he started with his friend, MarketMotive. He has way too much energy for 9am. He is 100% engaged right out of the gate.
He thinks back to the good old days when nobody looked at data and any decisions could be held up by “gut feeling”. Now he has more data than God wants him to have.
Just looking at user clickstream is crazy, there’s just too much data. That’s why people who do clickstream analysis die young. Everyone on the internet has ADD, and they’re all coming to our website.
If all you’re measuring is hits (how idiots measure success) you’re missing out. You need to measure what’s important. Things like bounce rate (aka I came, I puked, I left).
Also look at the top 20 homepages on your site. What about those pathetic pages that couldn’t even get 1 entry – kill them, and be glad about it.
He jumps into the SAS site where he points out that there a million places to download whitepapers – why? It’s text, why isn’t it on your site as text, why are your forcing me to download a white paper? It’s like it’s 1962 and you’re going to mail me information. It’s not just SAS though, all the big companies do this.
Question this type of behavior. “This sucks! This cabal of crap needs to end!”
Nobody likes a selfish lover.
Are you looking at your website? Avinash pulls up the old Expedia site where 1 small widget does all the work, and 80% of the site is just pimping other services. He talked with them and said that as a visitor, he needs to be satisfied first!
- Stink less
- Segment or die
- All data in aggregate is crap
- Love your loyalists
- Find the relationships between website purpose and customer intent
- Look at revenue, who is buying more than normal, and why?
- What’s important? Is your new campaign working?
How do you focus?
Intelligence. If we could get rid of the human factor, all of this would get so much easier
To get rid of humans, tools Google Analytics are just glorified data pukers. This drove Avinash crazy. He challenged the GA team to come up with a algorithm to bring out the most useful data – enter weighted data (aka magic button). By eliminating the stupidness from the equation, you can focus on the things that will give you the best possible ROI.
Try running a tag cloud for a graphical interpretation of what your site is delivering. How big should your brand be? How about main keyword?
He uses the Mormon church as a magnificent example of great search cloud. All the words they want associated with them are present, those they don’t want are nowhere to be found.
It’s very important to move beyond the top ten rows of data!
Try comparing content produced to content consumed. Fix where needed.
It’s important to have multiple goals for your website. Track both micro and macro conversions. Don’t forget to measure economic value. There are lots of other things happening that contribute.
Find out what channels drive what conversions. Even though channel X may not contribute to macro conversion, it opens up the funnel so that more people can find the macro conversion point.
Shout marketing. We yell our brains out and hope people listen – it scales when we can make more people listen. The problem is the old-school shout marketing world has passed.
What metrics are important? How about retweets as a way of showing value or the follower retweet %. Tweet things of value to your followers. Work on that!
On twitter only 10% of your followers see each tweet, you read that right, only 10%.
Listen, truly listen
How are you listening to your customers/followers/fans?
You can find out almost everything about your visitors by finding out the answers to three questions:
- Primary purpose: why are you here?
- Task Completion: could you complete your task?
- If you couldn’t complete your task, why not?
Did a Hippo create your site?
HIPPO = Highest paid person’s opinion
To compete with HIPPO’s you need to learn how to be wrong very quickly. Enter A/B testing.
Avinash Kaushik uses the Obama 2008 campaign site as an example of great A/B testing. From baseline to best was a 40% increase in conversion.
Don’t sacrifice functionality for sexy.
Don’t ask, “what channel gets credit for the sale?” – ask, “if I only had $100,000, how would I allocate it?” Figure out the right media mix. It’s all about experimentation. Experiment or die.
Q & A
I can’t summarize all the great questions, but here’s a couple.
If I’m not a full-time analytics analyst, how much time should I spend on data analysis?
Line of site model. Look at; Price Strategies. Cost Strategies. Market Share Strategies. Market Size Strategies. Then figure out the KPI for those strategies. Email him for the cool presentation handout.
Norm Cloutier from McClatchy challenged one hypothesis that scrolling is bad because his test shows that 70% of visitors scroll down. Avinash says, maybe they scroll because your site sucks. He admits that he could be 100% wrong. The only way to know for sure is to experiment.
What tools do you help you analyze Twitter analytics?