◎ Are We At Peak Networking Pricing ? What Will Commoditisation Deliver ?

Many people are predicting that networking will become a heavily commoditised with cheap white box hardware in the next few years. But I don’t believe commoditisation will happen the way that most people expect. So in this article I’m working through different ideas and concepts on product pricing and perceived value of networking as a whole. I would welcome discussion on this, I’m not completely confident that I’ve got my ideas locked down here. See you in the comments !

Is Networking Hardware and Software Overpriced ?

The evidence is clear  that networking customers are very strongly looking for ways to avoid expensive network equipment. This isn’t new or even surprising but there is a lot of evidence to suggest that price matters more than ever. The very high profit margins enjoyed by existing vendors, such as Cisco, Juniper and even Brocade, are definitely under threat and probably will reduce drastically in the years ahead. I also perceive this as a trend across of all of IT. We can see that Microsoft continues to lose premium positioning as the desktop market shrinks and new server technologies like Docker and KVM bypass Windows Server. For Microsoft the core product market shrinking and loss of technical leadership is serious blow. Even Microsoft Office is losing relevance as smartphones replace desktops while also offering cheap applications, supported by cloud services) that replace MS Office functions.

And professional services is changing too. IBM commoditised its own professional services by outsourcing to India choosing mostly third parties like Tata who are offering the same service to large companies directly. SaaS services like Salesforce, Dropbox, Atlassian are examples of professional services being outsourced. The actual software isn’t that important, it’s the professional services of internal staff and resellers that is being replaced here and, effectively, commoditising professional services.

Don’t mistake this as a blanket statement. The market will remain diverse for another decade at least. But these new products will be added to the market but the old ways of providing professional services are not going to grow.

IT is Losing Value To Business

Today, I believe that a significant underlying driver behind this loss of perceived value in products is that Enterprises and Corporate managers no longer perceive IT as worth significant amounts of investment. After two decades of of growth and value, IT is no longer producing good returns. A larger WAN connection does not improve business productivity, an bigger Internet connection doesn’t improve customer communication or reduce supplier purchasing times. Repeat for any other area of infrastructure.

Executives and managers are looking at consumer products like smartphones  and no longer perceive “computing” as hard or challenging. Equally, consumer software experiences like Dropbox, Facebook and Google+ create a perspective that software isn’t hard or complex. This perspective is reducing the value of corporate IT amongst businesses that don’t value IT or perceive IT as core to their success.

And the oft-repeated phrase of “IT is hard, complex and needs experts” from IT is wearing thin. IT is hard to get right and vendors haven’t been focussing on usability and simplicity. IT products will get simpler to use, operate and maintain in the years ahead because customers are demanding that. The result is  commoditisation in application and software. Often, commoditisation is a simplification of products by focusing on use.

For example, cars in 2010’s are much simpler to maintain than cars from the 1980’s. The last 20 years in car design has seen a solid focus in reducing the servicing costs – service intervals have typically doubled from 6000 to 12000 kilometres. Consumables have been reduced, servicing has been simplified and customers are pleased that their car are easier to own & operate. That’s an aspect of commoditisation.

Commoditisation In Networking

My current view is that networking will move to a commoditisation model equivalent to Apple’s iPhone and smartphones more generally. An Apple iPhone is a commodity product – made from off the shelf components that are generally available to any phone manufacturer. Apple adds value through carefully curation of its designs, some quality engineering  and excellent product management to appeal to a specific audience who have specific requirements.

Replace “Apple” with Cisco, Juniper, HP, Brocade et al and I see this as the commodity networking future.

Apple adds value to the commodity components through high quality and reliable software. Apple has innovative support programs that combine with best in class buying experience using direct sales instead of indirect sales through a channel. Same also for the incumbent vendors in networking who have their own software platforms – IOS,

Importantly, Apple charges fair prices for their product – the product is not cheap but they deliver fair value for the price charged. The majority of people who have owned and used an iPhone are open about their positive experience and brand loyalty. Customers will pay a premium where it’s deserved and reasonable.

Data Networking is a market that has specific requirements that are somewhat different from normal compute or storage. Vendors can still offer products that have value when using commodity components. A faulty network is much more likely to impact the entire business than a few failed servers or a problematic storage array. You can reboot a server or even restore a storage array but a dead network is a major event.

Peak Pricing for Networking

I doubt that few companies would regard the cost of networking as “fair value” today. In recent discussions, interviews and meetings with customers I’ve seen serious concern about the cost of networking and are actively looking for alternatives. We have all heard the “price debate” before and watched as the same person went ahead and bought the high priced option anyway but a market transition would happen if  just 5% of sales shifted to a cheaper option. A 5% gross revenue drop for any major vendor has major impact on share price and, in short order, the way the business operates.

