www.youtube.com/appliedinsights

So far we have some video blogs up there from the recent Emetrics conferences in San Francisco and London. Over time we will be adding to that with discussions and video seminars on various topics to do with digital marketing and predictive analytics, or even digital marketing predictive analytics!

Here’s a sample which is Neil catching up with Avinash Kaushik towards the end of Emetrics San Francisco 2008.

 

But then it gets a bit trickier, the questions start piling up. Like:

• What would a European WAA look like?
• How would it be organised?
• How would it work (or not) along side the existing WAA?
• What kind of legal status would it take?
• How would it be funded?

And probably a whole lot more that I haven’t thought of yet.

I have to declare an interest here. As a Board member on the Web Analytics Association, my responsibility is for “International”. I think last year was the first year that a Director on the Board had responsibility for “International”, though there has always been an International Committee ably co-chaired by Vicky Brock and Steve Jackson. Vicky, Steve and others have done a great job over the years helping to get activity happening at the local level in markets around the world. When I came onto the Board a year ago I agreed with Steve and Vicky that our priorities should be to continue to expand our international reach and also to look for ways to deliver more value to our international members.

A year on, have we done as much as we would have liked? Probably not. We are all volunteers, doing this is our own time. Most of us run our own businesses but I’m amazed at the amount of time that people do put in around the world on a volunteer basis.

Have we made any progress? Yes, we have. We have a new structure in place on the International Committee that should allow us to expand without losing focus and coordination. We have expanded into new markets by appointing country managers in places such as France, Spain, Russia, Argentina and Brazil. We are looking at how we can expand our activities into Asia. We are working on the structures and processes which will enable us to better help volunteer activity on the ground.

Could we be doing more? Absolutely. And it is a real case of “many hands making light work”. We need people to step up to the plate and get involved. I know it can be frustrating that sometimes we don’t seem to react in real time but as I said before that as a volunteer organisation it can take time to have the meetings, make the calls, to come to the decisions.

So back to the question… Does Europe need its own WAA? The answer is still “probably” but, the reality is that at this moment in time I don’t know. There is no doubt that International representation is getting stronger within the WAA. In 2006 there were no European Directors on the Board, in 2007 there were two. In 2008 hopefully more! As someone who spent many years working in the European divisions of US companies, I am well aware of the frustrations that can cause! I do think though that the WAA is becoming more internationally orientated and this debate about a European WAA is a great one to have. For me the next step is to work out how we get to the point of decision. There’s a lot of work to be done finding out what’s the best thing to do and how best to do it. We’ll be kicking that process off in San Francisco next week. After that I am sure we will be looking for all the help we can get! If you’re interested in helping out with the International activities and development of the WAA, let me know.

The web analytics space has many great bloggers, covering many different areas. Some are great on the technical side of explaining how to “do” web analytics, some cover industry developments and so on. I think it’s important to find some way to contribute to the debates without just adding to the noise. So, we’re going to give it another shot. John’s going to look at developments in the world of predictive analytics and data mining and I’m going to muse about digital marketing analytics and what’s happening here in Europe. We’ll be trying a few things out to see if they work, some will, some won’t, but that’s the name of the game. Stay tuned.

For more details and to register, click here.

During the nomination process we were asked to respond to some questions which might help people understand why we wanted to be on the Board. However, I can’t find this copy on the WAA website anywhere (and I have looked), only the profile that we were also asked to complete.

So I thought I would publish my responses to the questions here in case you wanted to understand my perspective on industry matters and motivation to be a part of the WAA team.

There are also many great people who are seeking election, some of whom I know and some of whom I don’t. Since you can vote for as many people as you want, I don’t feel too bad about this shameless canvassing! The greater the depth and breadth of experience and expertise the WAA has at its disposal the better I feel.

Anyway - here’s my “manifesto”.

Which current committee do you have a passion about? Or is there a committee you would like to see us undertake?

Coming from the UK I’m naturally interested in the International aspects of the organisation’s work. I’m also interested in raising the profile of the industry, so I guess that means Marketing!

What is the greatest challenge you see the Web Analytics industry facing?

I sometimes think that the industry defines itself in too narrow a position. I’m a great believer that you need to have a holistic approach to thinking about the effectiveness of their investments in the online channel. The industry needs to continue to broaden and integrate with other online marketing technologies and data providers if it is not to become a commodity data provide to the online marketing industry.

What major contribution will you bring to WAA and its membership?

I think I will bring a broader perspective to the WAA. I have worked in marketing analytics all of my career but in a number of different functions and environments. As a result I think I have a rounded perspective and approach to web analytics and what it means for organizations. I think this perspective will help the WAA to continue to move forward.

I also think that I can help support and develop the WAA over here in Europe.

Why should people vote for you?

Because they want to?

Where do you think the organization should be in the next year?

As I mentioned earlier I think that the organization needs to recognise and embrace the wider perspective when it comes to measuring and optimizing the online channel. It should look to forge partnerships with other associated bodies and organizations in areas such as usability, marketing research and surveys, advertising research and so on.

