Slicecast

Have been playing with Clicktale recently, and have been very impressed with their video-playback feature that allows us to look at how users interact on your site. While I've only been testing the tool on my site, which is perhaps not truly reflective of real user interaction, but playing back my video (shared below) gave me real insight into how visitors are struggling with content on the blog. The video below is only a sample, but I've not got a few optimsation options for the blog.

Am looking forward to testing this out on a 'real' client site now that I have validated the core functionality. 

Watch the Clicktale Video here

You have to love the tenacity of the Chinese, see a product and make it better. There's nothing more refreshing than walking around one of the fake goods markets in Shanghai to find a pair of classic Adidas that are better constructed than their original brethren.

Well, it appears the same is now happening with free analytics products - today I discovered the new product from Baidu, called Baidu Tongji,  (English translation here) that appears to be an almost carbon copy of Google Analytics. 

There are a few subtle differences yes, but on the whole the products are identical:

 

The image is pretty hard to see, but if you go to the live Baidu Tongji demo and have your own Google Analytics account from which to make the comparison you'll get the idea!

My thoughts are mixed, copying is bad but offering a Chinese language web analytics tool for free (this has to be verified as I believe the tools is under beta still) is a great thing for China.

Just one day after we start to digest the US$15b valuation of Facebook after Microsoft's 1.6% investment, we see what is possibly the most interesting acquisition in the digital analytics business - Omniture announcing its plans to acquire Visual Sciences for a stock and cash transaction estimated at US$394m. Sweet!

In a series of interesting buys of big fish buying little fish we saw WebSideStory acquire Visual Sciences earlier this year for US$57m; which potentially was going to create its own analytics powerhouse, yet now Omniture has minowed this deal..

There is no announcement on how Omniture will intregrate / merge / silence any of the varied products in the combined portfolio, and with the deal unlikely to be closed until Q2 2008 there is plenty of time for the management team to figure that out.

This decision is interesting for me for the following reasons:

1. Consolidation and better Consulting / Support. A combined OmniVisualWebSideSciences-ature will mean better customer service and client relationship handling for existing customers; by this I'm referring to the potential to provide improved consulting services to customers to help maximise the value of their digital analytics investment. With the various conversations in the blogging community about the "Web analytics is Easy", "Web analytics is Hard" there is no question that many customers of the web analytics vendors have been thrown in at the deep end - what I've seen frequently is a great vendor sales team who promise the world, but then leave clients alone once a 'basic' (or default tag) implementation has been rolled out. Let's hope the combined company can better partner with consultants, agencies and clients to educate and grow the industry.
   
   
2. Technology and Platform integration. The technology and data architecture of Visual Site is very interesting, and should be of great interest to the Omniture product engineers. We need better data, we need products like Omniture Discover, Visual Workstation, Visual HBX and the like to be the default analysis front end, not the 'deluxe' bolt on to a pretty, but standard reporting tool. Increasingly our clients are migrating their businesses to the digital space, and in doing so want better data integration - they don't want siloed products such as 'web analytics', 'adserving, 'crm systems' - we need the combined flexibility and plug 'n play features of Omniture Genesis married with the integration capabilities of Visual Site; we need open data architectures that allow us to customise, distribute and share data. A good example of this would be through the integration of testing and behavioural products into Site Catalyst through the integration of Offermatica and Touch Clarity. How about extending this? What I want to be able to present to my clients is collaborative ways of leveraging our customer data to build (for example) collaborative filtering systems to use customer preference data to improve the user experience. I want a data API I can hand over to our Web 2.0 engineers to give customers a truly interactive, responsive environment.
   
   
3. Moving into Analytics and away from Reporting. As I mentioned above we need better, easier access to 1-1 visitor data; the days of aggregated reports are vanishing - I want to be able to slice and dice my digital data; I want to be able to integrate multiple sources to get a 360 view of my customers and I want an open system for data analysis. Far too often clients ask relatively straightforward marketing questions that simply cannot be asked of the data without considerable customisation or requiring my client to "buy the advanced analysis plug-ins'. I want a non-hierarchical structure that allows us to define our web marketing and distributed media business dynamically when collecting data from sites.

The integration of two great companies gives us an opportunity to start addressing the points raised here, and to help move the industry in a positive direction that is matched with the speed of change that comes about from the paradigm shift we are seeing in the digital world (did we forget all media is going digital, so our analytics products need to plug-into the matrix..).

