The Cranky Admin
Does Public Cloud Spell Doom for Channel Partners?
Also, what it could mean for your IT career.
Companies that sell, support, manage and maintain computer systems face an existential threat from public cloud computing providers. If the organizations that buy hardware, software and services from them move all their workloads into the public cloud, what role remains for the companies that make up what we have traditionally called "the channel," and how does this impact IT practitioners?
Channel partners consist of distributors (disties), Value Added Resellers (VARs), Managed Service Providers (MSPs), Cloud Service Providers (CSPs), colocation facilities (colos) and a few other miscellaneous categories. Disties set up partnerships directly with vendors, buying hardware and software from them. Most vendors sell into the channel through disties, though some either bypass the channel or compete with it by selling directly to end customers. Dell is one example of a company with a complicated channel relationship.
Everyone else in the channel is downstream of the disties. VARs buy hardware and software from the disties, then "integrate" them. This integration -- the "value add" portion of VAR -- is usually taking multiple hardware and software components then installing, configuring and testing them to ensure everything works. MSPs are usually the organizations that install the equipment on site, manage and maintain it over time and otherwise act as an outsourced IT department.
A colo is a company that owns a datacenter. They have the square footage, power, cooling, fiber optic hookups and physical security to run hundreds or thousands of servers in a single facility. They rent out this capacity to those looking for space in a proper datacenter without building their own.
CSPs are companies that own a bunch of equipment and rent out slices of that equipment in a self-service fashion. You can light up a whole bare metal box or just a single VM. Whereas the big public cloud providers have datacenters scattered around the world dedicated to hosting millions of workloads, CSPs tend to be more regional in their focus.
Colos and CSPs are closely linked, and often the same organization. If you own a colo, why not toss some boxes in there and start selling off some capacity? And if you want to start your own CSP, why build your own datacenters to start? Just toss some boxes in someone else's colos and build your customer base from there.
In the same way, disties often perform basic VAR work, and the line between VARs and MSPs has gotten very blurry over the past few years. VARs and MSPs have both been evolving into CSPs in an attempt to pull some of that subscription revenue away from the large public cloud providers. More importantly, everyone is desperate to hang on to ownership of the customer relationship.
Microsoft's 'Cut Your Own Throat' Strategy
Ownership of the customer relationship is king, and this is something the large public cloud providers know all too well. If you carefully examine Microsoft's partner "strategy," for example, you'll quickly see that Microsoft is strongly encouraging their channel partners to cut their own throats.
Microsoft has pulled virtually all incentives for channel partners to sell on-premises software. Similarly, they have jacked up the Service Provider Licence Agreement (SPLA) rates and difficulty, making it harder for those who want to rent access to Microsoft software as CSPs.
If a Microsoft partner wants to make any money in the short term they must migrate their customers to Azure and/or Office 365. The channel partner will then get a small percentage of the public cloud revenues, decreasing over time until it's irrelevant.
The pittance channel partners get from being the Microsoft "partner of record" for a public cloud account isn't enough to run a company on in the long term. This means that in order to survive, a Microsoft channel partner has to go find more companies and move more of them into the public cloud.
This continual shovelling of organizations into the public cloud inferno is unsustainable: in relatively short order the channel will have moved everyone that wants to go public cloud into the public cloud. By feeding Microsoft their customer base, they will have survived a few extra years, but at the cost of a lack of a long-term business model.
The customer, of course, doesn't need to have a "partner of record." They can simply be a Microsoft customer without being a channel partner. Of course, without channel partners customers will have a hard time finding, buying and integrating solutions from multiple vendors, or creating a "hybrid" environment in which there are both on-premises and public cloud elements of the customer's IT.
Of the big public cloud providers, Microsoft is the one with the strongest hybrid cloud "story." Despite this, Microsoft is emphatically not interested in supporting customers continuing to deploy workloads on premises. If, for whatever reason, a customer no longer has a channel partner they can work with and is interacting directly with Microsoft, they are going to push public cloud offerings as the solution to everything.
