The innovation behind number portability is the creation, in the late 90’s, of a single database connecting numbers to devices that all carriers access when completing calls. That database now contains more than 500 million numbers and is queried over one million times every day.
But that profoundly useful system may now be at risk of a pointless and completely avoidable meltdown.
That’s because the contract to manage the database is set to expire next year, with the FCC dangerously behind in negotiating a new one even as its requirements are changing rapidly.
Two principal bidders have emerged. One is Neustar, which developed the existing system and has administered it all along. (According to The Post’s Catherine Ho, half of Neustar’s nearly $1 billion annual revenue comes from fees approved by the FCC and paid for by phone companies who use the system.)
The other bidder is a company named Telcordia, a subsidiary of Ericsson, which operates a similar system in India and is hoping to expand to the United States.
The FCC began the long process of negotiating a new contract back in 2010, hoping to finish by the end of 2012. The original schedule anticipated plenty of time, if necessary, to transition from Neustar’s system to a new one.
But the effort has been plagued by delays and accusations of misconduct by pretty much everyone involved. Requests for proposals didn’t go out until May 2013, and the North American Numbering Council, an industry advisory group that advises the FCC on the contract, has yet to make a recommendation for a winning bidder. Last week, NANC met in a closed session. So far, there has been no announcement about the outcome, if any, of that meeting.
As the contracting process drags on, critical technology changes continue to outpace it. In the midst of wrangling between Neustar and Telcordia, significantly, the FCC earlier this year okayed long-requested trials by the remaining wireline telephone providers to begin the shutdown of the old switched network. That transition that will ease the dwindling number of customers onto all-digital networks — the most dramatic revolution in the history of voice communications.
Trials, which could begin soon, include tests of new numbering systems that will both ensure continued operation of portability and test ways to improve the system for an all-digital world. The FCC held its first workshop on numbering trials just last month.
So no matter who gets the next portability contract, the bids already submitted can’t possibly have taken into account either the challenges or the opportunities of an all-digital phone system, a system certain to become reality over the course of the new agreement.
Former FCC chief technologist Stagg Newman, for one, sees some uncomfortable parallels here with the disastrous rollout of Healthcare.gov, where requirements were constantly changing even as contractors struggled to build a high-volume system using technologies that were evolving as they went along.
Newman’s concerns are valid. Healthcare.gov may have been the most visible example of the problem, but it is hardly unique. The government contracting process is easily derailed–abundantly evident in the new portability contract. Its slow speed and profound complexity is a regular source of collisions with the accelerating pace of disruptive improvements in information technology.
Indeed, expensive failures have plagued large-scale government systems projects since Herman Hollerith first proposed using punch cards to speed up the U.S. census in 1890.
The mismatch has only gotten worse. Not only do user requirements, design specifications and bidder gamesmanship pile up as contract negotiations drag on, but the capabilities of available technology change even more rapidly, the result of a bedrock principle of computing known as Moore’s Law.
As I learned first-hand in an earlier career building government systems, no matter how well an agency scopes out its requirements and no matter how efficiently it manages the contracting process, better and cheaper hardware and software solutions are certain to have become available in the interim. As users or simply funders, taxpayers never get the benefit of that improvement.
Worse, the contracting process encourages governments to extend the life of their technology investments long past their usefulness. When Microsoft recently announced the retirement of Windows XP, its 2001 operating system, the company discovered that millions of government computers still rely on it. Now those governments must pay Microsoft extra to keep supporting obsolete software.
Government contracting is designed to be slow. Part I of the current Federal Acquisition Regulation runs thousands of pages. And lawmakers and regulators can’t resist the addition of conditions and processes that advance other policy objectives, such as giving special consideration to minority or women-owned businesses. And choosing the lowest price doesn’t always mean getting the best value.
The FCC’s approved requirements for the portability system exacerbated the risk of a potential meltdown by requiring the winner to deliver a finished system all at once, and do so in a now greatly compressed timeframe. The winning bidder may have less than a year to build and transition to its new system.
How likely is that? According to a new report from The Standish Group, a consulting firm that studies large-scale IT successes and failures, a new system developed by someone other than Neustar has only a 6 percent chance of coming in on time and within budget. (Telcordia objects to a lack of specifics in the Standish report, but acknowledges that bid details are being kept secret.)
In retrospect, Congress and the FCC should have required the carriers, perhaps through a non-profit consortium, to build and maintain the number portability database, with appropriate incentives and penalties to ensure its smooth operation and with the FCC in the background as the final referee of any disputes.
That’s almost what happened in the creation of NANC. But in the end, the FCC gave itself final authority over approval of the contract and its renewal, inviting precisely the kind of mischief that has occurred.
Hindsight of course, is 20-20. For better or worse, we now have a working system owned and operated by a private contractor. Neustar’s incentives to keep the system state of the art may not be perfect, but there’s little sense in starting over with a new developer who will have to start from scratch. And who, in any case, would find themselves in exactly the same position when its own contract came up for renewal.
Add to that the uncertain impact of the on-going IP transition, and the 6 percent likelihood of success estimate starts to look positively cheerful.
Governments may never be able to align the clock speed of their contracting machinery with that of Moore’s Law. But they can certainly learn from past mistakes. The number portability database has worked exceptionally well so far, but will almost certainly require major revisions to accommodate the transition to all-digital voice networks.
The FCC should leave well enough alone, delaying any reconsideration of the current contract until the revolution in voice services reveals more about its uncertain — if exciting — denouement.
Downes is co-author with Paul Nunes of “Big Bang Disruption: Strategy in the Age of Devastating Innovation” (Portfolio 2014).