IBM Watson APIs is about how the future comes to everyone. Can IBM Watson save our predictive analytics future? Maybe

Deep Dive Moon-Audio Web Analytics
Spending the day (and probably most of the weekend) deep diving #googleanlaytics  listening for the music of http://www.Moon-Audio.com. Here are a few of the questions I like to ask every quarter:

* YTD vs. LYTD sales by product & brand. 
* Site heuristics (time on, pages viewed, bounce YTD vs. LYTD. 
* Top 10 Internal searches (compared to last time we looked). 
* Top content. 
* Referring sources. 
* New vs. Returning YTD vs. LYTD. 
* Growth in social. 
* Most linked, shared & reviewed content (will be their Dragon Audio Cable so check against last bench marks). 
* Review the funnel (top entry / exit, Recency, Frequency, Monetary RFM values). 
* 80:20 rule in product sales, brands and content.  

 Read more about analytics deep dives on G+: https://plus.google.com/+MartinWSmith/posts/EHq6vvSJhxf

Asking the right questions in the right way is key to online marketing success. How SMBs can compete with Amazon isn't as important as what is their why and how are they learning from Amazon's web marketing power.

See your Google organic visit data overlaid with Google's major algorithm updates, and then dig into your data.

Marty Note
One of our favorite Free Tools is Panguin. Panguin overlays Google algorithm changes on your Google Analytics taking the mystery out of what has happened to all of your traffic.

We've seen well established sites take a 30% traffic haircut due to all of Google's changes. Hope your site isn't in that group, but if it is know that working your way back is possible, but change is necessary (it doesn't get better on its own).

Visualizing Data
Inside every big data stack is a secret, a secret capable of helping your web marketing IF you find it. Our new Curagami Ecommerce Master Class discusses how to cut a mound of mud into meaninful patterns.

Steps are easy:

1. Start with clean data.

2. Start categorizing.
3. Sort by your categories.

4. Cut your detail out (remember to replace functions with numbers).
5. Graph your broad strokes.

We fly through how to create patterns in this 3 min video. If  you ae stuck or confused email martin(at)Curagami.com.

Good luck and make sure your weed whacker is working.

M

When Everything Is Equally Meaningless Anything Works Equally Well
Had a great conversation on GPlus today about the nature of web analytics. Here is the controversial point I made -

All web metrics are equally meaningless SO anything works. 

If you haven't read Taleb's Black Swan you should. Nicholas Taleb is a brilliant QUANT who is fascinated by the strange predictably irrational ways of man. Taleb is a QUANT used by hedge fund types to see into the float and make more money than you and I can imagine together (lol). 

Taleb looked HARD at what he does (or did) for a living and asked a hard question we should all ask - IS THERE ANY THERE THERE? 

His conclusion was meaning and patterns exist BUT not nearly as much as we think. The world spins and turns whether we understand it or not. Keep that realization in mind because we are about to go down the rabbit hole. 

Web Metrics Are Stupid
The rest of that sentence is "Web metrics are stupid, but they are the best stupid we have so we love them". The Rolling Stones are LOUD and Jagger is the devil as I write this and that feel appropriate. 

Web metrics are stupid and meaningless because of how poorly we understand online attribution, who does what and what that DOING really means. Attribuiton is goofystupid.

Sure if we had hedge fund resources we could hire QUANTS and momentarily touch the face of God but to what end?  The web is a sand castle on a beach where the tide is always coming in. 

The web works on a simple often overlooked idea - the law of large numbers and tiny differences coming at ever increasingly FASTER exponential increments. Google becomes Google by multiplication of tiny advantages across a LARGE number of people over TIME to become CASTLE GOOGLE. 

Because Google is LARGE and IN CHARGE doesn't change the nature of the game Taleb would explain. When white swans are all you see then you may be "sure" black swans don't AND CAN'T exist.

The nature of the game is the math of content networks described in the book LINKED: Why Everything Is Linked To Everything Else form Rules of Internet Marketing Chess. 

Be sure to read BURSTS by the same author (Barabasi) too. Then read Black Swan by Taleb. If Linked explains the RICH GET RICHER content network truth Google mined BURSTS explains how we are not as unique as we think. Our patterns are knowable, predictable in an Asmovian psycho-history kind of half creepy way. 

This is NOT to say Internet marketing is non-axiomatic, just a random chaotic ramble at all times. Nope, Internet marketing is a non-random wander masquerading as fledgling democracy. 

Post Google's BIG CHANGE the new boss is much the same as the old and RICH are getting RICHER. Two axiomatic truths all Internet marketer learn (eventually)?

* The only time that matters is NOW. 
* Patterns you think you see don't really exist and that's okay. 

That last bullet is the truly ADVANCED idea. When all data is EQUALLY meaningless any data will do. If you are sitting across the table from an Internet marketer who insists on their view they are an idiot. 

Any "my view" of things is an expense no true IMer can afford. Maybe there should be a third bullet:

* When in doubt return to the first bullet. 

:). M