On-Premise can't win (and neither can some cloud providers)

On-Premise can't win (and neither can some cloud providers)

Summary: The economics are getting clearer. Individual IT shops can't acquire and operate hardware with the same scale and economy as larger cloud providers. But, not all providers are the same. Some are clearly driving costs the right direction.

SHARE:
TOPICS: Amazon, CXO, Hardware
21

Can your IT solutions deliver 19 consecutive price reductions in 6 years?

Scale is a beautiful thing. Companies can use scale to create more streamlined processes, greater efficiency, lower per unit costs, etc. What these firms do as a result of these operational godsends is up to them. They can choose to:

  • Increase dividends paid to shareholders
  • Reinvest the savings into new product development, new plant & equipment, geographic expansion, etc.
  • Pass along the savings to customers

Companies often choose the latter when they operate in highly competitive markets and/or when they want to increase market share.

Recently, Amazon announced they are reducing prices for their AWS offerings. This is the 19th consecutive price reduction they've made in 6 years.  (see: here )

It's interesting to note the trend in play here. Amazon, along with other cloud providers Google and Microsoft, is lowering costs to customers of its cloud services. They are clearly dropping prices (because of scale efficiencies and dropping commodity hardware prices) to garner more customers, that will drive more volume and present more opportunities to scale, which will eventually generate more opportunities to lower costs further.

These vendors get it. They can grow these businesses further by relentlessly focusing on reducing costs. IT departments generally can't do this. Their IT footprint is what it is. Their scale opportunities are capped. And, as a result, they are facing a no-win situation. Cloud providers will likely continue to widen their cost advantage over in-house technology. It's simple math.

Smart CIOs already see this. And, the smarter ones are seeing that not all cloud providers are the same. Some, with extensive captive hardware businesses or proprietary software, just aren't cost competitive. Their scale opportunities are limited. Worse, some of them are years behind Amazon, et.al. The big boys already have large scale operations that are getting bigger and cheaper. Latter entrants don't have scale and will have a tough time building scale especially when they won't be price competitive for some time.

Who wins or loses in the new IT economy? Well, it looks like:

Winners

  • Cloud vendors of scale
  • Cloud vendors that embrace very low cost hardware and open source software
  • Early cloud provisioners
  • Cloud providers with low cost access to electrical power

Losers

  • Cloud vendors who'd prefer to sell you on-premise hardware (of their making) instead
  • Cloud vendors that make brand name hardware and proprietary software
  • Late entrants to the cloud space
  • Cloud providers without a focus on low-cost locations for cloud centers

What I didn't put on the list of winners and losers was anything to do with existing customer base.    Vendors with extensive CIO connections aren't necessarily in possession of any real advantage here. Yes, they may have long-standing contractual relationships with CIOs but will these ‘relationships' really carry the day? I don't think so.

CIOs are exceptionally logical. They also know how to shop. AND, contrary to what some software or hardware sales pros think, your firm just isn't that strategic to those CIOs. That's right - you won't win future business with them just because you did so in previous years. A while back I was contacted by a long-time software sales pro who was complaining that he could no longer get meetings with his CIO customers as his employer was no longer considered to be a 'strategic' vendor.  The leader board may be changing and some of the old guard vendors could find their previous place of importance has undergone a demotion.

If these sales pros want the possibility of bidding on these deals, they'll need to:

  • Really come with a sharp pencil
  • Prove they can continue to produce future price reductions (a la the ones mentioned previously) and be willing to put these reductions in writing
  • Convince prospective buyers that they can quickly catch up to Amazon, Google and Microsoft on scale, cost AND technical abilities
  • Change their stripes and become big users of open source, lower cost software (not of their own making)

This should be a fun time .....

Topics: Amazon, CXO, Hardware

About

Brian is currently CEO of TechVentive, a strategy consultancy serving technology providers and other firms. He is also a research analyst with Vital Analysis.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

21 comments
Log in or register to join the discussion
  • So what you are saying is that ...

    Microsoft will fail, unless they go open source with all their cloud software.

    I am sure Ballmer will be really happy to hear that. ;-)
    D.T.Long
    • I doubt Ballmer (or many) even cares what this blogger says

      Sharepoint was supposed to fail, yet billions of dollars worth of sales later, and getting India on board with their cloud based offerings.

      Just because you write it doesn't make it true. The blogger leaves out the fact that not every cloud needs to scale to the size of an Amazon, MS, Google, ect.
      William Farrel
      • Sharepoint is a failure

        The only reason it "Succeeds" is because MS gives part of it away with Enterprise Agreements. They hook you in and then by the time you realize how cumbersome it is you have to go forward.

        It sure isn't a good solution nor is it the cheapest but like Windows, the initial cost is low but the TCO is HUGE.

        Wasn't it MS that said for every $1 you spend with them on Sharepoint you spend $7 with others making it work?

