Standard and non-standard metrics
If you are reading this blog you probably know metrics like Unique Visitors, Visit, Pageview, etc. Those metrics are Standard metrics, which means that the way those metrics are being measured is based on a definition made by a recognized organization (the Web Analytics Association), allowing comparisons among different projects. Actually when someone tells you “I have 10.000 unique visitors” we all know the meaning of that (10.000 unique browsers being measured by a particular cookie).
Standard metrics are very important, actually the Web Analytics Association have a committee that works very hard trying to define those standards (The standardization committee), and very useful as well. Standard metrics are focused on external analytics, which means the analytics that relates your company/project with other companies/projects. So with standard metrics you can do external analytics for control, benchmarking (I don’t like so much external benchmarking but it does exist), etc. This means that you shouldn’t use standard metrics for Internal analytics ever? No, no, I’m not saying that. I’m just saying what’s the main focus of those metrics, but if any of those metrics is useful for you for Internal Analytics then just use it.
On the other hand you have the non-standard metrics. The non-standard metrics are those that you use for internal measurement and benchmarking. In this case it doesn’t matter that the metrics is not standard because you will only use it as a sensor that will allow you measure your own company’s performance.
Why I’m saying this? Because when looking for “sensors” to measure your performance you have to be creative, creating a metric able to capture the specific information that you or someone from your company will required.
So, when looking those metrics to measure your internal performance forget about all what you know and think about the ideal one and call it in a way that the users of the metrics will understand and don’t forget to write the description and formula of the metric everywhere the metric will be shown.
You can also download the Standard Metrics from the Web Analytics Association
Dad is back and will lift your conversions
It’s seems like nothing changed for so many years…take a look at this post “Desired Advertising – DAD – Is that possible?“ and tell me if it is not actual? It is, and it is for the reason that the base of it it is that advertising should be planned and executed as a service it self and not as an stupid and intrusive manner of stolling peoples’ attention.
With the above in mind and always following the Web Analytics Association’s Code of Ethics, Imagine how better could be your clients that you contact them considering their interests and needs, following the Dad model…
Do you think it is something sophisticated for your skills? It’s not…at all.
The only thing you need is an Analytics solution that allows tracking individuals and having a registration or login process in your website. To those that accept the tracking cookie you will have the information of the interest that that particular person has in a very specific moment. So you can create segments based on behavior and/or demographics and/or attitude and contact people with, for instance, a newsletter, that adds value to them. You can send them promotions based on what they are looking to buy, or events that could be of their interest among other benefits.
I test it with Yahoo! Web Analytics just because it was the closest one I have (and it’s a very good one!), but you can use any other with individual tracking. Test it and please let me know if you get any lift.
“Underestimate tools” is a great first step for an Analytics professional
In more than 14 years working with internet information I’m seeing the very same situation happening over and over again. The first contact of a person in the Analytics world is related to tools. I’ve being talking a lot about it, we all know how hard is for the average civilians 
separate analytics tools from the practices. We committed a huge error branding tools with the same name than the practice it self (same problem than the CRM peoplecommitted with their practices and tools).
So, when someone starts walking in the Analytics way their main focus is on learning using tools and later on learning how to implement tools (that’s the long term dream). The true is that Web Analytics (I added Web because the Web Analytics Association talks about Web Analytics in their post “Optimizing the Online Business Channel with Web Analytics“) refers to the collection, analysis and reporting of Web site usage by visitors and customers of a web site. So, tools are just a part of the process, but the most important part in the analytics process if getting insights from huge quantity of information to improve the decision making and business results. So, is it important having the tools’ know how? Yes, however I wouldn’t ever reject someone that wanted to work in my company Intellignos just because he is not a professional in any particular solution. Learning how to use tools is just about reading Helps or a FAQs, any person in the world can do that…is not different as learning how to use excel…but you wont be a good analyst if you don’t have the analytical mind that can find patterns and trends in the data that can be use for making efficient decisions. On the other hand tools are alive, they are constantly changing and getting improved so are you gonna base your core competence on something that could change or even disappear in the near future? Come on, is not about how a particular tool should be handle to get a particular set of data, it’s about having in mind what information do you need and then looking how should you handle to tool to get it. So, with the above in mind the best advice I could share with you is…if you are walking your first steps in this industry look at your Web Analytics tool at its face and tell it “I don’t care about you, I don’t need you, you are just a tool”
The value of an smile – Analytics allows you to counting the uncountable
If you only read the first sentences you will probably think that this post has nothing to do with Analytics (that may be good for some people, but let’s think that since this is an Analytics blog, that is something bad), so let me invite you to read the entire post.
