Features and Analysis

What do busy circuits, taxis and NFV have in common?

“Sorry, all circuits are busy.” When was the last time you heard those dreaded words on your phone? That universally used network congestion message was more common in the landline-only years, but it still pops up once in a while on my mobile. As operators make network functions virtualization (NFV) a key part of their strategy, it may soon be replaced forever – but will its successor be an improvement?

Taxis and telephones

Strangely enough, this came to mind when I read a recent post by alternative taxi company Uber board member, Bill Gurley, explaining the rationale for and operation of Uber’s dynamic pricing model.  You may have seen the flood of tweets, news and posts alternately praising and condemning Uber for their ‘surge pricing’ innovation, in which ride prices may increase by as much as ten times when the number of would-be riders greatly exceeds the number of available drivers.

As often happens, sensational sound bites from irate riders and preaching pundits have distracted attention from aspects of perhaps deeper interest.  Uber competes on ease of use, reliability, transparent pricing, and convenience.  But Uber also aspires to lead with low prices (contrary to the ‘black limo’ image with which they began, most of their cars and most often their prices are quite ordinary).  Does that sound familiar?  Replace taxis with telephones and we could be talking about a lot of modern operators.

The operator’s dilemma

As Bill Gurley tells it, surge pricing is the most realistic alternative to the taxi world’s congestion message: “Sorry, all drivers are busy!”  But Uber’s solution is different and arguably better than what was available to operators over the last century or so.  Where more circuits were simply impossible to create, more drivers can be brought or kept online by simply raising their incentives (they get 80 percent of the fare).  Of course, this increase in availability comes – literally – at a price.  Want your driver to forego their New Year’s Eve party so you can get home from yours?  Be prepared to pay – or to wait a lot longer for an Uber competitor with a less flexible fleet.

What are operators to make of this dilemma?  NFV will let them create more of almost any part of their infrastructure on the fly, and turn it off when no longer needed, but that too comes at a price – one that is proportionately much higher than what Uber faces.  While Uber needs only to make minimal investments in drivers who furnish their own cars, operators also have to buy the ‘cars’ up front, or at least reserve them from infrastructure providers.  And the business and technical agility that they need to amortize those costs more quickly with dynamic pricing is also expensive to acquire and maintain.

Marketing lessons, too

There are further lessons that operators – and other digital service providers able to implement NFV and dynamic pricing – can learn from this episode.  Regardless of their position on surge pricing per se, commenters generally agree that from a marketing and communications point of view its implementation was a stumble at best.  A typical headline at the height of the storm was, ironically, “Uber surge pricing: sound economic theory, bad business practice”.  How will digital service providers with millions of customers position new business practices of this kind?

Whose mistakes will you learn from?

There is an old saying that “experience is what you learn from your mistakes, but wisdom is what you learn from others’ mistakes”.  How wise will digital service providers prove to be when they are able to use dynamic pricing to create real-time supply in response to real-time demand?  Time will soon tell.


About The Author

Director IoT Projects

John is a TM Forum Distinguished Fellow who recently co-founded network functions virtualization (NFV) innovator Svarog Technology Group. He is well known for his work with information and process models and has held technical and thought leadership positions at Progress Software, Ceon, Evolving Systems, and DEC.

1 Comment

  1. Hewage Fernando on

    Agree and Disagree 
    NFV is really back to the future. In the old days, ie prior to the introduction of switches and routers into an Operators network not just to carry the traffic but also to decide how that traffic should be routed within the network based on local intelligence and conditions, this control function was done centrally. Take the PSTN world’s STP (Signaling Transfer Point) as an example. An STP at the centre of a Telco’s network was overseeing and controlling traffic flow in the entire network using the C7 (CCIS7) links, using the simple rules of supply and demand of the Network as a whole. The benefit of such a scheme to a Telco is obvious.
    So in summary: While I fully agree with the view “…NFV will let them create more of almost any part of their infrastructure on the fly, and turn it off when no longer needed…” I am struggling to understand whether the writer was right in saying “…This is different and arguably better than what was available to operators over the last century or so.” All I say is it is not new. It is old stuff coming back as future. Back to the future.

Leave A Reply

Back to top