        Yeah, a quality product right there.
        itguy10
      • LOL! I'm sure there are a ton of businesses that wish

        their products would "fail" like Sharepoint, itty!!

        Please, tell us that great "bed time story" again, where MS's developers and techs came onsite and could not get your Sharepoint server to work, ever!!

        That was always an amusing fairy tale!
        :)
        William Farrel
      • Try to apply a tiny bit of common sense

        Open source cloud: SW very cheap or free, hence service has lower cost

        Proprietary cloud: SW expensive, hence service has higher cost

        I guess such a simple analysis is beyond you, but MS fanbois have never been known for rational thinking.
        D.T.Long
  • It's not just about cost

    Public clouds are akin to public transportation. They make sense to those with limited resources. That is why small business is gravitating towards them.

    The server market is not just about saving money, it is about providing companies with a competitive edge. Windows Server 8 does so many things that are simply not possible with Open Source Software. Companies will spend money to acquire these capabilities.

    Also, server ecosystems, which include a substantial segment of individually owned equipment, will generate more wealth than those which don't: just as a transportation system where a lot of people own their vehicles, generates more wealth than one where they don't.
    P. Douglas
    • Utter nonsense

      Efficiency and productivity generates wealth. Who owns the equipment is irrelevant. Writing your own SW when you can use SW written by others for free to achieve the same results is a complete waste of resources and puts you at a competitive disadvantage, but I do not expect the average fanboi to grasp that well established and simple fact.
      D.T.Long
      • You missed the boat by *that* much

        First off, unless the situation being discussed is "what server software does Microsoft use for its in-house needs?", your comment about "writing your own SW" makes no sense, since only [b]Microsoft[/b] would be the ones "writing" the Windows Server 8 software P. Douglas mentioned in his comment.

        Secondly, his claim is that Windows Server 8 software provides features to its users that aren't available with open-source alternatives. So, before saying that open-source software can "achieve the same results", you might want to consider providing at least [b]one[/b] example before saying he's wrong.

        And finally, given the lack of examples in your response, yours has the greater "fanboi" feel to it...
        spdragoo@...
  • Cloud is expensive and slow

    I suspect there are lots of customers for whom public clouds are just fine. However, when I recently looked at what it would take and cost to move my company's operations to the cloud, it was going to be very expensive up front, expensive on-going, and suffered unpredictable latency, which our applications could not withstand. If you know how to operate at web-scale on a budget already, the cloud solves nothing. There are plenty of open-source recipes out there for doing this.
    ncted
    • Yeah but

      Your solution won't help out poor little Microsoft push their Window 8 server with its "Meto" UI. Because we all know that a server needs a cellphone inspired UI.
      Jumpin Jack Flash
  • They are clearly dropping prices (because of scale efficiencies...

    How on earth can you claim this? I am so sick of the wild assumptions ZDnet makes and then tries to pass them off as fact. I'm switching to InformationWeek. At least there the staff acts in a professional manner without all this wild hype.
    happyharry_z
    • do not go

      One reason for these articles is to get people to respond. Your response made me think... if you did not respond, would others have thought in return?...
      HypnoToad72
  • Wrong. Wrong. Wrong.

    "The economics are getting clearer. Individual IT shops cant acquire and operate hardware with the same scale and economy as larger cloud providers."
    Agreed - but that is because of the PRICING POLICIES of the original vendors - not the inability of the purchaser.

    "But, not all providers are the same. Some are clearly driving costs the right direction."
    Agreed - but the main people keeping costs at their HISTORICAL HIGH are the vendors and providers.

    What Sommer fails to understand is that the PRICE of (e.g. VMWARE) licensing is what keeps the little man in check, not his system design capability or operational competence.

    "Can your IT solutions deliver 19 consecutive price reductions in 6 years?"
    I don't have to: I just sit back and watch Intel's tick-tock ... the constant reduction in disk prices ... the rapid reduction in networking costs.
    THE ONLY THINGS HOLDING ME BACK ARE THE CONSTRICTIONS DELIBERATELY AND SYSTEMATICALLY FORCED ON ME BY MAJOR VENDORS.

    "Scale is a beautiful thing. Companies can ..."
    Indeed ... but all developments which exploit scale for the benefit of end customers are outlawed e.g. one license per PC, one connection per household, the bittorrent protocol, the number of VM's per processor/RAM, ...
    Companies can ... but inevitably choose to keep most of the benefit for themselves and shareholders.

