Analytics

This post originally appeared on Applied Insights’ blog. Foviance acquired Applied Insights in November 2008, with Neil Mason joining us as Director of Analytical Consulting. As part of this acquisition, we’ve incorporated Applied Insights’ blog into our own.

It’s been a busy couple of weeks! Over the past few days we have been completing the transaction by which Foviance has acquired the business assets and brand of Applied Insights. I have joined Foviance as Director of Analytical Consulting and will be taking a seat on the Foviance Board. John is pursuing his own consultancy activities and will continue to work with us in the future. You can see the official press release here.

So what’s the background and what does all this mean?

Well, I’ve known the guys at Foviance for some time. We have and have had many common clients. Foviance is an excellent company and has a strong reputation in the market for customer experience consulting. They have been aware for some time that data, analytics and insight are critical to understanding the user experience and getting it right. The acquisition of an analytics consultancy like Applied Insights is a natural move in terms of developing the overall proposition. I’m delighted to be joining them and although it’s been less than a week, it feels good. One of the reasons why the fit with Foviance feels right is because we have common values and beliefs about what “good” looks like.

I’m heading up the analytics consulting team which is called Foviance Applied Insights. Maybe not a totally original name but “it does what it says on the tin”! In the team I will be bringing together the existing analytical capabilities of Foviance and Applied Insights and will be scaling and expanding our services over the coming months. In addition to Applied Insights core competencies in strategic analytics consulting, customer analytics, predictive analytics and optimisation we will be growing the existing capabilities within the Foviance team in web analytics consulting and developing our services around some of the core technologies out there such as Omniture, Google Analytics and others. It’s going to remain busy!

For me this is a really exciting move. With a core team in place and with the Foviance infrastructure behind us I’m looking forward to extending our capabilities, developing our ideas and continuing to offer quality consulting services and products. Look out for news either on this blog or on the main Foviance site.

This post originally appeared on Applied Insights’ blog. Foviance acquired Applied Insights in November 2008, with Neil Mason joining us as Director of Analytical Consulting. As part of this acquisition, we’ve incorporated Applied Insights’ blog into our own.

London, UK, 6 November 2008 - Foviance, the expert in digital customer experience, today announced the completion of its acquisition of Applied Insights, a leading provider of analytical consulting solutions to blue chip businesses. The acquisition will accelerate the development of Foviance’s technology agnostic analytics division and supports Foviance’s strategy to provide a comprehensive customer experience solution to its global customer base.

The acquisition brings Applied Insight’s director and co-founder Neil Mason to the Foviance team as director of analytical consulting. With his vast experience, he will serve on Foviance’s board of directors and lead the company’s new initiative and analytical consulting practice. Mr. Mason is a renowned industry figure and brings Applied Insights best practices of predictive analytics, optimisation and measuring digital marketing effectiveness to Foviance. He also brings with him Applied Insights existing blue chip client base which includes Barclays, BP and Total Jobs and he will have a pivotal role in developing further Foviance’s strategic partners such as Omniture and Google.

Neil Mason joins Foviance with 25 years of in-depth industry experience in marketing analytics and strategy. Prior to founding Applied Insights, Mr. Mason has worked in a number of senior leadership roles for many major companies including QXL ricardo and Research International UK. He also currently serves on the board of directors of the Web Analytics Association, the global industry body for digital analytics professionals.

“The combination of Foviance’s expertise in delivering seamless cross-channel experiences with Applied Insight’s leadership in understanding data analytics, is ideal for our customers, both in providing innovative services and optimising usability and conversion rates” said Paul Blunden, CEO, Foviance. “Our combined global customer base now has a partner with proven expertise to deliver a strong blend of business, analytical and technology consulting capabilities.”

“Many e-businesses are looking to improve their online performance but are not clear about what they should be measuring or how to measure it effectively, our merger with Foviance presents us with an incredible opportunity to improve the digital usability experience for businesses globally,” said Neil Mason, director of analytical consulting, Foviance.

About Foviance

Foviance is a leading customer experience consultancy that works globally with some of the world’s best known brands to deliver measurable improvements in performance. .

Founded in 2001 and with a heritage in website usability and data analytics, Foviance delivers consultancy to its clients about the effectiveness of their individual channels, such as mobile, web and call centre and how they combine in a cross-channel environment. For many clients, insight is provided not only in their home market, but also internationally through Foviance extensive alliance network.