I think this is an interesting acquisition and perhaps not wholly unexpected. I often wonder about the future of European vendors in this space and what the long term prospects are for companies such as Speedtrap, Site Intelligence, Red Eye and others. The larger European vendors all have some distinctive feature to them such as a particular data collection methodology or visualisation tools, or they have a strong penetration within certain markets or channels.

With Instadia it looks like a bit of both. Obviously they had a strong presence in Scandinavia and were beginning to make moves into other markets such as the UK and it looks like this was a good fit for Omniture. However, I have had the opportunity over the past few months to take a good look at Instadia and there are some neat aspects to their technology as well which will hopefully manifest themselves in SiteCatayst at some point in the future.

One of the unique things about the Instadia product is their integrated survey tool. As you may know from other articles and blogs I’ve written I’m a big fan of using multiple sources of data including surveys to fully evaluate the performance of the online channel. A number of the web analytic tools can now “integrate” survey data and aloow users to do some basic filtering of the data based on survey responses. However, the Instadia tool has a survey tool built into it so that you can design, script and launch the survey from within the system itself and the data is fully integrated at the visitor level with the site behavioural data. I think this is very powerful.

If I had a small gripe about Instadia it was that I thought that the interface was a bit clunky. But I also tend to segment systems between tools that are predominantly reporting orientated and tools that are truly analytical. As a sweeping generalisation the latter group usually have clunky interfaces. It’s the nature of the beast. I think Instadia falls into “analytical” camp and the other feature that it has that I like is its “Filter Builder”. This is a visual way of creating filters and segments and it enables quite complex filters to be built very easily. It’s one of those neat ideas that works very well.

So I think that Omniture have picked up a little gem of a company with Instadia. From what I know of the two companies it should be a good fit. Good luck to Anders and the rest of the Instadia team. I hope that we see some of the more distinct Instadia features on the Omniture roadmap at some point. It would shame to loose them from the space.

As I mentioned in an earlier blog; if the primary data source is a data warehouse which contains well structured, rigorously cleaned and de-duped data, then this is usually the best starting point. But it is only that. The shape of the data tables in the warehouse will inevitably have been defined with a certain type of analysis in mind; most often to produce the standard business intelligence style of reports. You might get lucky and find you can use that data as-is for the kinds of predictive analysis you have in mind. The chances are that you won’t, and that you will have to re-structure the data in preparation for that analysis.
Furthermore it may well contain aggregated data, perhaps an OLAP structure of some sorts, which may allow you to produce time series forecasts but which will most likely contain data which is too summarised for most other kinds of predictive analysis. If this is the case then you’ll probably need to go back and locate the sources of the summary data. That might not be a trivial exercise.

How did we get here?
In the previous steps, discussed in other blog entries, we effectively designed the analyses which we intend to perform at the next step; Data Modelling. In Data Understanding we learnt all about the existing data structures, formats and sources and we started to look for patterns in those sources which are pertinent to the analytical objectives we defined at the start of the process. The truth is that, to perform the exploration, we would have had to prepare the data to some extent. But this is the point where we get serious and apply the necessary data management steps to get the data into the shape(s) required for the main task; predictive modelling.
At the top level this means we end up doing one or more of the following:

• Cleaning data
  This may not be necessary depending on how “clean” the original source is (though it is not unusual to find data problems when we start to analyse it in an unfamiliar way). Our previous exploration should have revealed any errors, or inconsistencies, which need to be corrected, or excluded.
• Merging data from multiple sources
  If you are lucky the data will be in a single data file, or a single table in a database. If you are unlucky it will be in a variety of disparate sources with different formats in various locations
• Shaping it for the analysis
  Often the most time consuming element. A classic example is where we have data with a sequence to it; typical if we are looking to predict the likelihood of a an event given a set of previous events. The starting point is typically data in a database which often contains all event transactions. In order to model it in a way which mimics how we will look to apply (deploy) the model we have to define an appropriate point in history as the baseline, e.g. if we are interested to know what will happen after March 2007 we might use March 2006 as that anchor point. We then have to restructure the incoming data to derive all the interesting predictors e.g. transaction frequency, transaction value in previous months, years, etc. from March 2006 backwards. We also need to have a separate data partition which contains the “what happened next” data for a period after March 2006 that corresponds to the period we want to predict into in 2007; so if we are interested to see which customers are likely to churn in April 2007, then April 2006 is likely to be the best month to look at it 2006. NB. Modelling and Evaluation (see later) will help test that hypothesis.
• Deriving new data elements
  Typically new fields(variables). In our exploration, for example, we may have found that there appears to be a strong relationship between the rate at which a customer buys products and the likelihood that they will churn. In many cases that rate will not exist as a separate measure in the current data, so we create it in this step.
• Describing it
  Labelling, formatting and generally documenting the data in a way which helps the analyst, or other viewers of the data, to understand its meaning.
  The outcome of the above is a set of tables, or data files, which are in the shape we believe we need for the modelling effort we have in mind.

