Monthly Archives: November 2008

I set up a website for my Father

I have been pestering my Father about the importance of communicating with your “Tribe” on the web for almost a year now, and of course I have to be the one to set things in motion. I went and bought a domain and a WordPress theme that I liked, and put together a community blog for his office last night.

It’s an alright start, great considering I have next to no experience doing any of the things I had to do to get it this far, but it still needs a lot of work. At least I will be able to have an educated discussion when we pay someone to redesign it and combine it with the main website. The biggest problem I ran into is that he just doesn’t have any sort of branded resources or logos available that are designed specifically to be used in a blog or social network profile. I am not in the mood to make things from scratch; I had to use pieces of images from his business card just to make the header logo (which I still am not totally satisfied with). Note to anyone trying to maintain an online identity: pay someone to design logos and handle branding; it is important.

Anyway I am having fun tinkering with it a bit, and adding more features and cousin sites like Facebook pages. I appreciate suggestions that any of you readers might have to offer.


Twitter Groups Business Model

I have been doing a ton of microblogging on Twitter lately, and I am disappointed that there still are no groups for Twitter. After seeing the effect Twitter had with their election page, and recently seeing Tweetmas you would think they would eventually roll out groups for everyone.

They seem like such a simple addition, but I think it has been delayed because Twitter has plan to monetize their service using groups. I mentioned before how I felt their acquisition of Summize was a path to turning a profit, and I can see them combining that service with groups for advertising purposes or charging for enterprise type groups as some of their clones have done.

Twitter’s most significant asset is how it captures word of mouth, which greatly enhances their ability to effectively serve ads. Additionally, studies have shown that the affinity of members within a social network is inversely proportional to CPM rates. Interest groups like the election hub constructed by Twitter will have relatively high CPM rates and a high value since their members would have a fairly low affinity. Of course, they would be totally worthless without driving users to the content, so a group/interest search and recommendation engine would also have to be implemented as well. There still could be traditional self-organizing groups, friends tweeting to friends and co-workers tweeting to co-workers, but unless Twitter is charging for that category of groups there is not much value in it.

Power User’s Guide to News

People appreciate news sites for their ability to provide an edited and selected overview of worthy topics. There is not a more classic embodiment of this than the DrudgeReport, and that is why the site draws tens of millions of viewers each day (Alltop has done a great job realizing this principle of news as well). However, what do you do when your interests are not represented at a singlular site like Drudge/Alltop, or you want to create a hub to easily monitor and correspond with your “Tribe”? You need to roll your own news by pooling together many different streams of content across the web and distilling them down.

Mostly because I enjoy learning new things, I comb through roughly a thousand articles each day in Google Reader alone, and I gradually have figured out what works to help me cope with the heavy load. Of everything I have tried, the Read-It-Later plug-in has been phenomenal.

Having separate sets of actionable items (Gmail starred items), and items you want to get around to reading (Twitter favorites, GoogReader starred items, etc.) is inefficient and goes against our instincts about news. You must pick one repository to aggregate, star, and tag articles. That means taking all of your various streams of information and having them pushed to one place. I chose Google Reader to aggregate content since the items I mark during regular browsing and feed items can be combined through Read-It-Later (FriendFeed might also be a solid option). An additional benefit of the Read-It-Later integration is that now I have multiple options for handling content in Google Reader; I can star items to an actionable list, tag them as before, or simply select articles for future reading. NewsGator is fantastic in this respect also….it probably is a more powerful option than Google Reader, but I am just used to Google Reader.

I will really have my act together when I work out the best way to apply supplemental tools like Yahoo Pipes and Yotify to my RSS feeds. Both are extremely useful for generating unique feeds and removing duplicates; although, most of the fine filtering still has to be done by you (quickly combing through your feeds and highlighting items of interest using Read-It-Later). For news, I don’t like the idea of applying filters that are too exclusionary since I don’t want to miss something unique. There are many neat services out there like AideRSS that try to simplfy the filtering process, but they shouldn’t be relied upon entirely for filtering out the noise. They do supply excellent metadata that have the potential to augment the pipes you construct on your own. The more data available, the more informed your decisions can be about your feeds.

I would really like to see effective implementations of customizable feed dashboards. I know you can accomplish this to some extent using a Netvibes page, but I think that it is something that is still underdeveloped.

I realize I have had a ton of coffee so this post might be a bit scattered, so I will summarize.

• Aggregate content feeds in an RSS Reader (scour web for useful sources of content and subscribe)

• Push feed items and discovered content to a dedicated reading list with Read-It-Later

• Tag useful knowledge and star actionable items (downloading apps, contacting people, etc.)

