Are you trying to understand your website traffic?

I am a big fan of a service called Quantcast

From their website:
Quantcast is a new media measurement service that enables advertisers to view audience reports for millions of sites and services to build their brands with confidence.

This service is incredibly useful for the public sector.  Here are just a few interesting uses:

1) Understand your own audience using the demographics tab
2) Understand how your audience overlaps with the audiences of other websites.  This allows you to identify websites you might want to target for collaboration (our company, GovDelivery, allows public sector clients to collaborate and cross-promote with each other so we use Quantcast to see audience overlaps between different websites)
3) See the frequency of your visits.  Quantcasts classifies visitors between addicts, regulars, and passers-by.


In my experience, the service undercounts website traffic by 30-60%, but it is still directionally accurate, and it gives you excellent information about your website and the user traffic of others.

I welcome any feedback on how you make use of this service and how the numbers you see from your own analytic tools compare with what Quantcast is telling you about your website.

Recently I came across an interesting paper from Princeton University titled "Government Data and the Invisible Hand." The paper proposes ways that government can do a better job making information more available to citizens. It's a good read and only ten pages: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1138083

The paper makes a strong statement that government web sites have failed to provide government data to the public, and that private industry can do a better job. I don't agree with this view, but I believe that the paper makes many insightful observations and provides useful direction to policy makers as well as government professionals involved in planning the future of the web and e-government within individual agencies.

 

The bottom line from my standpoint:

1)       As the authors of the paper suggest, government should put far more energy into publishing information in highly-structured formats so that the information can be repurposed for distribution by other organizations (private, public, non-profit, etc.).

2)       The authors think government should make structuring content the top priority, but they discount two areas where the government Web manager plays a critical role.  First, the government Web manager acts as an editor raising awareness of the most valuable content.  An agency may have 50 publications on its website on a certain topic, the government Web manager can feature the publication that is most current and has the most potential public valuable.  Second, government agencies are increasingly acting as citizen/customer relationship managers by learning what citizens are interested in, recommending complementary services and content, and creating ongoing ways of receiving updates direct from the government.  In my opinion, the biggest obstacle for a citizen trying to get content or services from the government is time.  Government Web managers are increasingly adept at making interaction with their agencies more efficient.

 

On the web, we value Amazon's ability to create an efficient and pleasant shopping experience over its ability to offer lots of different products.  Amazon features the most interesting (or maybe profitable) products, allows users to see suggestions based on what other users have purchased, and gathers information from users so it can notify them when new information and products of interest are available.  In working with federal agencies since 2003, I've seen enormous strides in how government Web managers understand their audience and play the role of trusted editor.  In addition, we've seen many government websites deploy strategies to use the website as the hub of the citizens relationship with the agency so that it serves not just as a content repository, but also as a place to signup for email updates and submit questions and comments.  In full disclosure, my company, GovDelivery, is a private sector organization that serves many governments with a digital communication platform.  This might make me bias, but it also gives me insight into how government websites have leveraged at least one platform created in the private sector to offer tens of thousands of different topics for subscription and send over 50 million digital messages out every month.  This seems to be the best of both worlds.  Government controls the content, the message, and the citizen relationship, but is able to leverage the best technology from the private sector.  Should GovDelivery fail to continue to innovate and meet government needs, another company will step in to fill the new needs of the government (this is exactly how companies move between technology providers over time). 

Highly-structured content on government websites has the benefit of making the websites easier to manage and facilitating easy deployment of new services that leverage the content in different ways.  This is a tremendous benefit, but expecting the private or non-profit sectors to manage and improve this type of service without a direct connection to the government agency is, in my opinion, like telling Amazon to outsource its homepage and its customer database.

 

Complaining about the TSA is like complaining about the weather in Minnesota.  Even when things run well, we assume bad things are on the horizon.

TSA is playing a cruel numbers game from a PR perspective.  They must inconvenience a lot of people to capture or deter a small number.  Not an easy job if you are trying to make friends, but they are taking many steps to be more effective.