Foviance engages with its customers wherever they are in their product lifecycle, and provides insight so they understand how to improve, create and deliver excellent customer experiences.

Foviance boasts 43 of the UK FTSE 100 companies among its client roster, including Barclays, BSkyB, and Sainsbury’s. In addition Foviance works with International brands such as Astra Zeneca, Dell and Nokia. For further information please visit: www.foviance.com

For further information:

Melanie Hesketh / Becky Cheers

Prompt Communications for Foviance

+44 208 8996 1638 / +44 208 8996 1636

Foviance@prompt-communications.com

Omniture, the online business optimisation specialist, recently announced that it had agreed to acquire search and merchandising assets from solutions provider Mercado.

Another day, another Omniture acquisition! How fun it is watching the leviathan consume all in its wake searching for the Holy Grail that is the perfect online business optimisation solution. Is this the final piece of the puzzle? If so, how will this fit in with the rest of the Omniture suite?

On the face of it, acquiring these central elements of Mercado’s business will allow Omniture to help its own customers market their products better. Online retailers are gaining a good deal of experience selling products and services, but they also want to be able to provide customers with pointers towards related products that best marry with their purchases.

Merchandising in this sense is an effective way of collaborating products together. Mercado’s technology will allow Omniture to record sales details, receptiveness to merchandising, track and compare customer data from external search with internal searches, and manage keyword and pay-per-click campaigns dynamically. The concept of merchandising is not just about collecting information about what customers look at in terms of site design, it is about analysing pure product focus information - what is hot and what is not?

Of course collecting data across online channels is one thing, but refining businesses based upon the information collected is more of a challenge. One of Omniture’s more interesting acquisitions of late was web optimisation company Offermatica. This flagged Omniture’s intent of moving from its analytical routes to something much more ambitious, a completely automated online business optimisation solution.

Test & Target, as it is now known to Omniture customers, provides the ability to conduct real-time multivariate testing - an extremely powerful tool that could provide the answer to what combination of content drives customers the most. But here’s the flaw; what happens if all the permutations Test & Target displays are bad? Just because ‘Layout A’ has the highest conversion rate does not mean it is the most you can achieve, maybe it is the information itself and not the means by which it is displayed. What if there was a means of determining the best content to serve on products to push?

Enter Mercado’s merchandising technology.

Merchandising encourages commercial activity via the promotion of content. This process involves analysing sources of data to determine which content sells and what doesn’t. Quite simply, Omniture is attempting to come up with an automated merchandising solution that analyses data sources and serving product/services information directly into Omniture Publish (online CMS system) via Omniture’s Test & Target.

Omniture started life as a highly-effective statistical reporting company. By acquiring new tools like Offermatica & Mercado, the Omniture suite is another step closer to delivering a fully automated online optimisation tool. The benefit of which will be the most powerful tool on the market giving businesses a ’switch to flick’, kick-starting a self-learning tool that will automatically increase the likelihood of turning visitors into a customers.

Google Analytics came out shouting “We are coming to get you” at the big 3 vendors (Oct 22nd) with the announcement of its new enterprise class features due to be released in the next coming weeks. Advanced Segmentation, custom reporting, a data export API and the long awaited integration with AdSense all feature on the list to really add some serious weight to the GA offering.

After IndexTools was acquired by Yahoo! and they announced that they would join Google in the “free tool” market space, we have been watching with eager eyes to see how Google would raise their game and from what we have seen so far it is a big improvement. The big win for most analysts will be the new Advanced Segmentation feature, which enables you to create your own dynamic custom segments based on multiple dimensions and/or metrics so you can properly slice and dice your data as you want it. Say goodbye to fixed dimensions and multiple profile settings and filters. If you want to find out more, Avinash Kaushik has written a great post teaching you how to become a segmentation ninja. Google Analytics Releases Advanced Segmentation: Now Be A Ninja!

Custom reporting is also another big benefit with a drag and drop interface similar to that of IndexTools allowing you to choose the dimensions and metrics which you want to see in your report. Concerns over data privacy will also be put to rest with the introduction of a data export API, giving users more transparency over the data which Google holds. There is also some new eye candy for all you marketers out there with a funky motion chart to give a new twist to data visualisation.

