Evolution of GPS

Mark Evans talks about the mainstream-ification of GPS in his recent post. As noted, GPS technology has been around for 30+ years but it is only until the past couple of years that it's really started catching on and becoming a mainstream technology. 

My first GPS experience was in the mid 90's when I was working for a GIS (Geographic Information System) software company. It has been quite interesting watching this industry grow and going from being a novelty where people would say "so what can you do with it?" to them saying "let me show you the directions on my GPS".

So what has been the catalyst for this industry. I firmly believe it is due to data availability on a couple fronts. Firstly, up until 2006 civilian GPS suffered from a built in selective availability (SA) feature. With the systems origins in the military, a random error was introduced into the system for apparently security purposes. This meant that for example anywhere in the US you'd have an error margin of about 10m in any direction.

In 2006 this was disabled (or set to 0) so finally it was possible that a $200 device can pinpoint you down the center of a street. Prior, it was difficult to explain to a user why the device was placing us in the middle of a field when we were standing in our backyard.

The second and probably as important catalyst to the growth of this industry is general data availability. Introduce MapQuest, Google Maps and numerous other available data sources, consumers can now see where they are on a map. 

Map data is expensive - very expensive to collect and maintain. Thus there was a period throughout the industry where any licensing of data was extremely expensive as producers were very protective of their data. Canada was particularly bad in having a cost recovery attitude to data and the US (starting with the USGS) was a pioneer in making data available to the public.

One of my greatest frustrations when working within the Geo industry was the lack of data. I am a firm believer that providing data and making data accessible at a very reasonable cost has much greater benefit. Data on it's own is quite dull until you can wrap it into some application. Consumers and users in general relate to data through an application. But the moment you restrict access to data, building an application to access it will be hindered as well. The end result, no data and no applications for consumers.

Like I said, my first GPS experience was almost 15 years ago though my parents got excited about geo-technology and map data when they downloaded Google Earth. Since then they purchased a GPS unit for their car and tell all their friends about it.

This same concept applies to any other data intensive industries including web metrics. Right now we are in a transition stage. Most trusted web analytics are buried in corporations taking a cost recovery approach though with a number of initiatives, there is a trend to opening up and making much more data available to users.

The problem with surveys

Surveys are broken. It's a bold statement but I'm not sure if it's far off from being true. Surveys are broken in part due to a lack of innovation in the industry. Market research and the survey related industries suffer from a lack of technical innovation and are generally engineering poor.

I would stand to argue that surveys have changed little over the past 50 years. Sure technology has made it somewhat easier and less expensive but let's take a look at a few changes in the survey as we know it.

Taking a look at some traditional survey techniques, we can start with the mall survey. The premise is that if you intercept people where they conduct a primary activity (shopping) that you'd get a valid sampling population. The basic survey mechanism involved an individual holding a clipboard to ask the unsuspecting shopper as they pass by if they would like to participate in a survey. If they said yes, the interviewer would ask a few basic screening question and if successful, conduct the interview based on a paper survey with question and logic notes. This method is slow, error prone, time consuming and expensive, though still a valid method.

Telephone polling changed the industry since the adoption of the telephone was near 100% in some areas. Using random digit dialing (RDD) a truer random sampling of individuals could be done and all you needed was a room full of phones and interviewers. Today this sector of the industry is facing many challenges due to do-not-call lists and the rise of cell phone use.  Still the survey hasn't changed much. In this case the survey was still mostly the same in either paper format or CATI.

The CATI (Computer Aided Telephone Interview) system added great efficiency to the telephone interviewer. Essentially a survey would be programmed and the interviewer would read off of a computer screen. This system supported basic survey logic (conditional skips) and computerized data collection.

Welcome the Internet and the CATI or paper survey can now be formatted in HTML and presented to an individual online with checkboxes, radio buttons, text boxes - all with logic, nice formatting and drastic cost savings.

Up until this point, the major breakthrough has been in the efficiency of deploying surveys. Online technologies drove the cost of implementing a survey to near commodity levels so to reach 500 or 1,000 individuals become very cost effective.

Though my real disappointment has been that the basic survey has not changed. The paper survey used by the mall or phone interviewer really hasn't changed much when it went live on the Internet.  The effect has largely been a direct translation of the paper survey to the web browser screen.

Surveys or market research in general has a fear of change. Any change in implementation creates fear and concern in validity of data. This fear has stifled innovation in methodologies and implementation. Why not challenge the boring typical linear survey? Sure there are those who are creating surveys with silly flash to flip pages or select items off of a shelf.

In my opinion the value or advantage of doing research online is being drastically under utilized. Take for example email panels. Instead of being innovative, everyone is still trying to build these massive 1,000,000+ email panels. I won't go on my feeling about panels, that's for another time.

