Analytics & Insight

Well, not quite, but Facebook has now overtaken Google to make it the number one most visited site in the US. Back in January I suggested that Social will soon become bigger than Search and I think that this will very soon be the case.

Facebook is quickly become an absolute internet behemoth, with 175 million people logging in every day - and Read more…

We love data here at iCrossing and all the juicy profit driving insights you can pull from it but sometimes it’s difficult to turn that dry spreadsheet into a simple graphic that really illustrates your point. Recently a few nice tools have been released that we thought we’d share with you.

Microsoft Pivot

Think of pivot tables on steroids. This new, highly visual data manipulation technology from Microsoft allows you to quickly zoom in and out of the data, cut and filter what is being shown all in a deliciously visual fashion. I could write reams of text on this, but as it’s a visual tool, I recommend watching the video below.

Click here to try it out

Google Data Explorer

A nice little tool that lets you look at public data and easily trend it over time. Motion charts are a great way of discovering patterns in the data and see how various elements interplay. The video below shows Read more…

This post is basically re-hashing a lot of the things that Avinash Kaushik said at SES on Tuesday (15^th Feb), but I wanted to share some of my own thoughts about attribution modelling, and highlight some of the great ideas Avinash talked about.

Attri-what?

All the discussion about attribution modelling comes from the understanding that most tracking software will generally attribute a conversion to the channel where the “last click” happened. So if I visit a website through the organic results, an affiliate, a regular referral, then finally through a paid search ad, my purchase at the end of this time will be attributed to the paid search click. I am oversimplifying this, but let’s assume that this is the case.

You can see that in reality many marketing channels have contributed to my purchase, but the last click is getting the value attributed to it. So how can we attribute the value of the purchase in a less biased way that gives a better picture of the value of the previous clicks? This is where the idea of attribution modelling comes from.

Image Credit: CC Semio

All this assumes that you actually have the technology that can give you the kind of data required here. These two posts from Will Critchlow on SEOMoz show how complex this can get using Google Analytics.

Should you even bother?
Read more…

Image credit: Creative Common Attribution: webtreats

2009 was the year social media became mainstream and your mum got a facebook account. Twitter took off in a big way (Twitter bigger then Amazon) and Facebook saw massive growth (Facebook bigger then Yahoo!) but search was still king of the hill, will that change in 2010?

Hitwise released some interesting stats last year that showed that social was now bigger then search in New Zealand and that at the end of the year Facebook overtook Google as the most visited site on the internet (albeit only for two days). So is this a trend that will likely continue into 2010? Well the growth of Social Media has not levelled off yet, people are now using the internet to blog, upload pictures and videos, socialise and network in greater numbers all over the world (click right for larger image). It looks like social media will soon surpass search in Australia and I would expect the same to happen in many countries this year.

So is search passé? Most definitely not, 2009 saw many of the big search engines integrate social data into their search results. You can now see real time Facebook and Twitter results in search pages, results based on your social graph and a whole litany of other social data (images, videos, news etc) right on your search page. The web is evolving, search results are becoming more personalised, social media is playing an ever increasing role online and as such online Read more…

Twitter bigger then Amazon

2009 was most certainly the year of Twitter and in the UK it is now bigger then internet superstar Amazon. Twitter quickly gained popularity at the beginning of the year when celebrities such as Stephen Fry started using and talking about the service. The service has become so popular that now even some beds have a twitter account and will tweet out a “Frenzy Index” when their owners are doing you know what.

LinkedIn bigger then Myspace

2009 has been a good year for LinkedIn, so much so that it is now more popular then Myspace (at least in the UK). Myspace the former social networking king has seen its visitor numbers continue to tumble, while business networking site Linkedin has gone from strength to strength. Myspace has seen its daily visitor numbers fall from 300,000 at the beginning of the year down to 80,000 at the start of December, whilst LinkedIn has seen strong growth signing its 50 millionth member in October.
Read more…

The use of Google Analytics is currently subject to a heated debate in Germany, dealing with web analytics and data protection. On Friday, November 27th the “Düsseldorfer Kreis”, a panel of the German supreme supervisory authorities for data protection, said that the collection of IP addresses violates current law. In particular, the creation of user profiles is caught in the line of fire. What will this mean for digital marketing?