Some of the features such as AdSense integration are still in private beta for the moment, but I can’t wait to get stuck in. Mamma’s got a brand new toy to play with!

This post originally appeared on Applied Insights’ blog. Foviance acquired Applied Insights in November 2008, with Neil Mason joining us as Director of Analytical Consulting. As part of this acquisition, we’ve incorporated Applied Insights’ blog into our own.

When we think about how to evaluate a predictive model the first thing we typically think of is how accurately does that model predict against the (unseen) test data. More often than not though when we develop models our business/research customers want more than that. They want to know how the algorithm got to the predictions i.e. they want to understand the model.

The more transparent predictive methods don’t just predict they also reveal the patterns that underlie them. The two main benefits of this are that

1. Subject Matter Experts (SMEs) typically on the business/research side - can assess the model’s validity by viewing these patterns, for example as rules or formulae. This way they can see if the inherent relationships make sense. Do they see any potential anomalies in the data that we didn’t pick up when we previously explored it?
2. And of course the patterns themselves may reveal useful insights. We often find specific segments of interest; demographic groups who have a higher propensity to convert through a given channel, or re-purchasers who have short, but potentially interesting and valuable, buying cycles.

The bottom line is that when we can see what a model is doing we can glean much more from it than the likelihood that the outcome of interest (convert, attrite, default, etc.) will happen.

To be frank most of our projects are like this. This is where Decision Tree methods often win out because the output let’s us visually explore the data to both understand the model and to examine other potential patterns of interest. They may not necessarily give us the most accurate predictions but often the SMEs care more about understanding than predicting. This is a classic trade-off in PA.

There are exceptions to this. The alternative view is that accuracy is paramount and it could be that the winning model is opaque. Neural Network models are a case in point. Depending on the software you are using you might see a ranked list of fields which contribute to the prediction along with the prediction itself and perhaps an associated confidence level. Even if the final network is displayed it doesn’t necessarily explain much more.

For the most part these are the two most typical scenarios however we are currently designing a 3^rd type - where opaqueness is the main objective (together with an acceptable level of predictive accuracy of course). We’re talking to a government department who don’t want to have to send sensitive data out and who don’t want our models to reveal any of that information either. So the gist of our approach is that we’ll develop black-box models on our data and let them deploy them on their database. They’ll give us addresses and predictive scores in return but in so doing we won’t know why a particular address was selected.

Anyone living in the UK will understand the political backdrop to this as there have been various high profile cases of data going AWOL (here is the latest one). We are hoping that a somewhat unorthodox application of Predictive Analytics might help the UK government provide a valuable public service without further compromising the confidentiality of its citizens. There’s many a slip twixt the cup and the lip mind you - we’ll keep you posted…

This post originally appeared on Applied Insights’ blog. Foviance acquired Applied Insights in November 2008, with Neil Mason joining us as Director of Analytical Consulting. As part of this acquisition, we’ve incorporated Applied Insights’ blog into our own.

In his series I have been looking at some of the basics of setting up a web analytics programme. Often when organisations come unstuck with getting their web analytics programme working effectively it’s because of issues around planning and processes rather than the technology itself. Last time I looked at campaign tracking and this week it’s the turn of understanding content consumption.

Typically when web analytics systems are deployed “out of the box” the reporting of what content is being looked at is at the page level. We are all familiar with reports such as the “Top Pages” report which shows which were the most popular pages on the website, but the problem with these types of reports is that they rarely change and from looking at these reports it’s difficult to understand overall patterns of content consumption. The data is too granular. Often it’s more useful to know what types of content are being consumed the most (or the least) rather than which individual pages.

The solution to this problem is to assign pages into “Content Groups”. A content group will represent pages that have something in common. For example, all news items might belong to a content group called “News” and there may also be sub-groups or a hierarchy such as “News: Domestic, News: International” for example. Once all the pages are assigned to the various content groups, it’s possible to take a look at how many people looked at a group of content, how long they looked at it, where they came from to reach that content and where they went to afterwards. For sites that are content reach and maybe don’t have much transactional activity, this is more useful and more important.

