Category Archives: Web Analytics

My thoughts on the X-Change conference 2013 Berlin

x-change conferenceDuring 12-13 of June 2013 I attended the X-Change conference held in Berlin. It is a conference mostly for web analysts with a bit of a twist. It is organised in discussion groups called huddles. These discussion groups contain a maximum of 15 people and you can choose multiple sessions during the two-day period. In total the conference had approximately 100-150 attendances from companies such as Zalando, Siemens, Virgin Media, Dell, Soundcloud and many more.


The good

  • To be able to meet fellow web analysts from all around the world. About 15 countries were represented.
  • Small, medium and large size companies were present.
  • The format is unique. Having discussion sessions is much better than listening to presentations. You get closer to the real problems people face.
  • Intimate session which means you talk to a lot of practitioners even after the end of the session.
  • I have never talked to so many people at a conference and never exchanged so many business cards.
  • The conference was well organized, and the sessions in general fun to attend.
  • The format allows you to get a more intimate insight of other companies problems and solutions when it comes to analytics and e-business.
  • No selling from consultants. Although one of the USP of the conference is no consultant, this was not true. I must say that the consultants attending were extremely well behaved. No sale pitch at all from anyone, which was a unique experience.
  • There were some really smart people attending. It was mostly web analysts but you could also find consultants and business people.
  • Businesses that can be considered to be in the forefront of the web development attended although the US true giants were missing.


The bad

  • How the sessions are experienced depends a lot on the moderator. A good moderator makes people talk and the conversation gets a natural flow.
  • Many of the discussions went too much in technicalities, which in the end of the day is more or less specific to each business model making it hard to extract knowledge from the discussion. I would rather see discussions based on the business value hence raising the bar of the topic.
  • To many people in the session which hampered the discussion. There is a sweet spot around 8-10 people, bigger than that and you have a hard time having a discussion. There was a tendency to have small monologues instead of discussion a bit like the presidential debates on TV.
  • There should have been more business value sessions. Sessions discussing the business value and how you can use technology to move your KPI needles.
  • The price of the conference. 1400 euros is a lot of money even for big companies.
  • The representative of Facebook cancelled (did it have anything to do with the Prism scandal which erupted the days before?).

Main takeaways

  • The mobile web holds a lot of focus for the companies. Many of the companies have native apps but few think about responsive web.
  • Identifying unique users on multiple devices is a problem. The most prominent use unique ”fingerprints” for various devices.
  • The general consent is that serving segmented content increases the conversion although only one company had numbers on it.
  • Privacy issues are a big concern for all companies. The coming data protection law that the EU is cooking is also seen with great concern.
  • A lot of the attended companies have done AB testing in one way or another. Only one company though (Zalando) had a team called Conversion Optimisation. I think that explicitly naming a team like that sends strong signals to the rest of the organisation. Most companies do AB testing under the web analytics umbrella.
  • Tracking is still a big issue for most organisations. The complexity of having multiple products and in multiple countries adds to the challenges companies face. Global organisations also ad another layer of complexity with teams and stakeholders scattered.


Will I attend again?

Well, this is a hard question. Returning customers is actually an extremely good measure of how good your product or service is. I enjoyed the conference and met many interesting people but I feel that the takeaway from the sessions were smaller than justified especially if you consider the price. I would say that I might return in a few years but if the format would improve a bit then I would definitely consider returning earlier. When I say that the format should improve I imply to the fact of having some sort of quality control for the huddles. It could be cutting the groups in half which will spur the conversation further, make the huddles more work shop like for example let people use post-it notes and the whiteboard, provide a summary in the end of each session, and finally lower the price below 1000 euros which would make it an extremely competitive package.

In general conferences are a mixed bag. You never know what you will get. My main mantra is to attend various conferences outside my area of expertise to learn about the various parts of the web like seo, ppc, mobile etc, which lets me grow my online knowledge further.

If you attended I love to hear your thoughts on it.

All web analysts will soon be fired

I think technology is an amazing phenomenon. It is so amazing it will eventually replace your job as a web analyst.

As a society we develop through technology. Every piece of tool, from the invention of the wheel, the hammer, to the car to the computer is defined as technology.

So what do I mean when I say that web analyst will soon be replaced?

As a web analysts you are dependent on various tools to do your job successfully. Web analytics tools are constantly evolving. They become more complex capturing more and more of a company’s business. Many tools allow you through the API send in offline data and combine it with the online information they show.

Looking at today´s tendencies we see that the integration of Business Intelligence and web analytics tools are becoming more prominent. A fellow analyst told me not long ago that web analytics is just a technology to collect the data that then is put into his BI tool. From that he controls all aspects of his business.

As tools develop so does the capacity to start predicting outcomes. Predictive analytics is strongly growing and more and more companies use it to create a data driven culture.