I have flown around 2 million miles since 1997.  Currently I take 4-6 flights per month (mostly from our corporate headquarters to DC, but occasionally to see clients in the U.K. and business partners around the country).  I have a theory that everyone is a little afraid of flying and that is why we all look for the worst in air travel.  What I find in reality is that security is reliably efficient across the country and that I am much safer in an airport or airplane than I am in a car.

The main reason I'm impressed with TSA is their approach to communication.  They are proactive at getting the word out and incredibly open.

Three examples:
1) The most open blog I've seen where the whole world can comment (and sometimes rant) about the TSA and air travel in general.
2) Every time they have a rule change, they immediately get the word out through any and all channels as well as through incredibly easy to understand (e.g., "3-1-1" for liquids)
3) They use email and RSS alerts to get the word out on new information.  (Full disclosure: my company, GovDelivery, provides the email alerts.)

Finally, they now put up posters saying what they have accomplished over the previous week.  These posters make the public aware of the work TSA is doing to improve security.   When I see what they are accomplishing, I'm more agreeable to taking my shoes off.  Plain and simple: when citizens know that government is getting its job done, we are more cooperative and appreciative. 
 



When we organize websites and determine what kind of content we're going to offer for subscription when communicating by email, there is a common question that needs to be answered.  Do we give people what they want or let them tell us who they are so we can give them what we think they need?

I've been talking with a number of agencies and cities recently about this challenge.

In the email world (where GovDelivery's work is focused), I see two major categories of content.   The problem is that I have a hard time categorizing them.  I'm hoping this blog entry forces me to come up with some labels.

Here it goes:

1) "Give me what I want" content is content that the user finds on your website and determines is of value.  The user says to you that she wants to know when a certain type or category of information is updated.  Examples of this might include: "Diabetes Publications"  "Updates to Park Hours"  "City Council Minutes" "Federal Reserve Board Rulings" or "Tax-related Press Releases"

2) "Give me what you think I need" content is content where the user self identifies as being interested in a certain issue or as having certain characteristics.  The user says to the content creator, "Please send me whatever you think will be useful to me."

The best example of this type of information is an e-newsletter.  (A recent favorite for me that is this one targeted at e-government professionals:  http://www.usa.gov/dotgovbuzz.html).  The IRS e-newsletter for Tax Professionals is another good example.

Another example would be where a local government, like San Bernardino County, CA gathers zip code or neighborhood from subscribers and then sends messages to those subscribers on any topic affecting that zip code.   We see this type of approach used for emergencies where it's hard for users to know ahead of time what the issue will be.  The user wants to say to the content creator, "Look, I live in this particular area.  If you ever think I need to know about something, please use your judgment about what to send me."

Both of these types of content play an important role.  We recommend to clients that they offer as many "Give me what I want" options as possible and limit "Give me what you think I need" content to where there is a clear stakeholder group that you think will trust you to create targeted content of interest.  The "Give me what you think I need" content takes more effort and requires that your audience trust your judgment (a diabetic might want to read the "top ten health tips" from the NIH, but not from a pharmaceutical company).  This type of content allows you to help the user filter through the vast amount of content on your website to see what you as the content creator or manager thinks matters most.

If you use the proper cross-promotional techniques, offering both types of content has another benefit: You will actually get more people signed up to all of your updates.  Why?  Because just like shoppers on e-commerce sites are open to buying related products when cross-promotion occurs at the right time, citizens looking for content are open to signing up for different types of content during the initial subscription process. 

A user who signs up for the "Monthly City Newsletter" might also sign up for updates on Park Hours and City Budget Announcements.  A user who signs up for Mens Health Publications from the CDC might also sign-up for the Obesity Prevention Newsletter.
Attachments are one of the most useful features of email. The general public uses email attachments to share photos with friends and family, businesses use them to exchange legal documents, and, in an emergency situation, a government could use attachments to convey important safety information. In a natural disaster, for example, if the government's Web site was unavailable, an email containing a file with important safety tips could be very valuable.
 
