back to article Why I'm sick of the new 'digital divide' between SMEs and the big boys

Recently I have been spending most of my time with enterprise CIOs and vendors' "product owners"; their dismissiveness of the needs of small and medium-sized businesses has finally got to me. As a rule, talking about the needs of corner cases is boring to the masses, but today I just need to get up on my soapbox and talk a …


This topic is closed for new posts.
  1. Canopus

    Could not agree more

    Hear, Hear ! I clap with all three hands (having had the bad luck of being designed by an armchair system architect) and salute the author !

    He is one of very few voices that seem to remember that - under the herds of spherical IT cowsultants and the mile high layer of spherical bovine excrement they have been trying to bury us under - there is one basic down to earth concept that should underlie any IT project (drumroll please) - it's called : fit for purpose !

    1. Anonymous Coward
      Anonymous Coward

      Re: Could not agree more

      Having heard the phrase uttered "I'm not a technical specialist in this field" repeatedly from architects and tda's this year when raising my concerns about aspects of solutions, I feel your pain. They're designing the entire thing, SURELY they should be the specialist in that field!

      1. Tony-A

        Re: Could not agree more

        Automated ignorance.

        And it's no joke.

      2. Roland6 Silver badge

        Re: Could not agree more

        >"I'm not a technical specialist in this field" repeatedly from architects and tda's ... They're designing the entire thing, SURELY they should be the specialist in that field!

        No as a TA/TDA, I create and adapt the solution 'framework' that provides opportunities for the specialists to make meaningful contributions. As (I hope) a good TA, I use their input to adapt and enhance the solution and make compromises where we find that the available products just can't do what the client wants them to do.

        I've found two major challenges working in the SME space after the enterprise space. Firstly 'Enterprise' solutions are generally an overkill, however SME's still need many of the features eg. business continuity, but it needs to be "Right scaled" and "right priced". Secondly, you don't get to deliver stuff with a large team, so get prepared to learn quickly.

    2. xyz Silver badge

      Re: Could not agree more

      As an IT consultant (don't laugh) the thing that gets me is that for all the talk of giving the client what he wants, which usually entails tearing him a new one with the cost of hardware or infrastructure and packs of "Agile" developers with their monthly sprint group wanks who spend most of their time CV programming rather than bloody finishing something. The key thing everyone seems to talk about and immediately forgets is that the customer wants "Shit That Works" (STW) without having his pants pulled down. The bloke just wants stuff that works and no one seems capable of delivering the aforementioned shit without major cost additions or delay. Yet everyone tells me they are experts. Does my head in.

      Oh and with reference to...

      .. governments sit on payments to suppliers far longer than they should, draining capital from smaller organisations.

      yup, I've just had to wait 4 months to get an big invoice paid. Nearly blew my cashflow out the water.

    3. jason 7
      Thumb Up

      Re: Could not agree more

      Fit for purpose! That's why with any IT project or requirements findings, I always started with the guys in the post room and worked up.

      Most just speak to the CIO/CEO and never bother going any further.

      Oh and remember it's not about giving the customer what they want. It's about giving them what they NEED.

      Folks forget that too.

      1. Nick Ryan Silver badge

        Re: Could not agree more

        Thank you for remembering that it's really about what customers NEED.

        Far too often I've been given what a customer wants, and spent (a relatively) a long time with them separating their requirements from their chosen solutions to get to the bottom of what they actually need. Usually they've been suckered by promises of automation solution nirvana (MS, IBM, Oracle, etc... they're all very guilty of this) where they lose sight of what they need. I've only ever once come across a client that where I've neatly redefined it like this didn't they appreciate the distinction... maybe I've been lucky or it's the way I've pitched it.

        On a few occasions I've had to challenge the customer to run their newly designed processes, that they want to automate, on paper first. While this may sound odd I've found that if a relatively small company can't run their new (basic) processes on paper, there's usually no hope of them ever running them on a computer system either. Naturally testing a scheme, even on paper, really shakes down requirements as well.

        1. jason 7

          Re: Could not agree more


          Indeed one of the first things we'd do on a new delivery project would be to get a length of brown paper about 40 feet long and 4 feet wide, a pile of post it notes and felt tip pens. Then we'd get a couple of Johnny Frontline staff and sitt for a week mapping out the entire business process that we were tasked with improving etc. from start to finish and back again.

          After that week or so we'd know the process back to front, often finding some big flaws (even to the surprise of the staff) and places for improvement. Then we could build up a process to improve upon and refer back to the chart over and over.

          So at the end of it all we would deliver a system that would cope with all the situations that could occur. It worked a treat and saved a lot of time and money later on.