Predictive analytics is in essence the marrying of data with statistical methods. When you do these interesting things start to happen. You can start understanding if you daily changes in metrics are for real or just normal.

For example Google Analytics uses this method to show you some of your metrics are significantly off the norm. Taking this a bit further the more data you put into a tool the more it will be able to understand the correlation between your data hence being able to predict what will happen if a value goes up and down. From this step it is not far to create a tool where you get recommendations what to do to boost your bottom line, today the job of a web analyst! You see where I am getting at? When the tools have been smart enough to start predicting and suggesting outcomes the next step is for management to start asking why they should have a web analytics department at all. This is the time when they will fire you.

Now, you might think this is a bit far-fetched conclusion but let me though put this into perspective through an example.

I read this interesting book about technology that leads to automation. Automation leads to loss of jobs in many different fields. It is also strong driving force for outsourcing. Without todays IT tools and Internet technology much of the outsourcing would be impossible. The most popular medical profession in the US is to become a radiologist. As a radiologist you work office hours and get paid a lot. Your job is to interpret scans of an x-ray machine. A big hospital today needs several people of this profession but more and more hospitals are outsourcing this to India. They send their scans through a secure Internet connection to India where doctors interpret them for a tenth of the salary a doctor in the US would do. Without the Internet technology this would of course not been possible.

Now you ask yourself, how does that relate to you as a web analyst? Well, it is already today possible to outsource part of a web analyst job to another country, but the biggest threat is not outsourcing it is software technology.

The next big leap in IT technology will not be hardware base but rather software base. Today’s software is rather primitive. The technology is moving so rapidly that if you start programing a software solution today and it will take you two years to complete when you are done it will probably already be out-dated.

Predicting the future is not an easy thing but almost all scientists within the IT technology field agree that within this century we will create artificial intelligence. That means computers that learn from their environment and their experience. Imagine what implications this will have for all areas of society and nonetheless for software development. A few years back ago I was discussing this option with a programmer. I told him to imagine a computer that can write millions of lines of code in a day and constantly review how the code works adjust it and deploy it all without the interference of humans. No need for a big team of developers that need to coordinate themselves. No need for human mistakes. Just clean beautiful code that works perfectly.

When the AI day comes, we will all be in trouble since it will touch all areas of society. But for now if you are a web analyst you should worry more about how your tools will develop within the next decade or so. To avoid being misplaced you need to figure out how to move up in the food chain. Understand how to create true value to your business which can probably be done in many different ways but keep an eye on jobs that need human creativity and idea generation. These are probably the last jobs to be replaced by a computer.

As a web analyst since you work with data you are probably in the best position to move into new fields. These new fields will probably combine the understanding of the power of data with the combination of creativity. So don’t be discouraged just keep an eye for change and you will probably do fine.

What makes a good web analyst?

A great web analyst can move the revenue needle.

I have seen a couple of posts about this subject (post 1, post 2) but most of them focus on the traits or actions a web analyst should have and take. They are mostly written from an outside in perspective. My aim with this post is to write the inside perspective, with that I mean how I as a web analyst think in my daily work. After reading this post you should be able to get into my shoes and view the world as I do.

It has been about four months since I changed company. I absolutely love my job and my new company. Since the online department is more or less brand new, we have several challenges in front of us.

If you have followed me you will know that I think a web analyst is in the absolute core of a company, at the very heart, since you can guide people to make great decisions. Being in this great position you need to make most of it. Therefore I think a lot of how I can help my company to excel to the next level. Putting my web analytics and conversion hat, glasses and gloves on you should think like this:

  • I should be on top of all web numbers. It is important for me to understand my site and the cycles that drive the business. I spend a bit of time looking a different numbers and contemplating about them from different angles. I should also control all measurement and expand it accordingly.
  • I should know the behavior of our users on the site. I think this is one of the most important points to be able to take your knowledge to the next level. I refuse to do reporting. Reporting is for dashboards and automatic reports; I should only do insights and recommendations. That is when my full potential is being used.
  • I should know what most of our users get triggered on when making a purchase. This user insight is worth its weight in gold and can only be found if you combine many different factors.
  • I should support key players in our organization to make better decisions. This might be through emails, reports with analysis, discussions, workshops etc. I think that as an individual I cannot succeed if my fellow colleagues cannot succeed.
  • I should guide marketing efforts towards the most valuable solutions. I deep dive into marketing actions to try to understand each component of that action. For example if it is email, I need to understand what the optimal form factor is, the optimal expressions etc. Each channel has its own mysteries that I should understand.
  • I should advise in all web page design creations, to get the conversion perspective in there. Every web page you create should have an aim and it should work with the rest of the site. Therefore you need to understand the various aims of each section of your site.
  • I should challenge key persons within the organization to push for new boundaries. With knowledge found in numbers one of my most important tasks is to push to explore new boundaries.
  • I should challenge partners we work with to be on top of their game. Measuring correctly gives me the power to know exactly how our partners perform.
  • I should own the conversion field. I know how we could use the web to drive more business. I use many different tools but the most powerful is AB testing. With that I can elevate the online business towards new highs.