Like most people, I tend to take attachments for granted; however, two recent questions caused me to examine attachments a bit more closely. A client wanted to know if email bounce rates were higher for emails with attachments, compared to emails without them. Another client wanted to know if there was general data about businesses blocking attachments from their email servers. While I couldn't find an answer to the second question, finding an answer to the first one was eye opening, and sheds some light on the second.
 
Nearly 50M emails are sent using GovDelivery each month. That data provides access to a lot of valuable data about how ISPs treat attachments. We examined the data and concluded that while senders who include attachments have slightly higher bounce rates, the differences are so small as to be almost negligible, far less than 1%. Getting to this answer uncovered some interesting details, which I'd like to share here.
 
The major email providers, including AOL, Gmail and Yahoo! have fairly liberal policies regarding attachments, even accepting rather large files, or more accurately large message sizes. (A message could include multiple files. The sum of the attachments and the email content makes up the message size.)
 
  • AOL 16MB
  • Gmail 20MB
  • Yahoo! 10MB (a premium edition supports larger file sizes)

The primary concern with email attachments for most ISPs and mail recipients is the risk of viruses. AOL's statement about attachments is clear on this point: "Attached files can contain virus programs or Trojan Horse programs that could damage the files on your computer or steal your password. Never download a file unless you know the person who sent you the file, and were expecting to receive that particular file from him/her. "
 
To protect their customers, AOL, Gmail and other providers block emails that contain attachments of some file types. Gmail's policy is "Gmail won't accept these types of files even if they are sent in a zipped (.zip, .tar, .tgz, .taz, .z, .gz) format. If this type of message is sent to your Gmail address, it is bounced back to the sender automatically."
 
Some other things to consider if you plan to send attachments, especially large attachments, regardless of the end-user's email provider is do they use dial-up and other slow Internet connections? These connections can prevent attachments from downloading completely. Other reasons a user might not receive your attachment include:
 
  • Full browser cache will prevent attachments from being downloaded
  • Full email box prevents delivery of the attachment. Email providers are expanding size limits all the time, but corporate and small ISPs may still have smaller limits.
 
Of the 50M emails sent through GovDelivery each month without attachments, 99.57% successfully reach their final destination. The remaining 1/2-percent are sent to invalid email addresses, defunct ISPs, or are filtered by overzealous filtering applications. Emails with attachments, reach their destination at only a slightly lower rate overall, 99.20%. GovDelivery allows users to send emails up to 3MB in size, including message content. Also, GovDelivery will accept any number of attachments, as long as the sum is less than 3MB. Judging by the delivery rates, this is sufficient to pass most ISP filters.
 
If you are concerned about your message getting through to the end user, a common alternative to an attachment is to post the content you'd normally include in the email to a Web page. Provide a link to the URL in the email. Other strategies include compressing large or multiple attachments using a program like WinZip. Lastly, always remember that common file types are best. Adobe Reader is free, so sending a PDF is a decent way to ensure that your message can be read, other file types may require expensive programs to run.
 
Attachments are still a useful and common part of email communication. Taking a few precautions will help ensure that the messages reach their final destination.
  
I've spent nearly 10 years working with public sector organizations and one question comes up over and over again particularly at the more senior levels of government: 

How can we be more _________  (fill in the blank with agile, results-oriented, customer-driven, efficient, or any other adjective out of the MBA vocabulary) like the private sector?

The reason this question makes a lot of sense to anyone coming out of the private sector was best explained by former Secretary of Treasury, Robert Rubin who arrived at the Treasury directly from his leadership role at Goldman Sachs:

"Most people I've known in the private sector are just much too accustomed to linear processes where they're clear decision makers and they make a decision and things more or less then happen.  In government, the decisions you make very often are much less direct in their effect."  - Robert Rubin

Source: Leadership in Government: An Interview with Robert E. Rubin from the McKinsey Quarterly, July 2007

Understanding the difference between public and private sector decision making is fundamental to allowing the tension between public sector best practice and private sector best practice to drive good public sector decisions.

For example, my focus at GovDelivery is on working with our city, county, state, and federal government clients to help adapt best practice communication practices from the private sector into the public sector environment.  In the private sector, communication has a linear and measurable objective which is typically summarized by two words: Get sales. 