Amazon, Google and IBM are no different. Having a customer's primary relationship directly with a public cloud vendor is akin to having an individual read their news only from a highly partisan extremist news site: everything becomes monofocused, outside ideas are rarely discussed, and when they are they are often vilified. On-premises becomes "legacy," and everything from the development languages chosen to the details of what "as a service" technologies are chosen become an almost religious "old vs. new" debate, with the "old" being treated as purest evil.
The Practical Channel
In tech, homogeneity leads inevitably to a mantra of new being important for the sake of newness. Practical considerations, economics, or even functionality become irrelevant. Only newness matters.
In the short term, this means great things for customers. The channel is in a frenzy trying to redefine itself and make arguments for why it should even still exist. Hardware vendors and Independent Software Vendors (ISVs) who sell solutions that don't rely on the public cloud are also scrambling for relevance. Great deals can be had with only minor haggling.
Our choices now, however, will determine the long-term viability of the channel as a model. The channel enables a vibrant ecosystem of competing hardware and software vendors tied up with local and regional service providers that deliver everything from a hands-off retail experience to truly white glove service. Even with recent consolidations, the channel is large enough for there to be multiple disties -- both big and small -- on every continent, with thousands of VARs, MSPs, CSPs and colos to choose from.
If too many organizations become "cloud only" or "cloud first," the channel won't be able to sustain itself. It will either collapse into a small oligopoly of disties that raise prices and drive down quality, or vanish altogether. Most likely one followed by the other.
There is room for some channel partners to evolve into effectively becoming MSPs that manage your public cloud workloads. Instead of today's MSPs that largely "handle the hardware" so that sysadmins can focus on software and management, they would effectively become outsourced IT departments that replaced internal IT teams.
Partially or fully outsourced IT of that magnitude goes in and out of fashion, but MSP plus public cloud provider offers a powerful -- if ultimately expensive -- combination that many organizations may find hard to resist.
Implications for IT Practitioners
If you're an in-house IT practitioner and you want to keep on being an in-house IT practitioner, it behooves you to give some real thought to where you spend your budget. Engaging with the public cloud providers is often a quick solution to an immediate problem, but every workload put into the public cloud is a nail in the channel's coffin. The right channel partners can probably provide you what you need for about the same price. It's essentially the same argument as shopping at the local book stores to prevent the encroachment of Amazon, or the local anything the prevent the global dominance of Wal-Mart.
The uncertain question of the channel's future also provides insight into what to do with our career development. Keeping a careful eye on public cloud adoption rates for various market segments can tell us who is likely to still be hiring in-house techs, and what verticals the channel will abandon first.
If you see the channel pulling out of a given segment -- education, for example -- because a public cloud provider like Google is coming to dominate that segment, this provides clear signals for career choices. Either get really, really good at the public cloud provider dominating that segment's offerings, or migrate out of that vertical entirely.
In the medium term, learning how to work with public cloud offerings enough to work at the public-cloud-focused MSPs is good money. There's 10-15 years' worth of solid work in migrating organizations to the public cloud and then wrangling their workloads once there.
Longer term, the planetary demand for operations teams either has already or will quickly plateau. The need for infrastructure nerds is diminishing. If you still want to work in IT consider shifting careers into development or robotics.
The Pendulum Swings
The pendulum will eventually swing back to on-premises IT again. The centralized-decentralized cyclical nature of our industry makes it inevitable. The unknowns are how much of the channel will make it through the eye of the needle, and how completely on-premises teams will collapse before that happens. If you want to know more, find an old mainframe greybeard/greybraid and ask them for the tale of their career.
Trevor Pott is a full-time nerd from Edmonton, Alberta, Canada. He splits his time between systems administration, technology writing, and consulting. As a consultant he helps Silicon Valley startups better understand systems administrators and how to sell to them.