Death by Google
Google has killed the web analytics software industry with the release of the new version of Google Analytics. The new version was released just under two months ago and is simply a quantum leap above any other analytics product on the planet.
In my opinion, Google Analytics does for the web metrics industry what the Google search engine did for online search: it kills everyone else off.
Google Analytics version 2 is not revolutionary. It does not extend web analytics software by providing new forms of analysis. Neither does it extend our understanding of websites by offering new approaches. What Google has done is simply take every feature in every product on the market and put them all into one system, and then make it available for free.
I am surprised by the range of features Google has added. I would have assumed some had been patented by the companies that created them. I can only conclude this is not the case. The range of features Google has borrowed from other products suggests the web analytics software industry managed to do 10 years of research and development without registering even one patent. This must be unique in the history of computing. If Google has stolen patented ideas, then I can only conclude they simply don't care and will rely on their massive cash reserves to sort it out later.
I say this as someone who, until this month, ran a company that produced web analytics software and directly competed with Google Analytics. No more. There is simply no way my organization can produce the range of features Google offers and make them available for nothing. We will keep the consulting arm going but use Google Analytics as the reporting system. Our techies don't want to see the product die completely, so they plan to convert it into Open Source. While Google Analytics is world class, we can see the need for there to be at least one Open Source page-based tracking product, and currently there are none. If you’d like to help, please email me and I’ll put you in contact with the team.
Under these circumstances, it should be obvious that I am not an automatic fan of Google Analytics. As a professional web analyst, I thought the previous version was strictly amateur. If you've read some of my previous articles, it's pretty clear I'm not a fan of Google either.
I have been converted to Google Analytics version 2 purely by the strength of the product. It is not just the range of features that is impressive, it is the integration and flexibility.
The way in which the system has been put together demonstrates an extremely sharp understanding of how to analyze websites. Web analytics systems can, and do, produce hundreds of metrics. This can completely overwhelm anyone. Experience teaches us that the secret is to focus on what really matters. We have learned that a limited number of measurements are critical key performance indicators of visitor behavior. In the industry, these are jokingly referred to as the "key key" performance indicators (or KKPIs). These KKPIs enable you to successfully see if a site is appealing to a particular audience. These are average pages per visit, the average time on the site, and (most importantly) bounce rate.
The importance of these numbers is determined by what percentage of the total readership any given group constitutes. A bad number for a tiny subset is rarely worth the money or time it would cost to improve. It is also useful to know what percentage of this group are repeat vs. new visitors. This is because most sites get their income primarily from one type or the other. In addition, well-designed sites have different layers of content for each type.
What Google has done is make these five numbers consistently available for almost every report. Next, they have taken every measurement that is technically possible to get with today's technology and allowed you to cross-reference these critical measurements against them.
For example, I can compare bounce rate by screen resolution to determine how my site design works on different screens. This is an important issue at present as designers push people to drop support for 800x600 so they can do more visually exciting work. I was able to look at this issue for a client recently with Google Analytics and show them that failing to support this resolution was costing them $100,000 a year in lost sales. Small percentages add up to big numbers on busy sites.
Coming from another angle, I can review bounce rate by country or city. This will tell me how my site appeals to different regions. In fact, I can compare bounce rate by keyword, by search engine, by affiliate site, by ad campaign and probably by eye color for all I know.
Throughout all this, Google Analytics shows me how the numbers for the group I am looking at vary from the overall averages for the site. This type of reporting enables me to identify which visitors are in need of attention. All the tables are clickable so that I can instantly drill down on the elements that stand out. For example, I recently analyzed the performance of a tourist site's listings in travel directories. I was able to drill down on specific directories and see which pages and descriptions were working and which were not. Within the same directory, I could see some listings that had a bounce rate of 9 percent and others with a bounce rate of 70 percent.
It is clear that Google Analytics supports the process of exploration extremely well. It is rare to enter into a site analysis with a complete picture of what you need to know. You may have specific activities or groups that you are keeping watch on, but exploration – jumping from group to group, drilling down, comparing – is the only way to discover the unexpected.
The other key elements of web analysis are present in Google Analytics 2.0 at the same level of depth and usability. A dedicated ecommerce section allows you to track conversion goals and display the conversion funnels pioneered by WebTrends. Simple code is provided for your shopping carts to enable financial analysis of sales if you wish.
Site design can be assessed via overlays that appear over your website and show which links were clicked and how they contributed to conversions. This system does deceive a little, however. It is not really analyzing each individual link. If multiple links or images go to the same page, they will appear individually analyzed, but in fact they are all treated as one and the same.
Management reporting is also catered for extremely well. Most people use web analytics not to manage and improve their site, but merely to create reports they never look at for their managers, who also never look at them. Google makes this pointless process very easy. You can create mailing lists so reports can be emailed, create user-access accounts so people can see the reports online and schedule regular report runs.
Thus in a few minutes, you can generate a complex reporting schedule and information flow that will fool the organization into thinking it has a handle on its websites.
While Google Analytics is a truly remarkable creation, combining all the analysis used in the industry today, there is room for improvement. There is absolutely no recognition of the existence of any form of pay-per-click advertising, except, of course, Google Adwords.
All traffic coming from Yahoo! PPC ads will be reported as coming from organic (search engine) listings in Yahoo. Yet there is an entire section devoted to Google AdWords analysis. Many of the more obvious (and useful) drill-downs are not present. I cannot, for example, drill down on my landing (or entrance) pages to see where people came from. This is a major gap. If I want to improve performance, I need to know where traffic came from and what keywords they used if they came from a search engine.
But despite its failings, the overall range and flexibility of Google Analytics, combined with the price (free), leads me to expect the new version to totally dominate the market and drive most competitors out of business. You need an extremely good reason, or three, to continue staying with any other product.
If there is to be any future in web analytics software for any competitor, that company will need to rapidly expand the scope of reporting available and seriously enhance flexibility and drill-down capabilities.
Industry consolidation is sure to follow, and I expect WebTrends to be one of the few companies with the pockets to pursue such a strategy. It is surprising that Microsoft has not produced a product to compete. They have a five-year strategy to produce a search engine to compete with Google, and they understand the importance of analysis and related tools to promote the appeal of their products. In addition, they have a complete range of web server products, yet no web analysis tool. If Microsoft is not actively talking to web analytics software vendors now, they should start soon (and Bill, if you read this, I just happen to have one I'm not using).
The shape of the web analytics software industry has greatly changed with the new version of Google Analytics. In my view, everyone else is dead, they just haven't stopped moving yet. If you haven't installed Google Analytics on a site and devoted some serious time to it, you're already part of the previous generation.