We have seen more than 20 innovative companies gain prominence in the private sector by offering different kinds of electronic communication products/platforms.  Where these products excel is in allowing marketers to target distinct groups of customers and potential customers based on demographics, buying behavior, and previous online actions.  In the public sector, this kind of targeting is too time consuming and presents real concerns about discrimination and privacy.  However, by helping our clients understand the benefits of a personalized message, we've worked with them to devise ways to offer the public much more choice in what they receive. 

This doesn't lead to more sales (because our clients generally don't sell anything), but it leads to outstanding and measurable results in other areas such as dramatically more citizens signing up for content updates (what we refer to as "more subscribers") due to higher level of personalization.

There are many more examples of this phenomenon.  What I enjoy about working in the public sector is that there are so many constituencies and considerations that the work we do is complex and often non-linear.  We have to be creative in how we learn from our colleagues operating in the private sector and adapt what they do well (drive towards clear goals) to our more complex environment. 

If you work in government now and this is frustrating to you, you might be one of the people Secretary Rubin was referring to who would be better off in the private sector.  Don't worry though, if you make that leap, we'll be watching what you do and adapting it, as best we can, where we think it can help do the public good.
I have a habit of returning my library books a few days after their due dates. I don't like the fines and returning the books late is disrespectful to other people on the waiting list. In order to make myself feel better (and save a few bucks), I signed up for the Hennepin County Library's email alert service because it promised to send me a reminder a few days before the books are due.

When the first alert arrived, I remembered the books sitting on my nightstand, but the most interesting thing about the email was the comment link next to each book. Clicking the link opened library's site and allowed me to write a review or leave a comment about the books for other library patrons to read.

The comment feature probably isn't new for the Hennepin County Library. Afterall, I have been writing and reading product and service reviews on sites like Amazon, Expedia, and Trip Advisor for years. The link did, however, remind me of an article I read in the most recent issue of Wired magazine (May 2008). In the "Information Overlord" article, the author discusses the virtues of a semantic Web applications. He writes:

"...sometimes social connections are less useful than semantic ones... My Facebook page attracts my friends, with whom I share social bonds. Meanwhile, my science blog attracts complete strangers, with whom I share a common interest in a topic... It's a semantic relationship, based on shared meaning."

The point is interesting and relates to the library link. When I'm planning a vacation, I read the customer reviews at Trip Advisor and Expedia. Before I buy a book, I read the customer reviews at Amazon.com. The library now provides the same convenience. Before I check out another book, I'll read other people's reviews before I spend my time on a book that won't meet my expectations.

While I trust my friends about some topics, books and vacations are a different story. I need to depend on my semantic relationships. Those real people who have already read the book will be more useful than some pretentious reviewer or uniformed friend.

One of the primary tenets of Web 2.0 is allowing people who don't possess special skills to update content regularly and distribute it easily. Blogs, wikis, and podcasts are examples of technologies that support this tenet, and YouTube, Flickr, and Facebook are examples of implementations.

Emails, phone calls, newspapers, newsletters, and websites are all great ways government entities have been distributing content to citizens. Widgets (a.k.a. widgets, gadgets, badges, modules, and many other monikers) are another possibility. Widgets are movable, mini-applications that can be installed on most web pages, making them a perfect example of a Web 2.0 technology.

Google, AOL, and Yahoo! all allow users to add widgets to their personalized home pages. Common uses are weather forecasts, stock price updates, and news feed widgets.

20080416-AddiGoogWidgets.png




Amazon also provides widgets that allow users to add product catalogs and other interactive devices to websites.


20080416-AmazonWidget.png






Widgets are extremely popular with web users, especially 18-25 year olds who are adding them to their MySpace and Facebook pages in large numbers. Widgets are also big business. Some of the most popular of the 14,000 widgets available to Facebook users are produced by Slide, a San Francisco-based company. Slide's widgets include SuperPoke and FunWall. Fidelity and T. Rowe Price recently invested $50M in Slide.

Users can find and add widgets through personalized portal applications, like Google, Yahoo!, AOL, and Facebook; directly from the creator, including Amazon; and from widget galleries, such as Widgetbox and SpringWidgets.