These are the key points in my view of web analytics and conversion. My current aim is that our department should be one of the sharpest online departments in Sweden so view the list from that perspective. Every company has its own challenges and as web analyst you have to adapt to your environment. Although I think most of the points above are universal, but what will make you a good web analyst is if you in the end can move the revenue needle no matter which method you use.

How to become a web analytics dictator

I am fascinated by human nature. I love to try to understand why we make choices as we do. Everything from choosing a particular product to choose our path in life. Recently there has been a lot of talk about dictators being overthrown around the world so I thought I would go against the current and write how to become a dictator within a company; a web analytics dictator.

Of course this post is on the funny side since we all need to rest our brains from all the data.

Statistically speaking there should be plenty of people with the same traits as a dictator working as web analysts. Actually there was a study done not long time ago that concluded among the CEOs of major corporation the narcissistic trait was overrepresented compared to the population as a whole.

So how do you go along to accomplish your task? Just imagine you just landed a job at a major corporation and want to see your move up the ranks. How do you do that? How do you become so powerful that people fear the sound of your walk? How do you topple your competitors? How do you get so hated that everyone wants to revolt against you? Well, follow these steps and you will be on the right path.

Step 1, be mentally prepared

The right mental mindset is crucial. Just like a professional football team that wants to win the world cup you need the mental preparedness or you will not succeed.

You need to believe you are entitled to unlimited praise and deference. You should also believe that your will and desire is the only game in town.

Focus on fantasies on unlimited success, power and intelligence. Visualize how you could rule the company and all the praise you will get.

Like your mother used to say, you are “special” and unique, and can only be understood by other special people with powers. Believe that your fellow web analysts are not on that level.

Getting this mindset might be difficult for an untrained brain but don’t lose hope. Everyone can become a dictator with a little bit of training so repeat the three paragraphs sentences a few hundred times in front of the mirror and you will get there fast.

Step 2, strive for attention

Be like a spoiled child. Continually claim for attention and admiration. In a meetings talk a lot. Even if it is lies that come out of your mouth continue to speak unrestrained. Get all the lights to focus on you. Dominate the meetings. Use your knowledge in data. Make your arguments extremely high tech so people cannot question you. Also tell a lot of impressive lies even of events that never occurred like how you went on a fishing trip with Obama.

Step 3, use people

Other people should be seen only as an extension of the self to be manipulated and/or eliminated as needed. You should develop an inability to relate to people as people. Suck up to those in power and kick on those below. Feed important people with only favorable key data while keep the others in the dark. Try to give other people wrong data and deny it when they claim otherwise. Leave no paper trail to point back at you.

Step 4, turn off your feelings

Practice to disconnect your feelings towards others. Ideally you should have a strong inability to recognize or identify with the feelings, needs, and viewpoints of others. Ignorance is bliss as they say. Your web analytics field is the most important game in town. When other people have an opinion, bring out your web analytics guns. Prove them wrong time after time with data that has been manipulated in your way.

If you are being criticized on your data or behavior be hypersensitive. Insults (real or imagined), criticism, or defeat should be met with rage, shame, and humiliation. Do not give your coworkers any breathing room.

Step 5, manipulate data

To get what you want you will need to be able to argument for your case, and with data you can turn it either way to support your case.

You should misrepresents facts and ignore data that conflicts with your fantasy world. Based on the data you should opportunistically shifts positions and be overly confident.

The game is to be perceived as an intelligent web analyst with superb performance. You should have control of the perception of data and when you are challenged you should respond as of though your very survival is at stake.

Step 6, surround yourself with yes sayers

Until you reach that unquestionable tyrant position you will be dependent on others. Surround yourself with codependents, enablers and followers and if anyone of them challenge your authority  just discard them as they never existed, because the victim only matters in relation to how he can support the grandiosity of your dreams; beyond that he is faceless nameless and worthless.

There it is. Your six step guide. Easy as 1-2-3. Now you have no reason not to pursue this path J.

A small note. Of course following this path will classify you as insane and I highly do not recommend you to follow it. On the good side maybe you can use it to identify an upcoming dictator within your web analytics group or your company as a whole.

eMetrcis Presentation, Stockholm September 2011

I had the opportunity to speak at the eMetrics Summit in Stockholm and it was fun. I was not nervous at all, at least in the beginning, since I knew my story by heart. Although I have done a lot of presentations I have never had a presentation in front of an international audience.