We talk about the death of the web page, but what we should actually be considering is the death of the website - distributed push / pull, closed / open networks where consumers and b2b transactions are occuring outside of the traditional concept of the website is happening. We need to therefore start rethinking the way data is collected - the old ways of scripted tagging need to be rethought...

We do live in interesting times, so I'm looking forward to continue working with the combined OmniVisualWebSideSciences (great term Eric!)....should be fun :)

I've been thinking about this for a while - search engine marketing has become pretty sophisticated over recent years. However, I think one of the core considerations of managing a successful search programme is to extend way beyond the search engine. By this I am referring to integrating web and media analytics into search and migrating from a tactical measure focus (clicks, bid cost, ctr, even conversions) - and creating a customer centric approach. I am referring to this as a universal measurement framework - a set of ideas and something that can be customised per your own requirements.

From the planning of a campaign (engine selection, keyword selection, copy testing) through to customer experience measurement (landing page selection and optimisation, website analytics, conversion and process management) lead to a six dimensional approach to successful search programme management.

Rather than go ahead and creating a whitepaper (too time consuming..) I decided to prepare my thoughts via powerpoint and use slideshare to host the presentation.

These are my thoughts:

What do you think? Do you apply a similar framework? Are there things I've missed from this? How do you approach SEM programme measurement? 

So the story goes, if a tree falls in the forest and no-one is around do we hear it? I feel the same thing is happening with Engagement.

Since arriving in New York in January 2007 I've not had much time to devote to this blog, which is embarrassing and yet insightful - the digital practice at Ogilvy is growing very quickly and at the same time the demand for digital analytics is increasing. As part of this, there is demand for insight into the fabled definition of engagement. What is it, how do we measure it, who owns it; and other equally vague questions. This is not a criticism of Ogilvy, no it is a criticism of the industry.

The point I wanted to raise, and I think relates well to user considerations for measurement strategies is that engagement by its very dictionary definition requires a two way (return path anyone?) process to allow consumers or BDM's or TDM's or C-Level execs to participate. If the content is not there, then we can measure widgets, WOM, blogs, buzz all we like but the volume won't be there. As digital marketers our role is to develop channels and programmes that facilitate this.

Of course there are exceptions, most notably the Dove Evolution campaign which has been a consumer engagement case sudy for all to be interested in - but this is rare - and such marketing effectiveness has a foundation set long before the youtube video was uploaded.

 So if a tree falls over and no-one hears it, does it make a noise?

If a campaign expects 'engagement' and there's no content, does it work?

Maybe. If you're lucky. 

I noticed today that Google Adwords now offers a fairly useful cross-channel conversion tracking system that provides click based conversion tracking for any kind ofdigital campaign you might be running. This may be very old - but it's the first time I've seen it!

The tool is intended for low end marketers and is certainly not going to be a replacement for Atlas Search, DFA, DartSearch, HBX or SiteCatalyst. But that's fine; I'm more interested in what makes this cool - that is creating the ability for any business who wants to use the web for increasing sales / lead generation / demand generation to be able to track and analyse their campaigns without additional investment.

I am 100% in favour of the concept of 'analytics for the masses' (that's an awful way to describe it..) because this does two key things:

1. Increases the likelihood that smaller business will decide digital marketing is relevant, further accelerating the change that we're seeing at the high end of the market and improving the digital dollar shift.
   
2. Help all businesses to be accountable for their investments supporting (a) the growth of the web analytics business at all ends of the spectrum and (b) better channel planning decision making for SMB's without budget to support an agency.

The tool is basic - it's not going to blow the socks off anyone who works in web analytics - but like I said this isn't a problem because it's horses for courses.

As the process diagram above shows us, there's nothing new here - but for people who have only dreamed about being able to track click-throughs all the way through to sale or registration this is nirvana.

Select your ChannelSo, with Google opening the doors on Adwords and allowing users to track their emails, banners (and for smarter innovators how about tracking - widgets, blog entries, rss feeds - in fact anything that has a click event associated with it..) across any other site - even going as far as to list MSN, Yahoo, Lycos etc as alternative advertising channels this really does open the doors to, well, just helping people to run their businesses more efficiently.

The campaign creation process is very straightforward, although it's a shame there aren't ways of allowing you to collect more useful data (for example adding extra meta-fields such as placement, ad name etc) - but then this is what I've come to expect from an enterprise tool. This is free. So fair enough.