• Generate relevant feeds and remove duplicates using Yahoo Pipes/Yotify/Google Alerts

• Use all available data to augment the discovery power of filters

• Break down your reading into chunks to avoid exhaustion

• Try to organize things into a dashboard so you can prioritize your reading activity

Reblog this post [with Zemanta]

So Busy

I have been completely swamped fabricating dentures as of late (they make us do 90% of the work here in school), and my room has gone to complete shit. Fortunately I had the excellent idea of paying other people to clean it for me. I cannot tell you how amazing it feels to leave a disgustingly messy room in the morning and return at night to a spotless one.

Anyone living in San Francisco interested a solid cleaning company should look up Wizard’s CleanTeam. For a fairly large apartment I think 250 bucks was pretty reasonable, and they washed my clothes in the awful quarter washing machines and everything. It is so clean it looks like I just moved in.

In a related note, one of the best ways to find reliable cleaning services is to call quality/expensive hotels and find out who does their cleaning. Every hotel I know of outsources their maid and cleaning services, so they know exactly who to contact.

Apple CRM

People use Macs for the clean user interface and software synergy. However, both are accomplished through software and formats that are proprietary and incompatible with most everything non-Mac including the web. You can imagine most enterprise software isn’t specific for the Mac. Recently, I have been looking into options to more effectively handle office tasks, and, being that we should be switching to a Mac-based practice management system soon, I have been looking into options that are native to Apple.

From what I have found there are a growing number of options available for Mac users. Here is a smattering of some cool stuff out there in the way of CRM and project management….

Elements soon to have an iPhone app, developer APIs, and is feature rich

Daylite been around for a while and looks great

Contactizer great user interface from what I can tell

Nighthawk more of a personalized option

There are also a ton of web-based solutions that I found interesting, but software plus services is really what it is all about. You need both power and flexibility. Yes, mobility has grown even more important recently, but even with mobile access it is so important to have a dedicated app that leverages the abilities of the specific phone/platform you’re using. Going with a fully web-based service is limiting in many ways.

APIs will save us all one day

The New York Times released its API for campaign finance a week or so ago. It has been the in thing for a while. I wouldn’t say everyone has entered the era of Big Data yet, but organizations are definitely seeing the value of opening their doors for people to build something even greater with their data and services.

It is a shame that health care is not nearly technologically evolved enough to take that path also. I imagine a future where doctors’ professional records are fully transparent to the public through PUSH APIs providing treatment data to various web services/communities. There is huge value in health professional data, but you can’t just wish an effective health care IT system into existence with some laws and government money. It needs to be grown organically through repeated cycles of innovation.

I spoke recently with several people about personal health records’ importance to the demand for health care IT. I believe in a few years personal health records will become mainstream and drive doctors to adopt digital records and others to innovate.

In the future, open medical data will improve care through greater accountability, and increase the public’s understanding and awareness. More than one person has observed that abundant data has the potential to change the way we make conclusions. Better decisions become cheaper and cheaper. Freely accessible data has the potential to eliminate ethical dilemmas in drug testing and lower costs in post approval clinical pharmaceutical trials by changing the way studies are conducted. I am excited about opportunities to apply medical data for public benefit.

Driving efficiency

I have started to make more of an effort to remember the random conversations and ideas I have during the day, so that I can share them….

I was riding in the car with a classmate to our geriatric clinic rotation at the Laguna Honda Hospital, and we got to talking about GPS navigation and how newer systems are taking into account traffic levels to route people more effectively.

At some point during our discussion about traffic, he mentioned that he read around 30% of traffic in cities is due to delivery vehicles like UPS, FedEx, movers, and pizza. I am a firm believer that more data beats better algorithms; factoring in data on the routes and locations of delivery vehicles, would undoubtedly lead to better routes. I doubt many pizza delivery places have their fleet of cars networked with a GPS navigation system, but I pretty damn sure UPS, FedEx, and other commercial couriers do.

In many ways the present systems already measure this indirectly. They monitor current and/or historical traffic patterns and attempt to generate the best route. Although measuring outcomes is the easiest and most readily available method, supplementation with data on traffic’s causative factors would create a more forward looking model, and more accurately predict the best route.

I am waiting to see GPS navigation services integrate data on construction, parades, delivery vehicles, bicyclists (cars must slow and give them a safe distance), and pedestrians (prevent right and left hand turns if in crosswalk) into their models.

Reblog this post [with Zemanta]