On conferences I always enjoy mostly the sessions on user cases. Therefore I set my aim to present my own case based on my position as a web analyst and conversion specialist. My topic was to show how we moved from being very primitive in our optimization work towards integrating it into the business development. This path could be divided into three phases.

In the first phase we were growing, but we were very primitive when it comes to using web analytics and optimization. We did not possess the knowledge on how to use our tools in the most effective way. The ab testing world was new and we needed to gain some experience first.

The second phase, we were still growing as a business but not as strongly as before. This was the waking up phase. We had some experience with our tools and we had extracted some knowledge of our users. We also started to ask more and more questions about our business.

The third phase the growth had stopped. Management wanted me to come up with a plan to set the website on a straight course. This led to several important steps towards a new thinking and double digit increase of our most important pages and elements hence sending more quality traffic to our partners.

The conclusion of my speech was that only when you integrate web analytics and optimization in the business development process you will get the double digit increase of conversion, revenues or whatever your goal is. Many companies do not understand this and give up to early in their optimization quest. As a web analyst you need to help your business owners understand the business in several dimensions, everything from the vision to the user behavior and recommend actions to capitalize on it. Of course this process is not as easy as it sounds but it gives you ammunition to do the right things for a long time.

I have attached the presentation if you want to have a look. Download e-metrics presentation

Web analyst part of the revolution

I have been working with data in various ways for about ten years by now. I have worked both with collecting data, processing surveys, managing workshops, analyzing strategic data and now lately as a web analyst and conversion specialist. Being in the business for a while I have seen the transformation towards a more data driven society. For example the last two years the web analytics field has literary exploded. New tools and features are being introduced weekly. Companies are craving to handle their data and more and more companies are in the process of hiring web analytics people. The data revolution is hitting every sector of the economy.

In an outstanding article in the Economist about the data revolution one could read:

“Revolutions in science have often been preceded by revolutions in measurement,” says Sinan Aral, a business professor at New York University. Just as the microscope transformed biology by exposing germs, and the electron microscope changed physics, all these data are turning the social sciences upside down, he explains.

The Internet has contributed to this explosion of data as the article continues with Wal-Marts huge data sets:

Wal-Mart, a retail giant, handles more than 1m customer transactions every hour, feeding databases estimated at more than 2.5 pet bytes—the equivalent of 167 times the books in America’s Library of Congress (see article for an explanation of how data are quantified). Facebook, a social-networking website, is home to 40 billion photos. And decoding the human genome involves analyzing 3 billion base pairs—which took ten years the first time it was done, in 2003, but can now be achieved in one week.

Of course no human is able to handle these huge data sets therefore a lot of resources are being put on tools.  The amount of digital information increases tenfold every five years and the industry grows with over 10% yearly, roughly twice the software industry. The tools are getting more and more sophisticated and analytical methods are being built into the tools. For example Microsoft’s search engine Bing are examining 225 billion flight and price records to advice customers whether to buy an airline ticket now or wait for the price to come down (Farecast).

The article in Economist supports my general feeling I have had lately that the role of data interpreter, call it web analyst, BI, statistician or something else is heading towards a crossway. In the companies that are in the forefront of the data integration the need for a traditional web analyst is morphing towards a more converting type of analyst or even towards a strategic business developer.

From my perspective I see two major trends:

Trend A – More and more companies realize that data is a critical part of their business and the world spins faster and faster. This puts a lot strain on their data collection and their data use in making the correct business decisions. In this case the role of the web analyst is usually more of technical sort, handling the tools, understanding their requirements and supporting the organization on KPI´s and creating ad hoc reports.

Trend B – The data tools are getting smarter by the year. For example the field of web analytics is merging more and more with traditional BI. As the tools get smarter they are doing more and more the work of a traditional web analyst. KPI reports are handled automatically, ad hoc reports can be subtracted instantly without any special knowledge, correlation reports and basic insight is generated by the tool. In this case the role of the web analyst is either more like a statistician who handles all data input or a more evangelist like where one interpret or support business decisions.

For a web analyst both of these trends are favorable since the occupation becomes more popular but they demand two different types of skillets. In both cases the value of data lies in how we interpret our data. A combination of good tools and smart people will always be needed but the role of generalist is becoming more and scarcer, since more people will need to have special skills in various areas.

How many companies have web analytics or conversion departments today? I see this changing in the future. The BI and web analytics function will merge and the department will become a key player within the company.

As a web analyst in the future you have to consider what type of trait you bring to the table when the tools will do most of today’s traditional web analyst job. Focusing on the business value is a must but in what way? I keep pondering on how this development will go and I have no clear answer at the moment. Maybe you do?

“The data-centered economy is just nascent,” admits Mr. Mundie of Microsoft. “You can see the outlines of it, but the technical, infrastructural and even business-model implications are not well understood right now.” – Economist, Feb 25th 2010.