For the storage of personal data there are strict requirements in Germany. They may be collected only with the consent of the user. For Internet Service Providers it’s possible to identify a user on the basis of the IP address. Advertisers who use web analytics software possibly store IP addresses; however, as long as the user is not logged in to in to a web service, it is not possible to retrieve any information about the person behind the IP address.

Special attention is currently devoted to Google Analytics. The German data protection commissioners worry that Google uses web analytics data to create user profiles. In the terms of service  (TOS) (1) Google assured they won’t link the IP address to other Google data, but they also retain the right to change the TOS at any time. Millions of websites are affected. According to a recent article of Zeit online(2) more than 13 percent of all German websites use Google Analytics.

The controversial question still is: Are IP addresses personal data? Now the German courts have to decide. The legal situation is still unclear.

For digital marketing agencies like iCrossing, web analytics are very important. We use tools like Google Analytics to measure visits, page views, bounce rates and other performance indicators. The behaviour of an individual user is not relevant for us, so we have no interest in storing the user’s personal data.

I personally consider IP addresses no personal data. An IP address can change at any time, it can be shared and you never really know who the person behind the screen is.

Nevertheless: If the Germans courts decide that IP addresses are personal data this would be a slap in the face for the whole online industry.

(1) http://www.google.com/intl/de/analytics/tos.html
(2) http://www.zeit.de/digital/datenschutz/2009-11/google-analytics-datenschutz

About a month ago Google released Google Analytics Intelligence and we’ve been having a lot of fun playing with it here at iCrossing. In a nutshell it does statistical analysis on your analytics data and highlights areas where you might want to look, thereby saving you time and potentially highlighting things you would not have normally noticed. The system does a lot of the data crunching for you allowing you to hone in on the interesting parts and pull our great insights quickly and easily.

Read more…

In what some say is a reaction to an industry that refused to self-regulate, the EU has jumped into the problem of data privacy and cookies, and voted to pass a ruling that requires a user’s ‘explicit consent’ before they receive a cookie. There appears to be some exceptions to this ruling, including cookies used for logging in to sites and for shopping carts, but mostly, permission will have to be sought.

This will have significant impact on all types of digital marketing, including display, affiliate marketing, email, analytics and search; how much impact will depend on each member state’s interpretation of the ruling when they implement the law, something that has to be complete by April 26th 2011. We must not take this as a soft threat - the EU has taken action before in relation to Phorm, the ever controversial ad network that worked on deep packet inspection - and never got off the ground in the UK.

Behavioural Targeting techniques like Phorm are probably the cause of the fuss - thanks to the media exchanges we are going through a mini revolution in display media where we can target consumers more accurately than ever before - but how much is too much? Clearly the EU felt that Phorm had crossed the line and are now redrawing the battle plans to say that ALL Behavioural Targeting is too far without a user’s consent.

I wrote about this on my personal blog as the news broke in the US.  Read more thoughts on the new EU law, the implications for the USA and how this may get implemented.

I’m currently researching some data for a budgeting and savings planning calculator for one of our clients, and during the course of my research, I’ve been struck once again by the wealth of free data that is available to every citizen in the UK at the click of a few links.

As digital marketers I’m sure none of us need to be told about the power that intelligently analysed and appropriately presented data hold. After all, where would our businesses be without web analytics? But did you know that there are huge, shimmering pools of astoundingly robust and statistically accurate data sitting on government servers just waiting for intelligent minds to dive in? And all for free?

Here are my two top sources of free, public data - so why not hold your nose and take the plunge?
Read more…

Well not quite but it has added forecasting to its Insights for Search research tool.  The tool will now show you a 12 month forecast of what it expects the keyword volumes to look like over the next year as well as past historic trends for the keyword.  This can be quite useful if you want to look at what the seasonality will likely look like over the next 12 months for a given area/keyword as well as whether that space is likely going to grow, shrink or stay the same.  Where this tool falls over is where there is no easily discernable pattern, where the pattern has changed recently or the search volumes are too low to show a pattern.

The graph below shows historic search trends for ‘Hotels’ within the UK.  It has a nice simple pattern that repeats every year, with a high in January and August of each year, a smooth bowl like curve between and a sharp drop off at the end of the year.  You can see that Google have extrapolated this pattern out over the next 12 months on the right of the graph.  They have also taken into consideration that each year search for ‘Hotels’ as a proportion of web search is decreasing and have reflected this in their forecast.

Read more…