          However, over the years, new hotshot pink shirted project managers/contractors would arrive and they all laughed at my brown paper charts. Cue a lot of failed projects that just focused on one part of the over all process and didn't join up with any other part much to the annoyance of the customer.

          All too quick to buy the lovely shiny hardware, didn't want to actually speak to the poor buggers who would be using it.

          Still they all got paid a lot for failure.

  2. Anonymous Coward
    Anonymous Coward

    So what is "proper" IT design?

    "proper" IT is about matching the solution to the problem and that it rarely ends up being the same exact solution twice

    It seems you already know the answer to your question Trevor.

    Do try to calm down a bit, some of that sounds like you were very nearly foaming at the mouth.

  3. ElNumbre

    Fresh out of the Box smell

    New Enterprise equipment tends to be priced in the multiple 000's when a lot of it is some fairly standard electronics with a bit of custom software loaded onto it. But generally you're not paying for the tin, you're paying for a large team of developers, sales jocks, middle managers and executive golf days, plus a shiny building in the South East.

    If you need something that is 'enterprise' quality, but can't afford the fresh out of the box smell, how about the 'last-generation' version? Depending on what you're buying, you may be able to get several of them in case one breaks and you can't get support on it. The only spanner in that works is where something needs to be continually licensed, and the OEM refuses to issue them. But then, many third parties will still happily sell you support and maintenance packs if its a product they can support.

    The other option is off the shelf tin coupled with 'liberally' licenced software. A lot of enterprise products have similar options created by the community but at a lower (or even nil) upfront and ongoing cost. Interestingly, the only thing I've not encountered as a viable alternative is WAN acceleration - the FOSS equivalent to Riverbed et al. These solutions may mean you have to have support people capable of administering a Linux or BSD solution, but this is all about compromises.

    1. Anonymous Coward
      Anonymous Coward

      Re: Fresh out of the Box smell

      Yes, that's what we do (we're a charity that doesn't do fund-raising... budgets are small).

      Recent "new" technology has been:

      * a number of HP ML110G7s from when they had a killer cashback offer last year (£120 per server, at the end of the day)

      * Samsung 830 SSDs, similar cashback deal just before the 840s came out.

      * previous generation managed access points bought off ebay

      * Argos refurb. laptops

      Having a highly constrained budget is challenging - but as long as there is *some* budget, it is challenging in a good way.

      Servers are all Linux, the time saved on administering CALs is spent on bargain-hunting.

      1. keithpeter Silver badge

        Put a page up... Re: Fresh out of the Box smell

        "Yes, that's what we do (we're a charity that doesn't do fund-raising... budgets are small)."

        Suggestion: Stick a Web page or blog post up somewhere with a diagram and basic (non-specific) details of what the system does, how the users use it and a rough idea of how much it costs you to run. Just an outline. So proprietors/managers know what can be done, and technical people get a pointer.

        I think that charities could play a part in documenting lower cost/good fit systems as they do not have quite the same competitive pressures that SMEs (as we call them in the UK) do.

        The end-user tramp: I keep thinking about buying a new PC. But this ancient HP workstation just keeps chugging away...

    2. ukaudiophile
      Thumb Up

      Re: Fresh out of the Box smell

      Excellent, valuable advice that should be taught and echoed at every opportunity.

      Just because it's last year's (or 2 years old) doesn't mean it won't be fit for purpose and resolve the problem you have.

      Likewise the idea of using lower or zero cost community versions of enterprise systems is a good one with, as is mentioned, the caveat that you need the technical people capable of implementing and supporting it.

      It's sad that too many people with 'Director' and 'Manager' in their title don't realise that the product to fix their problem doesn't always come from someone wearing an expensive suit, a designer watch and who takes them for £100+ lunches.

  4. Anonymous Coward
    Anonymous Coward

    Whatever happened to "rate this article"?

    And in its absence, perhaps you'd like to upvote this if you liked that article.

    Upvote this multiple times if you liked it a lot (hmmm, might be tricky).

    I liked it a lot, but I can't vote at all.

  5. Don Jefe


    Good article!

    Everything you describe is a direct effect of marketing. Everyone has been sold on the 'five 9's' idea and efficiency and ROI but truth be told the majority of SME's don't need that, but they've convinced themselves they do because they're constantly inundated with industry literature, web articles, and forum comments telling them they'll go out of business if they don't buy/do (x).

    Enterprise level product advertisements aren't targeted at enterprise decision makers. They are targeted at SME's in order to move them up a few steps from bare bones deployments. Actual enterprise marketing is handled entirely differently. A side effect of all this is the same marketing used to drive sales in the SME market inflates the egos of the actual enterprise crowd and gets them talking about how great (x) is. It boils down to extraordinarily effective free marketing.