Ten days ago I was having vacations after two years. When you don’t have vacations for such a long time you have the expectations that everything goes perfect.
The second day of vacations my kids asked me to visit a Dinosaurus thematic restaurant…. I went to a restaurant with my family. It hasn’t the best food in world nor the best prices, but the idea was having a good time with the kids.
So we arrived to the restaurant and asked for a table, the answers was “your table will be ready in 45 minutes” (first metric). My first reaction was, let’s go to another place but my kids were not agree with that idea, at all, so I had to stay there for 45 minutes. I remained inside of the restaurant since outside was cold, until someone from the restaurant told me, in a very bad manner, that I shouldn’t stay there but in the gift shop or outside. So I waited for 45 minutes in the gift shop (can you imagine how hard is staying in a gift shop with two kids for 45 minutes?) and came back to the “front desk”. After waiting like 5 minutes (second metric, and we are building our Analytics model ) count someone told me, pretty ungently, that they still don’t have a table for me. So I waited for like 10 minutes (third one) and finally a very gentle waitress came to us with a big smile and invited us to have a sit .
I told her that before ordering I wanted the complaints book. I told her, you are the first person smiling in this restaurant. I understand that everybody here is working hard because the restaurant is full but that is not excuse for being rude or not gentle. We went there just for having a good time. She apologies for the situation and offered me to call the manager. I accepted. The manager was very gentle and always smiling. She said that probably their employees were pretty stress for their work and they “forgot to smile”. She told me that that is not an excuse, so they invited us to have a great time and to forgot about the past. We accepted, and finally we had a great time there.
Before leaving the waitress brought the check with a 20% discount for the inconvenience. It was a good way to say, the service was lower than the expected, so we reduce the price. The lower the price, the lower the expectations (price= product or service expectations – product or service provided). But what I kept thinking was, at the end of the day it is possible to measure uncountable things.
My complain was simple, nobody was smiling there. Even when I wait 45 minutes, even when they haven’t my table after the 45 minuties, the problem was…nobody was smiling. They understood that the lack of smiles in the restaurant worth 20% of the service. So, just imagine that the restaurant begins using that metric as one of the mains KPI’s and explain their employees the importance of smiling…”Everytime you don’t smile we lose 20% of the ticket, and in some cases even a client”. Actually is even a great deal for the restaurant paying like a performance fee. “If you smile and the client is happy, we pay you a and additional 2% of the ticket value to your salary”. Remember the quote “tell me how you measure me and I’ll tell you how I’ll behave. Measure me in an illogical way and don’t complaint if I behave illogically“. Well, the “Smile rate” is very clear. All the people from the restaurant can understand the relationship between both variables (Income & smile) and if that is clear it’s easier to accomplish the followed results. Wouldn’t be awesome do an Anova model Analysis with the variables Income and Smiles? Never say this is uncountable as the first answer, be creative, every flow in your system is countable. If you don’t get the answer, probably you only need to dig a little more.
Constraint management vs direct attribution
The last month I attended eMetrics New York, probably the most important and relevant event in our industry, which was incredible well organized by amazing people like Jim Sterne (a little history of it) and with a range of topics to satisfy even the most particular taste.
I wasn’t able to listen all the presentations because I was at our booth, but the average quality of presentations was pretty good. The only thing is that, apparently, we have a new fashion word, let me introduce you the best, the most important and relevant word from the Online marketing industry…with you…”Attribution“. Now, seriously…I’m surprised on how people try to identify the attribution to a particular object of their system. It’s like thinking that all the merits for a goal in a soccer game are from the person that sets the goal…come on…
Take a look at the following presentation…
The problem of direct conversion attribution
So my suggestion is managing restrictions instead (Yes, the Eli Goldratt model). Why? Take a look at the following slides:
So, if we can identify the system interactions and based on data series understand the dependence among variables why don’t creating a model that allows you to identify the percentage of conversion or sales attribution from that each object of the system? Let’s leave this for another post…I’ve been working on it
Analytical Claustrophobia
The possibility of having two projects based on similar systems are almost null. I mean, even when you can have an eCommerce that sells Golf stuff and your competitor has the very same thing, both are not gonna work in the same way. All the elements of both systems will not interact in the same way so the experience will not be the same and the way the elements interact in order to reach the system’s objective will definitely not even similar.
With the above mentioned differences is easy to understand why we need that the information platforms (Website Analytics, Social Media Analytics, etc) have analytical flexibility. Every time I use an information platform I feel an Analytical Claustrophobia…I mean, I need to use the information with some freedom but instead I feel that the current flexibility it’s still not enough.
