IPUrbia

I’ve said it once and I’ll say it again: too many tech startups spend way too much time ridiculing the big guys and too little time actually keeping a close eye on the competition and, well, competing.

AT&T is rolling out voice over IP as part of its U-Verse offering. This isn’t all that interesting: after all, residential voice service, regardless as to how digitized, packetized, or processized it gets, is still, well, residential voice service and is pretty boring stuff.

Still, what is interesting is that we are increasingly seeing signs of where AT&T (and presumably the other big service providers) are heading: integration. So, AT&T U-Verse customers will now see caller ID on their TV screens; push video from their cell phones to their television, and so forth. More and more, AT&T is leveraging their multiple platforms to offer features beyond just “single bill”. Nothing earth-shattering, nothing revolutionary; however, it creates competitive features that are hard to match because doing so requires access, if not ownership, of multiple pipes (voice, video, data, and mobile).

Increasingly, it’s beginning to get harder for the basement operations to match the offering of the big guys. 10 years ago, any redneck with a T1, PRI and a Portmaster could offer dialup service that was more or less comparable with that of the Baby Bells. Today, the barriers to entry are higher. The same thing will happen in the application space…remember the evolution of the PC software industry? The gaming industry? Browse the list of applications that Google and Yahoo offer these days? In the end, the ability to bring innovative ideas to market and surpass the status-quo requires, to some degree, the ability to match the offering of the status-quo along with the innovative improvements.

To put the problem more succinctly, the time is fast arriving where it is no longer possible to bootstrap small startups creating an end-to-end solution.

The answer, of course, is simple: leverage open tools and frameworks, open APIs, create business deals and so forth. In other words, don’t create massive end-to-end solutions from scratch.

Open tools, frameworks, and standards have been driving much of Web 1.0 and 2.0. They drove the startup cost down to next to nothing and allowed for startups to quickly have access to countless years of platform and software development via a quick download.

However, it bothers me that such an approach may also disappear. Not that open source and open standards are struggling–nay, they are in use everywhere. LAMP (Linux, Apache, MySQL, and PHP) and its various variations is a buzzword of Web 2.0; the open networking protocols underlying the Internet make it possible to use commodity, generic PCs outfitted with a couple of pieces of software to access virtually all the content on the Internet, Asterisk and OpenSER have been behind dozens of VoIP startups, etc., etc., etc…

And the applications? Well…the truth is that most of the web 2.0 developers out there take open standards and open source and develop these nice neat applications that are 20x more closed and incompatible than anything Microsoft dreamed of. Damn them. Damn them all.

At least with Windows, there was an API and a common filesystem to allow interoperability between applications.

Will open source die from neglect or from unacknowledged, unreciprocated abuse?

The time is coming when many of these web 2.0 startups are going to realize, at best, that they are making widgets. Nice widgets with pretty AJAXY interfaces, to be sure, that do very neat functions that no one ever had quite thought was needed (or, in most cases, ever will think is needed), but widgets nevertheless.

Once the novelty wears off (and believe me, it is wearing off quicker and quicker these days), the usefulness of a widget is how neatly it fits into the rest of our widgets. Closed and alone, it will perish a quiet, quick meaningless death, to become part of the web 2.0 bust jokes during the web 3.0 bubble. Open and integrated, it will become part of exciting and innovative solutions that compete head-on and win against the big guys.

Develop and publish APIs for software. Open source the platform. Develop open, because your business depends on a vibrant community that lives and breathes open standards, even (especially!) in its applications.

A comment to my previous post elicited an anonymous response by an individual who “works closely with E-verify” and took umbrage over my analysis that the program unfairly labels the person a criminal. I am posting their full comment below at the end of this posting for those interested in it.

In general, their response typifies some of the problems in policy decisions these days. The anonymous poster argues that workers are protected by “ample time to resolve the issue,” “legal protection against firings”, and “feedback” on their status.

Unfortunately, I tend to think that these “protections” are somewhat idealistic and naive, for the following reasons.

1. Legal Protection: While there is legal protection against being fired over inaccuracies in the database; it nevertheless does happen: in industries where workers are low skilled and plentiful; many employers have no interest in resolving this with the employee and will simply fire the worker. The audio clip of the NPR segment linked to in the original article (and available here) has a story of this exact scenario and indicates that this IS (however nicely the process is politicized) not an isolated incident. While the employee in the story was able to get their job back, doing so required legal action which is simply an undue burden, especially since most of the people involved in such situations will have quite limited financial resources. Criminalize them all and let the courts sort it out is not responsible governance. BTW, what legal action has the federal government taken against the employer in the above segment for violating employment law on several counts?