    These strategies are obviously loosing steam, as evidenced by dismal industry sales figures, but I expect the fractious nature of the two groups to get worse before it gets better. I predict the industry will move in a more truly compartmentalized direction where the SME level customer won't even know the enterprise stuff exists. As in the auto industry I reckon they'll erect barriers to entry (or even viewing) that will keep the groups truly separated. Right now the edges of all the marketing and sales overlap and I expect that to end.

  6. Anonymous Coward
    Anonymous Coward

    Point 1 of that Redmond Mag link made me nearly explode with rage.

    Some users can still barely work their way around a UI that's been standard for 18 years. You can find so many examples on the web of people that get stumped by the new UI ... hell, I'm an IT pro and even for me, the logic behind some of the design decisions are incomprehensible to me.

  7. PyLETS

    compromising cheap with reliable

    I've run an interesting variety of small charity and community organisation services, some Internet standards based, some custom webapps and sites, on a £15/month virtual machine running Debian stable which has dist-upgraded successfully for 10 years now with better than 99.9% scheduled availability (i.e. no more than 4 hours average unscheduled downtime per year over that period.

    If I wanted better than that, I'd rent 2 of these on different networks with different providers and figure out how to synchronise them and do round robin DNS. But that would be an extra level of complexity I don't need. The main cost is the time of yours truly who admins, develops and maintains it all, but the knowledge gained is what I sell, so it's effective revenue not cost as far as my own bank balance is concerned.

  8. Anonymous Coward
    Anonymous Coward

    margins etc,

    Completely agree. I have some other perspectives that may help.

    Firstly look at the suppliers margins, the higher they are, the less you will get for your money. The high margin model, IBM etc, only work at the very big end, but they (IBM for example) try to apply it at the SME level and so you get nothing much for your money. I liken it to battle ships and row boats. Need a battleship? go to a battleship maker and you of course will not be at all concerned about the price. Want a row boat? and you get it from a battleship maker and you will get a very poor deal. Get if from a specialist row boat maker and you will get a great deal and a great row boat. Do a search on IBM and margins. They boast about them. In the days of EDS, their margins were very low but their expertise was very high. EDS provided a LOT for your money. These sorts of firms, like EDS get swallowed up by the likes of HP, another high margin operation. Even so, the cloud is genuinely lowering costs for those who can stand the outages etc. but it is some what increasing risk for many. There usually are operators who can provide a great deal for the SME, but if your CIO thinks he has to pay big money to the battleship makers, I think he is in the wrong job and will contribute to or cause the failure of his firm as you appear to imply. Of course in Government, it just contributes to massive cost blowouts and failed projects. Governments don't often go broke and the high margin boys love them.

    1. TeeCee Gold badge

      Re: margins etc,

      EDS provided a LOT for your money.

      Hark! What is that sound I hear? Sounds awfully like the public sector IT lads having a collective attack of apoplexy.

      1. Anonymous Coward
        Anonymous Coward

        Re: margins etc,

        "EDS provided a LOT for your money"

        Ha ha ha ha ha (etc)

        A/C as a former contractor on an EDS-led project that was the worse managed project I've ever seen. Programme Manager in question was later seconded to a large public sector IT project which failed with spectacular losses. Was he sued for recovery of his "earnings"? Was he hell ...

        Actually the thing that really winds me up on the "divide" between SMEs and big boys is when you see large vendors pushing "Small to Medium" solutions (don't get me started on that word) that are just way out of the budget of any genuine small to medium business (average size fewer than fifty people, and in some areas, fewer than ten, so why even dream of saying a £50k platform is suitable?)

        1. jason 7

          Re: margins etc,

          Well my other half worked as a civvy in the Police Force for a while. She was given the job of keeping an eye on the many projects going on at the time. So the usual thing was knock on the door of some 50 something manager, ask him how the project was doing to be told -

          "Why are you bothering me? It's a Govt project, we'll just write up a report the day before its due telling them what they want to hear!"

          The other oft used phrase to any project that was just plain mad/incomprehensible/pointless was -

          "It doesn't's only tax payers money!"

          Yep, that's going on all over the country folks. They really don't give a sh*t.

          At least my other half had the decency to hand her notice in after 6 weeks. She couldn't tolerate the thought that she might end up like them too.

    2. Denarius Silver badge
      Thumb Up

      Re: margins etc,

      @AC, for once I agree, at least in southern hemisphere. EDS staff where I worked tried very hard to deliver working solid systems. I was pleasantly surprised. The PHBs count was low, so many skill sets were run effectively. Of course, the PHB power of overseas offices was also a clear and present danger. Then HP bought EDS out. The usual gutting commenced.