I know, the more flexible the solution the less easy to use. I also know that it is important that everyone in the company can use it and not every one should know how to run database queries, but come on, you can definitely have a very easy to use front end and ALSO a database access that allow you to run queries.
The more information, the more information freedom we need to get valuable and actionable insights. The problem is that vendor a not focused on that because the people that are supposed to analyze the information are constantly asking for easier to use platforms, and vendors instead of getting smarter and find the way of satisfying both requirements they just go for one of them.
You can also say that allowing people to querying the database increase the platform costs a lot, but…who cares? I mean investing in information is much more simple than what you can imagine, it is about “how much does it cost” and “what results can I generate with it”…you can say, how can I be sure if I’ll able to achieve those results and the answer is, you can’t, but at least you will be making decisions under less risk…actually you will be able to know the level of risk that you will be assuming. Since some platforms are free it seems like all of them should be free but the price must be related to a benefit, it doesn’t seems very smart using standard or structured information that doesn’t answer your questions just because you don’t wanted to pay for that…
The more I know information platforms the more I love Excel and Databases…
The real value of Text Mining and Sentiment Analysis for Analytics proposes
We talked a lot in this blog about the different type of information and the value of each of them.
One type of information is “Behavioral information“, useful to answer the “What are people doing”. Even when some people mix up Analytics with Website Analytics, the true is that Analytics it is much more than that. You can answer “What are people doing” with behavioral information, but not “Why they are doing what they are doing”. That it’s a huge problem…I mean, at least if you want to do something with the information like make a decision that drives your company closer to it’s objective. The coolest information to know “Why the people are doing what they are doing” is attitudinal and not behavioral. For instance if the people is not buying your products, that doesn’t mean that your products are not good, or your campaigns are not well targeted, or any other thing, that is something that you know with attitudinal information but not with behavioral. If you wanna know why they are not buying your products just go and ask them…why assuming the risk of making decisions with inferred information? Exactly…there is no reason…
Behavioral is cool but it laks of own value by it self, it requires to be used with any other type of information. We have the very same situation with the other type of information like Attitudinal. Before knowing why people is doing something you should what are they doing. So, when you mix both types the magic appears and you feel like having an epiphany is something possible and is not so far away.
However, even when you feel that the above is something clear to you, the possibility to measure all the attitudinal information required to give relevance or to answer the “WHY” to the Behavioral information is pretty new. I mean we have polls and surveys insite since…well, we always had them, but surveys and polls cannot capture the total reality of the behavior unless the survey allow open answers, but if so you have another problem…make that information useful or turn it into qualitative information, otherwise you have like a short list of things that you are measuring with the survey, so the possibilities of explaining the “WHY’s” are almost null.
Having said that is easy to understand the importance and power of the Text mining and Sentiment analysis for Analytics. However I’m not sure that we are using that technology in a properly way. Every day I receive four or five emails about a new “solution” for measuring qualitative information about what people are saying online and it seems to me like we are all accepting some kind of freaky schema where if you have more tweets and likes then it means that your company is being successful, so companies that develops technology for “listening” what is being said delivers reports that are focused exactly on that, but unless someone pays you for size of you ego your company won’t get any benefit from this.
As a person that begin as an analyst and a researcher I must said that I found lot more value on this technology than measuring likes and tweets. I mean, I found extremely fabulous using Text Mining and Sentiment analysis for Research Proposes more than anything else. I’m not yet talking about “Social Media Analytics”, “Social Media Monitoring” or “Social Media Engagement” platforms, I’m just talking about the potencial of this technology for business.
Have you ever heard about a person called Michael Porter? He talks about a model called value chain, almost every person in the world that studied management knows exactly what I’m talking about, if you don’t just take a look at Michael Porter by Wikipedia.
Mr. Porter (not Potter…don’t be such a nerd person) says that the margin that a company gets from its operations is the result of:
1. Primary Activities: Inbound logistics, operations, outbound logistics, Marketing and Sales and Post Sales support.
2. Secondary Activities: Firm Infrastructure, Human Resource Management, Technology and Procurement.
The better you work on those areas, the better your company goes. The better your company goes, the better margin you gets. So, if you want that your company gets more margin you can just provide each area with the possibility to listen what people are saying about them and let them work on give people, what people want it. Isn’t it soooo sexy? I mean, who cares on how many tweets or likes or even mentions you have when you can listen, in real time, what people are saying about yours, or your competitors’ products?