2. Feedback: The anonymous poster makes a couple of contradictory statements about feedback:

Third, the employee does indeed receive ‘feedback’ on why E-Verify could not confirm his or her status – within seconds of the query, the program transmits a message stating either ‘work authorized’ or ‘tentative non-confirmation, SSA unable to confirm’ or ‘tentative non-confirmation, DHS unable to confirm,’ and instructs the employee on how to resolve it. Thus, the employer instantly knows which agency was not able to verify employment authorization and provides the employee with this information so the employee can resolve the issue.

While the initial claim is that the employee receives feedback “within seconds of the query”; the poster goes on to state that the employer (NOT the employee) is the one actually receives the feedback and is responsible to relaying that on to the employee. How many people will just be “let go” because of this and not be told the reason?

3. Ample time to resolve the issue: This is the only valid point; the government does allow 3 months for employees to work through these issues and validate their information and citizenship. Still, given the above two problems, there are and will be people (too many people) who don’t ever get the opportunity.

In the end, these protections might be ample IF the database was reasonably accurate. However, given the 10% inaccuracy rate, it is downright irresponsible of the Federal Government to proceed with this program. While I generally dislike these programs in general; the government could have at least implemented this program in a responsible manner which involves, at the very least, making the database accurate to at least 99% (and 99.9% should really be the target) and clarifying inaccuracies directly with the employee, NOT through the employer.

Further, if the Bush Administration really were interested in dealing with illegal immigration, they would proceed in a manner that is consistent with their love of “supply side economics” and target the businesses that knowingly employ illegal immigrants with no pretense of false documents or employment taxes; I can supply a list of at least a dozen to get the process started. However, given the relative political clout, campaign donations, and so forth of business owners compared to the average employee, I fully expect their ill-advised policies to continue full steam under the usual politicized pretenses and no real expectation of actual results despite the billions of dollars of tax dollars and the human costs of innocent citizens mistakenly caught up in the overly aggressive dragnet.

Yes, there are protections. Yes, there are recourses. But, for the people who get mistakenly caught up in this mess, it is simply an undue burden and, given the rate of inaccuracies, it is simply too much of a burden on too many people.

It should not be our burden to clean up your mess.

Now, as promised, is the comment:

I think clarification on your post is needed because many of your points are simply not accurate. In particular, the comment, “..90% accuracy for a database that will instantly label you a criminal and drop you from the potential employment pool, most likely without any feedback as to the source of your miseries?” is problematic and very misleading. As someone who works closely with E-Verify, I can attest that the program does not “instantly label you a criminal..”

First, there are many reasons as to why an employee’s information would cannot be confirmed immediately, including becoming a citizen and neglecting to update your information with the Social Security Administration, getting married and neglecting to update your information, etc. etc. In no way does the program make the assumption that the employee is not authorized to work in the U.S. or label the employee a ‘criminal.’ In fact, in most cases the initial non confirmation occurred because the employee neglected to update his or her own records.

Secondly, if the initial query does not result in a message stating the employee is work authorized, the employee has ample time to resolve the issue with either Social Security Administration or Dept of Homeland Security. In the majority of cases, the case is resolved within 24 hours. Since the employer is required to allow the employee to continue working during this timeframe, employees are not adversely affected whatsoever, other than being inconvenienced with going to the local Social Security office to update their records (which they are required to do regardless of whether their employer uses E-Verify).

Third, the employee does indeed receive ‘feedback’ on why E-Verify could not confirm his or her status – within seconds of the query, the program transmits a message stating either ‘work authorized’ or ‘tentative non-confirmation, SSA unable to confirm’ or ‘tentative non-confirmatin, DHS unable to confirm,’ and instructs the employee on how to resolve it. Thus, the employer instantly knows which agency was not able to verify employment authorization and provides the employee with this information so the employee can resolve the issue. I realize that many are not familiar with E-Verify, and as a result misconceptions abound, but I believe that one should know the facts before casting criticism, or he or she risks the unfortunate circumstance of making false claims.

The blurb is a little scanty, but while out and about this evening I caught a news clip on NPR about a fight between the Federal government and the State of Illinois over a worker verification program created by the Department of Homeland [Pseudo-]Security.

