5 Reasons to Ignore Automation

By this point, you may already be growing tired of the automation hype cycle. That's too bad, because as hype cycles go, this one is just getting started. By this time next year, you'll probably be feeling nostalgic for simpler days, when NFV was the leader of all hype, with 5G running a reasonably close second.

Ever mindful of its role as the telecom industry's most reliable trendspotter, Heavy Reading has just released a report that puts telecom automation in perspective -- or make that perspectives. Telecom Automation: Heavy Reading Perspectives delivers observations and insights from Heavy Reading's analyst team on the coming new automation age and how it will affect telecom industry ecosystems in these areas:

  • Emerging technologies and services, such as IoT
  • B/OSS transformation
  • Analytics
  • NFV MANO initiatives
  • Wide-area network deployments (SD-WAN)
  • 5G
  • Network security

That's a lot of ground to cover. And some of that ground is going to feel painful to tread for one simple reason: The telecom industry's road to automation is nothing but an uphill climb, and that journey is just beginning. And it's a steep, steep hill. Tree lines will be involved. And an encounter with the occasional yeti isn't out of the question.

Reasonable people facing a daunting, seemingly unending, and probably thankless task would ask a simple question: Why bother? As with all daunting, seemingly unending and probably thankless tasks, there are plenty of reasons to forgo the pain and leave the work to somebody else. These five almost immediately come to mind:

  1. Automation is hard. Really hard. Compared with true automation, virtualizing piece parts of a network is about as difficult as opening a screw-top bottle of Merlot. Automation will require bringing tens of thousands of those virtualized piece-parts together to create a machine that would let go of itself.

  2. It's going to take a long time to build an autonomous network. To borrow the well-used line from John Maynard Keynes: In the long run we are all dead -- or enjoying the golden years scouting the early-bird specials and complaining about how much better the world used to be. If you remember what things were like before Al Gore invented the Internet, you may be put out to pasture before the job is done. So why do all this work to make things easier for some kids you don't even know?

  3. The web guys have a huge head start. CSPs may be milling around at the bottom of Mount Automation, but the Web giants have already set up their base camps and are ready for the final climb to the summit. The race to the top may have already been lost.

  4. Investors are not going to be happy. The people who own shares of your company's stock don't like it when lots of money is spent with little or no immediate return. And automation will require a big up-front investment that's going to put more of a squeeze on margins for the next few years at least.

  5. You don't want to be a service animal for a machine. Even if you stick around long enough to see the job completed, the reward for spending years to build a self-driving network may likely be a trip to the unemployment -- or at least underemployment -- line. A truly autonomous network will be able to get by with a minimal number of human maintenance engineers.

So maybe CSPs that don't automate will end up in some virtual tar pit sometime in the late 2020s. It's been a good run. Maybe it's time to let somebody else worry about the future.

— Dennis Mendyk, SVP of Research, Heavy Reading

Page 1 / 2   >   >>
mendyk 10/27/2017 | 11:11:43 AM
Re: Watch out Agree about when and not if regarding telecom automation. But the bigger question may be, who?
Phil_Britt 10/27/2017 | 8:16:06 AM
Re: Watch out Even though "automation is hard" and may take longer than expected, it continues to become more pervasive throughout industries, not just telecom. So the question is "when" not "if."
Duh! 10/26/2017 | 11:14:59 AM
Re: Perverse disincentives Opex reduction is a sure bet for the C-suite. If you take out some number of heads - loaded with health care, pensions (the unions haven't let go of those), OT, etc. plus strike risk - you know what it's going to do to your income and cash flow statements over the next few years.

If you create a new service, you're betting that its value proposition is as good as you think it is, that you've priced it right, that your competition isn't going to respond, that its margins will compensate for revenues lost by cannibalizing other services, and that your team is going to execute. If you're smart, you also factor in lots of confirmation bias on all the above.

Of course, you need some of both. But if your main objective is to keep raising the dividend to keep your share price up, which business case is going to be easier to buy into?
mendyk 10/26/2017 | 10:41:54 AM
Re: Perverse disincentives The walking off the bridge onto land part would be getting acquired by one of the Websters. Brought to you by AT&T, an Alphabet company.
Dissertationcorp 10/26/2017 | 8:24:01 AM
Re: Perverse disincentives This is really niuve we need help to make sure that we have some solutions in our life styles
Joe Stanganelli 10/26/2017 | 7:05:09 AM
Re: Perverse disincentives Playing it safe by staying on a disintegrating bridge. It's nice to have a companion in stretching analogies to torturous limits. ;)

Taking it yet further: For all the hype of digital transformation tech as we know it presently, the real winners are going to be those who innovate yet a third option that is better than what we have now and renders it moot by just walking off the bridge safely to land -- leaving everyone else to their morbidly no-win choice.
Joe Stanganelli 10/26/2017 | 6:54:09 AM
Re: Perverse disincentives @Sterling: Excellent point/reminder -- and it reminds me of a corollary effect: That when you save your organization as a whole -- including other departments -- $X, then the leaders of those other departments may get credited for coming in under budget, while you as the innovating department leader stand to get reprimanded if your CAPEX is too high.

Or, as it is put much more succinctly: (link)
yarn 10/25/2017 | 4:33:08 PM
Re: Perverse disincentives The same arguments are applied to ignore other difficult challegences such as climate change. The power of ignorence is strong as there are few issues that can't be ignored for a very long time. Many will just go away, but others exacerbate and become more costly to solve later. Without automation few people would be able to afford a car and there wouldn't be a case for building a national highway system, gas stations and massive oil production. But there would be less global warming...

Hm, perhaps there are 6 reasons to ignore autmation:-)
Sterling Perrin 10/25/2017 | 2:06:11 PM
Re: Perverse disincentives Brooks,

The other side of the NFV coin is something that generates revenue - then everybody seems to buy-in bc revenues increase and budgets increase. The CTL comments help explain why capex/opex saving use cases go nowhere but new revenue use cases move forward. There is a place for automation in revenue-generation use cases. In the report, we cite SD-WAN as a prime example of that.

brooks7 10/25/2017 | 2:05:57 PM
Re: Perverse disincentives Duh!,

Easier said than done.  Reducing Opex does not mean that there is a project approved.  It means opex was reduced.  Here we are talking about something where one would think that whole departments will disapear over time.  Something more like:  Reduce operational headcount by 50% in 2 years.  People look up and ask...Will I have a job?


Page 1 / 2   >   >>
Sign In