Back from Google Analytics Summit 2011
It was a very intense week at the Computer History Museum. Jesse Nichols, the tireless Program Partner Manager for Google Analytics and Google Optimizar, did a great job with his team, arranging an event for more than 500 people from all over the world in in a very exquisite way.
I’m not sure what it is and what it is not under NDA so I wont go in depth about new launches and developments. The most important things I wanted to mention are:
1. The first day was basically an overall presentation about the platform improvements. All of them where Woowww… not all of them really useful, but the ones that does are awesome…at least are the ones I was expecting for such a long time.
2. The second day, Thursday, we went in depth about the Partner program and Technical aspects of the current and future launches. I had the honor of being part of the Panel with Justin Cutroni (Cardinal Path), Timo Aden (Trakken) and Russel Sutton (ConversionWorks), moderated by the awesome Sophie Chesters (Google).
At the end of the event it was a Web Analytics Thursday with a Rock Band, networking, drinks and food…anything else? Oh, yes…amazing people from all over the world willing to talk about their experiences without constraints. See you there next year mates!!!
Google Analytics changed the session (visit) definition
Last week Google Analytics changed the way the session (or visit) is calculated. Why is this so important? Because your past information will not change, just from now on, ergo when you analyze trends you will identify a variation in you reports comparing before and after the new visit definition, so the first thing you have to do is loading a new event note in your Google Analytics platform so when you analyze the information in the future it will be easy to understand the cause of the variation.
Last visit or session definition:
More than 30 minutes have elapsed between pageview for a single visitor. No changes.
At the end of a day. No Changes.
When a visitor closes their browser: It changes by “When any traffic source value for the user changes. Traffic source information includes: utm_source, utm_medium, utm_term, utm_content, utm_id, utm_campaign, and gclid”.
How is this new session definition different?
Basically if the person leave your site and come back from a different source (any of the above mentioned values changed) this will count as a new session. According to Google, this change will just generate a variation of not more than 1%.
Important Note from Google, remember add a note in your platform:
Update 8/17/2011 2:10 PM PST:
“We identified an issue responsible for unexpected traffic changes following our recent update to how sessions are defined in Google Analytics. A fix was released at 2pm PST Tuesday August 16th”.
The issue affected some sites using the following configurations:
1. If a user comes to a customer’s site with a space in some part of their traffic source data, then revisit the same landing page during that session by refreshing the page or later pressing the back button, a new session will be created for every hit to that page. (Clicking a link elsewhere on the site that leads back to the page should not matter.)
2. Google Analytics implementations using multiple trackers (an unsupported configuration) are also affected when a space is included in the traffic source data. These sites will see fewer visits from new visitors, and more visits from returning visitors (with some variation due to different implementations).
Dennis Mortensen’s Visual Revenues secures Series A seed founding
Yesterday I’ve received great news from Dennis Mortensen (Remember, the guy from Indextools, acquired by Yahoo!, today Y!WA), a guy that I admire for his great and pragmatic analytical mind. His company Visual Revenue secured Series A seed founding for USD512k led by Lerer Ventures, SV Angel, Kima Ventures and NYC Seed and joined by 10 individual Angel investors. The investment will be used to accelerate product development of company’s Predictive Analytics Platform for Media
Visual Revenue offers media companies a first-of-its-kind Front Page Decision Support System for online Editors. The solution, which can predict the performance of a piece of content about 15 min. into the future, provides real-time recommendations on what content to place in which position on a Front Page and for how long. Taking these recommendations results in a dramatic increase in visitor engagement, content relevancy and revenue, whilst adding a cost saving opportunity at the same time.
Some comments from the main actors:
“We’ve seen first-hand the growing importance of data for publishers and jumped at the opportunity to invest in Visual Revenue”, said Eric Hippeau of Lerer Ventures. “In a fast paced newsroom, Editors need recommendations on what content to place where, not more data to interpret. Visual Revenue’s unique ability to talk directly to the Editor is a huge win.”
”Online Editors have access to a lot of data but this typically paints an unclear or misleading picture” said Dennis R. Mortensen, CEO and founder of Visual Revenue. “We are leading the way by providing online Editors with advanced, yet simple, decision support tools they need to take informed, real-time content placement decisions that impact the future.”
“I am extremely happy with the group of investors we’ve brought together and their media industry knowledge. This investment will allow us to accelerate the development of the platform and further expand our offering,” said Mortensen.
Visual Revenue is platform independent and provides real-time recommendations on content placement for a publisher’s web front pages, mobile and tablet editions and their syndicated feeds. Visual Revenue works with all Front Page layouts, requires no change to existing processes and has an extremely lightweight integration methodology.
So, congratulations to Dennis and his team!