Though there is some potential of advancement. Behavioral research is getting a fair bit of attention, and a few companies like Brain Juicer, UThink, comScore  Peanut Labs and HotSpex are getting some recognition for doing innovative things.

Online research is a sector that has been growing incredibly fast and is quite healthy. Though, due to this lack of innovation there are many concerns in it's direction. Some major firms are even starting to pull away from using online research due to the rise of professional survey takers, poor methodologies, weak panels and questionable sampling techniques.

Nearly 10 years ago, we were all trying to justify online research to clients. Today just about everyone is online because it's the place to be. The next generation of research is not far around the corner. It really only takes a little innovation to lead to big changes.

Is Mozilla to enter the metrics game?

There is a lot of activity in the realm of web analytics with the likes of Compete, Hitwise, comScore, Alexa, and Quantcast. The latest announcement with Mozilla potentially entering this game can make things interesting.

Apparently Mozilla is working on an internal project simply called "Data" where they are proposing to invite the 170 million FireFox users to opt-in to anonymous data collection. Up take would only need to be a small fraction of the 170 million and the worldwide distribution is even more impressive.

One of the first criticisms that everyone will hear is that FireFox users are different from IE, Opera and other users. Thanks to comScore, this difference has already been explored in their Firefox vs. Internet Explorer post. Based on the study findings, a greater proportion of users are male, are more likely to have a household income of over $75,000, and younger than the average internet user.

These known differences are easily accounted for and further, are key demographics that are hard to reach. I would argue this to be a benefit and not a criticism.

Back to what I've been saying for quite some time, web metrics require a lot of heavy lifting, particularly for companies who take the brute force approach. What is interesting and I believe can really change the landscape of web analytics is that this would be the 'other' thing that a company Mozilla can do by leveraging what they already do well. The other analytics companies are reliant on providing some value proposition to users. The problem here is that perceived value changes frequently so keeping value at a minimum level is a constant challenge.

Mozilla has the benefit of already having a great and loyal community and I wouldn't be surprised if the uptake would be quite decent.

So for everyone else in the game, things will only get tougher.

Another great reason for being part of a start-up

It's tough and anyone who's done it will surely agree. It's tough enough that if it wasn't largely driven by the passion of it's founders, employees and even external investors (if you've been fortunate enough), the chances for success are very limited. 

Max Kalehoff has a great post on Why Passion Matters in a hyper-competitive market and outlines a few key places in business that are sensitive to passion:

In my experience, there are a few places in business especially sensitive to passion:

• Listening and understanding your customers and the market.
• Innovating based on your market insight and intuition.
• Building your product with quality and speed.
• Ensuring the highest aesthetic and usability.
• Refining your product over and over and over again, until it’s better and better and better.
• Paying attention to all the details and signals that comprise the experience.
• Inspiring your employees, customers, investors and other stakeholders.
• Engaging and collaborating with customers.
• Fixing things quickly when they go wrong - and then making them far better.
• Using your product yourself and recommending it to friends because you truly believe it’s the best.

Businesses with passion tend to excel in these areas, while businesses that don’t tend to just get by or break. I know - this is all obvious. But the irony is that most businesses and brands I encounter come up short.

Maybe this is part of the reason why start-ups are so supportive of each other. When you meet someone passionate about that they are doing, you naturally support them. Are you passionate about what you do? 

A great reason for being part of a start-up

A quote by Dick Costolo that rings very true and close to home. One of the great reasons for being part of a start-up.  

"In the beginning, you are the most limited in terms of resources but the least limited in terms of range of motion. It's easier to innovate and change directions in the early days, but it's harder to do nine things at once. This is probably the founding team's favorite part of the lifecycle of the company. Everything is possible, and you can really focus on building the most awesome product/service possible"

Greatest product of all time . . . .

R2-D2 DVD Projector has to be the coolest product of all time.  Be sure to watch the entire clip . . .

LinkedIn to start market research polling

Tom Anderson talks about how Facebook and Linkedin are entering the market research game. Linkedin seems to be the soon-to-be latest entry, though as Tom points out polling is not robust enough to provide real consumer insights.

Polls are the logical entry for any social network due to their simplicity to implement. Further, any more thorough true market research provides a much higher level of maintenance and sales cycle between the audience/respondent and research professional.  This is in part due to the historical high use of incentives for market research and as a result, becomes an administrative nightmare. 

Polls are single questions, no logic, and little burden for the respondent - thus no real value proposition is needed. I would argue that if you get people at the right time, there is an incentive for respondents to provide insights beyond a simple poll. We just need to get outside of the 'email panel' or 'registrant' way of thinking. More to come on this . . .