WAN bandwidth, hardware maintenance, software reliability and lead times are all major concerns for customers and each is costly recurrent budget expense. Do customers perceive firewall and security tools as secure and valuable ? Is the network equipment actually reliable and predictable ?

Pricing of networking equipment and support is likely to fall to match customers expectations and it doubtful that marketing or value adding will change this. I make the logical conclusion that we have reached “peak pricing for networking” as reasonable. Once again I think there are lessons from the car industry where car pricing has been stable for the last decade and roughly moved inline with the local economies – effectively cars have reached “peak pricing” and there is no more value that can be added that increases the cost. ( Well, except premium cars of course that have any number of largely pointless features but that’s a relatively small

On the other hand, we need a lot more networking. Cloud computing has many new requirements, not least is the bandwidth increase. So while I believe that unit prices will decrease, it’s likely that overall revenue will stay about the same. And that’s confusing for some people. Logically,  networking hardware will be cheaper through commoditisation of silicon but you will buy more software networking to scale your private cloud or provide better connectivity to the public cloud.

EtherealMind View

As I said in the introduction, I look forward to discussing your views in the comments. Do you agree with this ideas and logic ? Do you think networking is too expensive ? Is there more value that networking vendors could add that would encourage you to pay more ?

What are the issues that would lead you to buy white box instead of branded ? Would you and your company actually make a change ?  What  about “peak pricing” – could the vendors increase product pricing (and their revenues) or will your budgets and purchasing refuse to increase the value ?

  • Steve J

    “IT products will get simpler to use, operate and maintain in the years ahead because customers are demanding that.” This changed when? Since when did anyone WANT it to be complex? Yet it is complex. Replacing a Cisco router with an iPhone sounds absurd and it is. Color me skeptical.

    • http://northlandboy.com/ Lindsay Hill

      “Since when did anyone WANT it to be complex?”

      We don’t directly ask for more complexity, but because we have historically demanded more features and more knobs to twiddle, that’s what we’ve ended up with.

      How often have you sat in a meeting and argued for the removal of features/options? Usually it’s easier to just keep adding features, few HW/SW designers have a strong enough vision to take things out. I find it interesting that the Nexus 9K has around half the codebase of the 7K, presumably with many features removed. I wonder how many people will jump up and down about those missing features, vs how many will never even notice. No doubt the feature creep will come back, when big enough customers demand them.

  • James

    At the very least, I agree that we’re really not getting a fair value in networking today. I’m cautiously optimistic about the future. I don’t see OpenFlow/SDN as a solution, but I see it as an intermediate step kick starting innovation. The productization is going to happen further down the road than most think, and it’s going to look very different than they expect.

    The goal of networking is simple – you get data from A to B. It’s all the other aspects like dealing with performance, scalability, reliability, and security that complicates things. You pick from a pile of technology — VLANs, VRFs, MPLS, firewalls/ACLs, IPS/UTM, QoS, packet shaping, WAN optimizers, 802.1x, routing protocols, SPB/Trill, FHRP, MLAG, Wifi controllers, NetFlow/IPFIX, service insertion, etc, etc, etc. Then you piecemeal it together into a “coherent” solution that’s unique to your organization.

    I need a network, not a technology bolted onto the existing kludge promising to zerg creep through over 5 years that, by the way, still has lots of unresolved caveats. That’s not going to make the situation better. I’d love a curated solution like an iPhone that’s uniform, but I’m a bit skeptical about any promises of that.

    There’s a lot of conflicting market pressure that create all these different pieces to *your* puzzle. When we make all of our puzzles look the same, we’ll get bored and figure out how to make them look different again.

    Still, I look forward to a new puzzle, this one is getting a bit arduous.

  • J Max

    Something very simple I will always remember. in hard time money vs anything(Networking, servers, Cloud, etc…..) Money always wins.

  • XiaoPongTze

    Your analogy really lost me since the iPhone is the opposite of a commodity. Mobile phones in general may be. Perhaps you should read the wiki page on commodity and try to come up with a better analogy? I don’t think IT has lost value, I think that some organizations are better than others at getting value out of IT and spending on the right thing. Certainly big data has lots of value to many.

  • Paul A. Leroux

    a customer once told me. “I like what I use so why would I buy Cisco at 3x the price?”. “We don’t give data centre tours, our facilities are secure, we aren’t even allowed to disclose where or how many data centres we have. So what does it matter what is pushing the packets around as long as I know it works”