You [almost always] never get it right first time
We’ve mentioned it before but it is worth re-stating that much of the CRISP process is iterative. Quite often we will get into the modelling step, for example, and discover a potential relationship that looks interesting but which we have to go back to the preparation process to derive. Frequently, because we are often building complex data handling processes from scratch, we just make mistakes which need to be corrected.
With large datasets the preparation time can be significant; It can take hours … sometimes days, so mistakes and re-runs can be costly. Hence wherever possible it is a good idea to test the process using data subsets, ideally random, or at least representative, samples. Samples can also be used to boost productivity when we get into the analysis…more on that next time.

An example
Data collected in the web channel is a great illustration of this point. We work with a lot of this kind of data typically for web sites with large numbers of visitors; usually millions per week. These sites inevitably have a web analytics tool which they use to analyse key metrics of site performance. Most often we are interested to apply predictive and/or segmentation methods to the site data. This typically involves:

• Extracting behavioural data from the data warehouse (underlying the web analytics tool) or via a data feed that the analytics vendor provides. More often than not we extract this data to a number of text format files.
• For our Customer Journey Framework we usually have an additional data source in the form of on-line surveys. Depending on the analytics tool that the client is using we have developed a number of ways of linking the data that the visitor provides as a respondent in the survey to the behavioural data which maps that visitors journey through the site.
• The data we end up with can be at various levels but more often than not it is at the individual page or individual click level (remember these sites have millions of visitors so the number of records gets multiplied up). We take this data and aggregate it over a period of time to end up with tables for analysis which are at the visit and/or visitor level. Each of the resulting records will contain fields of interest; e.g. site content viewed, visit intentions and conversion goals which we will use for analysis.

For a typical site processing a weeks worth of data into the shape needed for analysis can take 4-6 hours.

Which tools?
As is often the case the choice of tool for data management comes down to those that the analyst/data is familiar with. Database tools are all about this type of work and often the best approach is to aim to construct data mining tables inside a relational database. This can be achieved using a combination of SQL, ETL tools and other database utilities.
Generally speaking; the more sophisticated the predictive tool itself the greater the data management capabilities which are built in. So SPSS, SPSS Clementine, SAS and SAS Enterprise Miner offer a broad range of data handling procedures.

So much for progressEven though we have more and better tools, and faster hardware, with which to manipulate data these days this is offset by the increasing volume of data, complexity of structures and number of sources. Hence the old adage that data management consumes more of a data analysis effort than the analysis itself typically holds as much today as it ever did. But it is a necessary pain to get us to the point where we can get to the next step which is at the core of the predictive process; Data Modelling.

I’m also thinking that there are some other issues that you should consider when thinking about which system is best for you, particularly in the UK at the moment. It’s nothing to do with KPIs, funnels and the like. It’s about “fit” and “partnership”.

Let’s have a look at what’s going on at the moment. In the US, the web analytics systems market is beginning to mature. Most companies who are doing any serious business online have already invested in a web analytics system. In fact they may be on their third or fourth. As the online business has become more important, the web analytics system has become a bit more “mission critical”. Efforts have begun to integrate the web analytics system with other corporate systems. So, the cost and pain of switching from one system to another is getting higher and higher. The net result, as I see it, is that in the US, more companies have a system in place and organisations are becoming increasingly reluctant to switch and so growth opportunities for the vendors are diminishing.

This is not the case over here in the UK and the rest of Europe. Many organisations are still in the process of sorting out their online channel measurement strategies and are looking to upgrade their capabilities. Across Europe existing localised solutions are being centralised to help manage costs and enable consistency of measurement across markets. There is a lot of “buzz” in the market and the various vendors are busy responding to proposal requests, doing pitches and making the sale.

This is all great but the trouble is there aren’t enough people around to satisfy the current demand and I think that this is having an impact on service levels in the industry. Going back to my comments about what else to look out for when choosing a vendor, most of the main systems these days are basically functionally equivalent. One system might have some “bells” and another one might have some “whistles”. Or you might prefer the way that one system does its conversion funnel over that of another. These may be big enough differences from which to make a decision but they also may not. So how else can you decide? Well, it might be on price or it might be on service.

If you’re making serious investments in your web analytics capabilities, both in the system you’re buying and the resources to manage it, then I think you need to look at the kind of service and support you can expect after the salesperson has done the deal. I don’t think that’s it just about whether they have extensive online support capabilities or whether they have a department called “client services” or “best practices”, it’s more about whether they seem interested in you as a business rather than you as a sale. Ask to meet the people who are going to be managing your account before you sign the deal and see what you make of them. Are they people you want to do business with?

For all the reasons that I mentioned earlier about the current market situation in the US, if you are going through a vendor selection programme now, you don’t want to find yourself in the positron of wanting to make a switch in 18 months time. So it’s important to get it right this time round and that means not only making sure that the system is functionally fit-for-purpose but also that the vendor is a company that you feel comfortable working with for a number of years.