The jist of the program, called “E-Verify,” is that employers can run quick checks on the social security numbers of potential employees to verify citizenship and authenticity of the social security number. However, the State of Illinois recently passed a law banning use of the “E-Verify” program in the State of Illinois, citing the high inaccuracy rate of the database and the havoc that these inaccuracies can wreck on the lives of citizens that get mislabeled, often for seemingly small problems like misspellings and typos.

The news clip was interesting for three reasons. The first was the quote by an official of the Department of Homeland [Pseudo-]Security defending the accuracy of the database, saying that while nothing is perfect, they do have over 90% accuracy and any errors can usually be corrected in days.

90% accuracy for a database that will instantly label you a criminal and drop you from the potential employment pool, most likely without any feedback as to the source of your miseries? Innocent until proven guilty is now being replaced by “guilty until name is spelled accurately and typed correctly by an entry level bureaucrat”.

I can’t think of any commercial offering outside of commercial mailing lists that would even think of operating with such an incredibly low reliability rate. This is compounded by the recent government fetish for relying on such databases to catch “terrorists” from flying to see Grandma Jones or, now, to get a job at the local poultry factory (the missing link of the September 11th bombers???). If 90% accuracy is typical across all of these databases (and, given various media accounts about the no-fly lists, it does not seem unlikely), by the time you get 5 or 6 of these databases going, life becomes hell for about 1 in 2 Americans. Even 1 in a 1000 innocent victims is unacceptable.

The other issue, though, that this brings to mind is the sure insanity of the whole concept and just about every program that DHS has launched since its inception. In the aftermath of September 11th, the reoccurring theme in the media analysis was that the various security agencies of the Federal Government HAD all the information needed to predict and thwart the attack. However, due to what can best be described as a “high signal to noise ratio”, the warning signs were (understandably) overlooked as “noise”.

Given that the failure of the security agencies to prevent September 11th was their inability to hear the warning signs over all the noise, it is absolutely ridiculous to me that the governments answer is to implement various databases and spying programs to gather in MORE information. This is roughly analogous to, upon having difficulty making out the other end of a cell-phone conversation in a coffee shop, walking down the road to a rock concert.

It is understandable, inevitable, and, to some degree even admirable that the government is trying to apply technology in order to better perform its various roles and functions. Nevertheless, many of these steps are at best misguided, some are destructive, and a few are downright Orwellian, and a lot miss the target all together. Kudos for Illinois for at least standing up and demanding that the Federal government not implement [pseudo-]security measures that put the burden of auditing on innocent citizens that get chewed up by the process. Given the power that these all encompasing databases end up wielding in our lives (credit score is a notable example), this is something that we need to watch very carefully. Technology is NOT the magic bullet here, especially when the bullet misses and hits you instead.

For those of you who just flat out disagree with most of these programs, objecting on “accuracy” reasons may have a lot more efficacy in the end than objecting on merit. This is akin to lawyers fighting court battles on procedural grounds–it may not be for the right reason, but, in the end, it works.

Interesting article on the concept of the “professional manager” versus the master craftsman.

“The ‘professional’ manager earns no respect from the workforce because he can’t show them how to build a product or to meet their immediate concerns,” Mr Protzman wrote in an e-mail to Kenneth and William Hopper, authors of the recently published management history The Puritan Gift. “The workers are frustrated because no one will listen to them. He typically ends up trying to make friends with them, in an attempt to persuade them to do his bidding. Most workers have no incentive to co-operate because they figure they can wear him out, or outlast him, or both. Our managerial culture is so far behind Toyota’s that it’s just not funny.”

But surely today’s business elite dismisses such talk as sentimental, and relishes the speed and versatility of the generalist super-manager? Actually, no. Listen to the views of McKinsey director Lowell Bryan and his colleague Claudia Joyce, who offer this analysis in their new book Mobilizing Minds (to be reviewed in these pages on Thursday).

“In the digital age, professional work is often what produces the most wealth. It is also often the work that is most difficult to manage well. And indeed, by its nature, it is more ‘craft’ work than production work,” Mr Bryan and Ms Joyce write. “As such, the frontline manager of such work needs to be not only a good manager but also actually a ‘master craftsperson’. Scientists need to be led by chief scientists, investment bankers need to led by master bankers, and lawyers need to be led by a general counsel, and so on.”

Read it…