Channel AnalyticsAnalytics is also basic - just reporting on clicks and conversions with a conversion rate metric applied - and same comments apply - for new users this is more than enough functionality whereas for users of SiteCatalyst, DFA etc you'd be looking for more.

One market where I see this potentially being used across the board is in China, where clients are reluctant to make the investment in analytics tools. This way for both agency and client it could be a win win situation. For the agency it's a good way to demonstrate that digital media analytics is a great way to show better ROI and helping to make better channel placement decisions. For the client there's no upfront software / tool cost, so if you like what the agency is providing and see a value in it then becomes the time to look at increasing the scale of your analytics.

More screenshots

Edit Details
Get tracking codeSetup complete

 

 

 

 

 

 

 

For further information:

• Review the Google Conversion Tracking training presentation.
  
• Review the Google Cross-channel guide (you may need to login to adwords to see this)
  

CX NOW! is a free, lightweight version of the excellent Crystal Xcelsius tool offered by Business Objects.  As a free tool it obviously does not include the depth of visualisation tools offered by its parent, but still offers a nice opportunity to increase the beauty of your Excel reports.

This free version allows three different visualisation tools:

• Selectors (List, Label, Toggle)
  
• Basic Charts (Line, Column, Pie, Bar)
• Single Values (Gauge, Dial, Slider, Value)

A great example use of the tool would be in conjunction with web analytics reports produced with a tool such as Websidestory's HBX Report Builder; that offers a more robust analysis platform and access to your data via a datawarehouse plug-in to Excel. Using CX NOW you are much better able to recreate the look and feel of the visualisation offered in the HBX web interface if your viewing audience is more used to that presentation style.

A most excellent post by Chris at trackingtechniques.com today provides a reference to best practice in hiring web analytics staff. Hop on over to his site to read the full article, which will be a valuable 10 minutes spent.

Click to read more ...

It's sad I know, but I obsess over visual analysis. Having spent three years working in Asia one of my biggest daily challenges is presenting data to internal teams and clients. This has led to me becoming increasingly compulsive over how data is presented leading to fascinations over creative presentation of data. Making the data speak for itself is immeasurably more valuable than lots of text in a powerpoint that uses overly complex terminology that simply won't be understood by people who speak English as a second language.

Some of the key lessons I've taken from this experience are:

1. Simplicity - only show crucial data points, and refrain from incorporating complex data tables unless there is a valid reason. Determine the point you need to make and make sure that is highlighted.
   
   
2. Visual Tonality - use colour to make a point, for example using reds for negative and green for positive. Use colour palettes consistently across your analysis, rather than picking your favourite colours. Good techniques include using neutrals (grey, pastel) to highlight previous data over bolds (primary colours) and increasing the depth of field by increasing the thickness of lines.
   
   
3. Size - there is nothing worse than an ill-prepared chart that uses too small a font for annotation or is centred in a slide with huge amounts of white space. Additionally one the the most ugly mistakes to make is stretching a chart to make it will the space. Use Excel to keep the proportions consistent.
   
   
4. Format - experiment with different chart formats, but certainly avoid the use of 3d charts at all cost. You will end up spending too long formatting them and the result will not benefit you. Look to the point of Simplicity and think about the point you want to get across - how can you visualise this. A key lesson here is to use custom charts; once you have a format that works save the formatting - this will save you time in reproducing content.
   
   
5. Style - make sure you don't over analyse, for example make sure your presentation is only as accurate as your source (do not use large decimal accuracy like 5.264% if your source only refers to 5.3%). Do you need gridlines? How important are the scales?
   

Examples of sites where you can find nice charting art (look for ease of use and consistency..) are:

• Forrester
• eMarketer
• BBC News
  

Two examples I've found make these points. Firstly we have a chart from the BBC news site and it's nice - simple to understand, very clear and does not require thinking about. The second example is a generic chart (taken from Jon Peltier's excellent Excel site) - it's a 3d pie chart and has been referred to as the 'trailer trash of charts', beautiful!

1. BBC Chart Art
2. Trailer trash chart 

From the presentation Eric T Peterson gave at the emetrics summit in October. It's brilliant, and summarises what I have been trying to sell since I started in Web Analytics. Thanks Eric for such brilliant work!