    "Recently, Amazon announced they are reducing prices for their AWS offerings. This is the 19th consecutive price reduction theyve made in 6 years. (see: here )"
    So what? I've heard of Moore's law too! [b]So friggin' what![/b]

    "Its interesting to note the trend in play here. Amazon, along with other cloud providers Google and Microsoft, is lowering costs to customers of its cloud services. They are clearly dropping prices (because of scale efficiencies and dropping commodity hardware prices) to garner more customers, that will drive more volume and present more opportunities to scale, which will eventually generate more opportunities to lower costs further."
    Supplier costs drop - passes on reduction to customer - I consider this UNINTERESTING.
    A purely business perspective - I am a customer - where is the customer value proposition? Sommer omitted it altogether ... that's why his post is so lop-sided and in error.

    "These vendors get it. They can grow these businesses further by relentlessly focusing on reducing costs. IT departments generally cant do this. Their IT footprint is what it is. Their scale opportunities are capped. And, as a result, they are facing a no-win situation. Cloud providers will likely continue to widen their cost advantage over in-house technology. Its simple math."
    IT departments can't do anything because the cost of licenses is artificially high.
    There is a natural reluctance to sharing and fedeartion ... but is it really any different to multi-tenancy?

    "This should be a fun time ..."
    ... not for most: the usual suspects will increase their stranglehold, while blind pundits like Sommer applaud. Losers all.

    I've waded through Robin Harris's abortive iMAC purchase; Kingsley Hughers crap Photoshop power PC (without SSD); Bott's Windows Home Server support (deprecated by the simple addition of the Spaces facility in Windows 8, following the development of RAID in 1988) ... is there ANYONE on ZDNET who has done computer system design ... even for a few days?
    jacksonjohn
    • Arithmetic

      OK, so I can buy a mid range DELL quad-core PC with 300GB disk for ??450.
      I can buy Office Home and Student for 3 people at ??85.
      For arguments sake let's say I could share an equivalent cloud computing facility between 3 people and that I spread the costs over 3 years.
      That's ??535 divided by 9 pa, say ??60 pa or ??5 per month.

      True, I haven't included power and operations maintenance ... but you clever cloud people are going to be so effiecient at that it will only cost me ...
      ... well you tell me then, what's your best offer Sommer? You, for whom 'the economics have become clearer' but who includes not a single calculation in his whole presentation!

      And if you can match my costs with a cloud facility ... upon what endpoint I am I going to run it? And how much will that cost?
      Presumably not a iPAD ;-)
      jacksonjohn
  • funny

    For all of you hellbent on returning to server-side computing (minus SNA and 3270 emulation thanks to TCP/IP) more power to you. However you must realize you can only horizontally scale so long before wintel/microslyme TCO will surpass TCO of industrial-strength processing platforms. This fact is demonstrated in the financial industry. Consider your use of a debit card if there was a microslyme platform providing your card authorizations for funds disbursements. Cloud is all just a little bit of history repeating ... Use industrial-strength horsepower platforms.
    WSguru
  • Cloud is not always an option.

    At least not a very affordable option, especially if you are in a location without much access to high-speed Internet. As an example, my employer is paying a very large monthly bill for bonded T1s with a total speed of 3Gbps while I am able to get 20Mbps at home through the cable company for a fraction of that; the reason being that the plant is in a less populated area and T1 is all that's available. Sure we could order more T1s bonded into the current pair, for an even more ridiculous monthly bill.

    Local infrastructure, servers and storage are our only option right now. Maybe someday, when bandwidth is more available, we can consider cloud.
    PepperdotNet
  • Depends on scale

    For SMBs, cloud makes perfect sense. There is no way that an SMB I/T shop can compete with a big cloud provider on efficiency.

    Large enterprise on the other hand can still easily generate the scale needed to match cloud offerings with their own internal solutions and you don't have to pay someone else for the privilege.
    SlithyTove
  • By the way

    It's "On-premises" and not "On-premise". That one for sure can't win.
    MattPV
  • BTW beat me to it!

    [b]premise[/b]
    [i]noun[/i]
    the premise that human life consists of a series of choices: proposition, assumption, hypothesis, thesis, presupposition, postulation, postulate, supposition, presumption, surmise, conjecture, speculation, assertion, belief.

    [i]verb[/i]
    they premised that the cosmos is indestructible: postulate, hypothesize, conjecture, posit, theorize, suppose, presuppose, surmise, assume.

    The verb use is not common in [i]Australian[/i] English, but Septics are forever *[i]verb-izing[/i]* nouns, so it will be added here too eventually (ah, the cultural [and economic] imperialist that is the USA!).

    [b]premises[/b] |[i]premsz[/i]|
    [i]plural noun[/i]
    a house or building, together with its land and outbuildings, occupied by a business or considered in an official context : business premises | supplying alcoholic liquor for consumption on the premises.

    [b][i]From OSX Dictionary[/b][/i]
    mediumcool
  • No such thing as "on-premise"

    My college professors were infuriated by people incorrectly using the word "premise". We were told that we would fail the program if we used "on-premise" to refer to on-site hosting...

    http://en.wikipedia.org/wiki/Premises
    Bit-Smacker