Governments are also participating in the widget revolution, check out the widgets offered by FBI, NASA, and Veterans Administration.

The ROI for creating and distributing widgets is still difficult to quantify, but reaching citizens where they live, work, and play is always valuable to government. GovDelivery is experimenting with widgets that will allow our government clients to distribute their content across the web. We're excited about the future of this content distribution technique and will continue to monitor its evolution.

Email is clearly the most efficient way for governments to reach citizens. According to David Daniels, Vice President of JupiterResearch, "87% of Consumers online time is spent reading their emails" (Dec. 2007).

Not surprisingly, most citizens use common operating systems, browsers, and email clients to register for subscriptions and read their email.

More than 95% of citizens registering for government subscription options hosted by GovDelivery use a computer equipped with a Microsoft Windows operating system, and more than 85% use Internet Explorer. The following chart shows the browser operating system combinations for one federal government department.


02-BlogBrowserOS2.png

Email client and ISP choices are a bit more diverse, but there are still clear winners. Citizen subscribers, at a rate of 65.7%, receive government emails sent through GovDelivery at one of the top 5 email domains, including Yahoo!, AOL, Hotmail, Gmail, and Comcast. We can only assume that the vast majority of citizens are using the email clients provided by these vendors to read the messages, and not forwarding the messages to a separate email client, such as Mozilla Thunderbird or Microsoft Outlook. The following chart shows the ISP usage statistics for all GovDelivery subscribers.

01-BlogEmailClientUsage.png

The GovDelivery statistics are fairly consistent with overall US Internet usage. While it may be tempting to disregard the small minority of users who use Macintosh, Linux, or PalmOS, government serves everyone. When it is viable, I know that governments want to accommodate these users.

Accommodating email clients and operating system/browser combinations includes avoiding commonly unsupported content, like Flash, javascript (or other scripting languages), or large attachments. It also includes sending emails using common protocols, like Multi-Part MIME or Plain Text that most email clients can render. As technology evolves, and more citizens use more mobile devices to access their email, being flexible will be even more important.

I hate having to make choices all the time.  As empowering as it should be to have options... they are usually just annoying. 

I think that is why someone invented suits-- business casual is more comfortable, but it takes me an extra 5 minutes to get dressed because I have to make a bunch of choices.  (Quick side note on the benefit of suits: I didn't even know that I was color blind until I started going business casual.)

I spent 3 minutes yesterday trying to pick which kind of apple I was going to buy.  Why does Whole Foods give me 10 different choices of apples?  If they have to do this, could they at least have the "recommended apple" sitting out in a prominent location?

Microsoft has figured out how to make the default count.  That's why they have "fast setup" which requires few choices and "advanced setup" which, in all honesty, I have never used so I don't really know what it does though I imagine it would give me a bunch of annoying options that I don't need.

When I provide input into software and Web design at GovDelivery, I always push for the same approach.  I want clients to know that if they take the path of least resistance, it will probably be the right path.

A recent article in the New Republic titled, "Easy Does It: How to Make Lazy People do the Right Thing" (unfortunately requires a subscription) addresses this opportunity in the public policy context. 

For example, did you know that by allowing employers to make 401K plans opt-out instead of opt-in, the government has been able to greatly improve participation in 401K plans?  Turns out, the average non-saver isn't choosing not to participate in their 401K plan, they are choosing not to fill out any forms.  Make them fill out a form to opt-out of saving, and the same choice (don't fill out any forms) leads to a better outcome (401K participation).

The only thing I reject about this article is the reference to the public as "Lazy."  I don't think I'm lazy, and I don't think you're lazy either, but in this day and age, almost everyone is busy and making choices takes time. 

When you're communicating with the public, it can be scary to recommend something, but you can do the public a favor in your web design and software design by thinking of your "default option" as the "recommended option."

Hopefully, this slight shift in thinking will lead you to put more thought into the default and will raise your design standards so you're not just making it easy to make choices, you're also making it easy for users to succeed without making any choices at all.