Content grouping is fine in theory but how does it work in practice? Different web analytics systems tackle content grouping in different ways and some have more flexibility than others. If you are looking at different systems, this might be relevant to your decision making. Content grouping in some systems is dependent on the URL and folder structure and is usually fixed in the reporting interface. Google Analytics has an example of this approach where it is possible to use the Content Drilldown report to look at content consumption at each level in the folder structure. This approach can work well for sites where the content is organised with a neat folder structure but for many sites this isn’t the case and a different approach is required.

An alterative approach to content grouping is to assign pages to groups in the data collection tag. This approach is more flexible. Content groups can be defined independently from the folder structure of the website and in some cases a different hierarchy can be developed as well. Pages can then be assigned to content groups by customising the page tag. But flexibility comes at a cost and that cost is in development and maintenance.

First of all at the point of implementation, a plan is needed of what the content group structure is going to look like and how it is going to be implemented. For a large site with lots of content, this can be quite a significant exercise and require a good deal of planning. It’s also something that needs to be considered for site refreshes or rebuilds. The implementation approach will depend on the technology behind the site such as the content management system being used but ideally there will be some rules based approach which will help with the ease of implementation. At the end of the day there may be trade offs that need to be made between what the content groupings might look like in an ideal world and those that can be achieved in practice.

After implementation there is also the matter of maintenance. Most sites are dynamic in the sense that content is regularly being updated and changed. Pages are added, changed or deleted. In order to maintain the integrity of the data processes will need to be put in place to ensure that as pages or sections of content are added that the content grouping is managed at the same time. So, it needs to be worked out who is going to own the process, who’s going to manage it and who’s going to be responsible for doing it.

As with campaign tracking (which I looked at last time) the success of measuring content consumption on the website is not just down to what technology you’ve got but also down the planning skills and the maintenance resources that you put behind it.

This post originally appeared on Applied Insights’ blog. Foviance acquired Applied Insights in November 2008, with Neil Mason joining us as Director of Analytical Consulting. As part of this acquisition, we’ve incorporated Applied Insights’ blog into our own.

In my last column I outlined how organisations can come unstuck with their web analytics if they don’t pay sufficient attention in general to the integrity of the data they are reporting. It can seriously impact on the decisions that the organisation is making. One of the areas in particular that I have seen organisations struggle with using their web analytics tools is campaign tracking and once again it’s often the processes and not the technologies that are the root cause of the problem.

The ability to track marketing campaigns is now a standard component of any web analytics tool. We don’t need to worry anymore about having to set up specific landing pages and tracking referrals to the page. Most web analytics tools now use the same principle of campaign tracking. This involves of adding a tracking parameter to the end of the landing page URL to identify the piece of marketing activity. The web analytics tool is then configured to recognise the tracking parameter at the end of the landing page URL as a visit generated by a campaign and then populate the database and reports as appropriate. Simple enough in theory but often trickier in practice.

Some of the common pitfalls that lead to poor quality campaign tracking data are:

• Campaign data is not properly structured
• Campaigns are not consistently tagged
• Campaigns are not consistently tagged consistently

The first of these pitfalls is a planning issue. The second two are process issues.

Most web analytics tools have a framework or structure for campaign reporting. This is where a specific piece of activity is identified by a series of attributes. These attributes are then used to provide different levels of reporting. If we take Google Analytics purely as an example, then a piece of activity can be described using up to five different attributes (Source, Medium, Term, Content and Name). Part of the campaign tracking implementation process is to determine what these attributes mean for your own campaigns and how detailed you want to be. It’s important to think ahead about what activity you might want to run in the future as well and how that might fit into the framework. For example, you might be running only one type of email newsletter at the moment but if you develop your email marketing strategy to include different types of more targeted emails, will your campaign tracking approach allow you to identify how each of the different types of emails are working?

Whilst the underlying principle of campaign tracking is generally the same across most web analytics tools, the framework for reporting does differ from system to system. Some tools are more flexible in their approach than others. Whatever the tool though, proper planning is required to ensure that the right kind of reports come out the other end.

After planning comes process. Having decided how you want to structure and report on your campaigns, the campaign landing page URLs need to have tracking parameters attached to them. Sometimes this is an automated process but more often than not there is a degree of manual intervention and that’s where the problems usually start.

First of all, all campaigns need to be tagged to be tracked. This might seem like a statement of the obvious but it is surprising how often in the heat of the moment to get a campaign live, the tracking parameters are forgotten. I know that this doesn’t happen in your organisation but it does in others? Once the campaign has gone live without the correct tracking parameters attached you can’t go back and recover the data. It just doesn’t exist. And the time that you really want to know how a campaign is performing is when it goes live. So, you need to have management processes in place to ensure that all campaign landing page URLs have tracking parameters.

You also need to ensure that the landing page parameters have the right tracking parameters in place as well. For example, if you have an attribute which is “email” to identify all visits coming from emails, then it needs to be used consistently as “email” as opposed to “Email” or “e-mail” or “E-mail”. Lack of consistency in tagging resulting is poor data integrity in reporting. Again, this may seem obvious but the challenge comes when you may have different people or agencies responsible for management different types of campaign. They all need to tag the campaigns in the same way and a degree of process and control is required. This can be helped by having a centralised approach or using campaign management technologies.

So, planning and process are the watchwords for campaign tracking success.

This post originally appeared on Applied Insights’ blog. Foviance acquired Applied Insights in November 2008, with Neil Mason joining us as Director of Analytical Consulting. As part of this acquisition, we’ve incorporated Applied Insights’ blog into our own.

It was one of those moments. I was working on a client’s data and I began to suspect that something was wrong. Not with the client’s business but with the data, but the potential business implications were very significant. Sure enough, as I dug deeper and deeper into the issue it became evident that there was something seriously wrong and I got that sinking feeling.

The story behind the story was that the business in questions was looking to aggressively improve the effectiveness of the digital channel and had been focussing on conversion optimisation as traffic levels were quite buoyant. They had implemented a satisfaction tracking survey to understand visitor intent and satisfaction; they had commissioned usability testing to understand the user experience in more detail and they had started a testing and experimentation programme. It was all the right stuff, the problem was that there wasn’t any strong evidence that conversion was actually improving. So it needed a deeper dive into the data to find out what was happening and that’s when the problem emerged. Without going into the gruesome details the impact was that the conversion rate was being underestimated and that the degree of underestimation had been getting worse over time. This meant it was a case of “What do you want first, the good news or the bad news?”. The bad news was that the historical data from the web analytics tool was wrong on some key metrics, the good news was that the conversion rate was better than previously thought.

The really bad news was actually that the business had potentially been focussing on the wrong problem. Whilst all the activity on conversion optimisation was good stuff, the revised data highlighted that other issues may have been more pressing. The other bad news was that the credibility of the data was seriously undermined and to some extent the team as well. For conversion optimisation it was taking two steps forward and one or two steps back.

The point of this case study is that it reinforces the need to get the right numbers right and to keeping them right. When it comes to marketing optimisation good quality data is a core component. Getting good quality data that allows better decision making is a key step on the journey. That might seem like an obvious statement but it is not a process that should be underestimated and nor it is a one off set up event. When a new system is implemented that is inevitably a focus on the data it is generating and that might be reconciled against other data sources. That’s great but those checking processes need to be repeated at regular processes to ensure that the data integrity remains high. If this isn’t a managed, ongoing process then there is the possibility that the integrity will decline over time until something happens which causes the data to be questioned by which time it might be too late.

Managing data integrity is a messy job but someone has to do it. Good processes will certainly help ensure that all pages get tagged; campaigns are tracked properly and so on. Technology can also help with solutions out there that will check for tags on the site, as well as solutions that help address tag management challenges. It also needs a keen eye to be looking at the data for trends and patterns that may not be a true reflection of what’s going on. I actually think this is a skill but it’s a skill that can be learned. A good marketing analysts can sense when something doesn’t look right and in my own experience if something looks odd , then it probably is odd and isn’t real behaviour. Sudden changes in trends, steps in the data, spikes and dips are all potentially symptomatic of artificial impacts on data and if they cannot be explained by real world events, then it’s worth digging into the data to see if there is anything untoward happening like changes to the tool’s configuration, new site monitoring tools being put in place, changes to the hosting environment and so on.

So, don’t see getting good data integrity as a one off event but as an ongoing process. Be wary of the potential impact that changes to your site or your tracking environment will have on your data and plan accordingly. Take time to reconcile your data on a regular basis to see if there are any divergent trends. Hopefully with these basic processes in place you can avoid that sinking feeling at some point in the future.