  9. GreenJimll

    Enterprise often means half baked

    The marketing folk and salesmen seem to say "enterprise" when they want to stress something is rock solid reliable, scalable to large deployments and has a long supported shelf life. When we actually buy "enterprise" gear off them we often find its half baked code that needs multiple revisions to get it anywhere near stable, the scalability of their idea of "enterprise" doesn't actually match the size needs of our "enterprise" and support is a twisty maze of support calls all the same.

    So if you're doing IT development or support for SMEs I wouldn't be too quick to run to spend cash you don't have on "enterprise" solutions anyway: they might not be as five-nines as you'd hope they'd be for the money.

  10. Frances Banana
    Thumb Up

    Good article

    Being an owner of a SME that lives off making custom solutions for customers - I must admit - this article is sooo true! Bigger players and marketing zombies work from a "white paper to a white paper" - and it resembles pretty much a crusade when somebody in the group has the guts to say "amm, but maybe we don't need a canon to shoot this fly?".

    The last "white paper attack" happened to me last week - I was called "old fashioned" and "not up to the 21st century" simply because I had the balls to question a bunch of "CIO/CEO/whatever" architects from a much bigger player (that thing called "common sense" and "let us do some testing").

    Small companies cannot afford downtime - and small companies delivering service to other small companies cannot afford to deliver crap service - because then it's a quick decision and you lose a customer.

  11. Anonymous Coward
    Anonymous Coward

    What's real and what's for sale.....

    "Competence" in our industry has become synonymous with regurgitating corporate messaging and deploying the newest solutions instead of those most fit for purpose."

    Amen to that! It reminds me of the legacy the mocking US motto--- 'What's real and what's for sale'.....

  12. oiseau
    Thumb Up

    Bravo !

    "I believe that "proper" IT is about matching the solution to the problem and that it rarely ends up being the same exact solution twice. Your mileage may vary. Debate, as always, in the comments."

    Very well said and quite acurate asessment of the actual state of affairs.



  13. Destroy All Monsters Silver badge

    14 Reasons To Fire Your IT Staff

    Reason 6: An administrator offers virtual desktop infrastructure (VDI) solutions without understanding Remote Desktop Services (RDS). RDS is a cheaper, denser, simpler version of VDI that's also nearly two decades old. RDS doesn't work for every application -- but it does work for most.

    Could links to entitled pomposity blogposts be specially marked? "Not Safe For Rage" for example?

    Greg Shields is a senior partner and principal technologist with Concentrated Technology. ... Greg has been a multiple recipient of both the Microsoft Most Valuable Professional and VMware vExpert award.

    I bet he did. I bet he did...

    1. frank ly

      re. 14 reasons ....

      I think Reason 1. was thrown in by Steve Balmer at the last minute.

    2. Trevor_Pott Gold badge

      BY the same token, I'm a vExpert as well. Shockingly, I'm not an MVP....

    3. Danvighar

      14 reasons

      Wow. A lot of those are, well, stunningly ignorant. Most of the valid complaints are better addressed to security or the beancounters, and the rest are based on having an obsession with Teh Cloud!!!!111!!!eleventy or "Nothing could possibly go wrong with just giving me what I want. You have monitoring tools and backups! Who cares if the beancounters wouldn't buy the tools you need, since they could find a 'functionally equivalent' POS."

    4. John Smith 19 Gold badge

      Bear with me because "I'm no expert....."

      But he does rather sound well up his own colon.

      10% possibly useful information

      90% Buy new MS stuff.

      Possible contender for an opponent for Eadon in a cage fight deathmatch?

      1. Don Jefe

        Isn't buying new stuff the way to fix everything?

  14. Fenton

    Don't shoot the messenger

    The biggest problem is nobody wants to spend the money up front during the design phase and get all of the "relvant" subject matter experts in to do the design.

    Just because I'm an Application specific technical architect with a bit of HW knowledge it does not mean I can architect and entire landscape, from storage, servers, networks to resliant DR fabric.

    I can scream all I like "I need a storage Architect so work on the Array architecture", all you get is.

    Sorry no budget. So out comes the "white paper" and if you can find it the latest manual.

    Can I try things out whilst I'm designing.

    "No you're an Architect, you're not allowed to actually touch anything, only the engineers"

    Alas the Engineers are all stuck in the far east and have never seen the latest and greatest piece of kit that was ordered. They also want step by step instructions on how to configure the beast, just giving them a rule on what needs to be configured is not enough. If they had their way they'd have the architect writing every single command step by step so if it goes wrong it's not their fault.

    1. John Smith 19 Gold badge

      Re: Don't shoot the messenger

      ""No you're an Architect, you're not allowed to actually touch anything, only the engineers""

      Time to set up that home test lab?

      1. Trevor_Pott Gold badge

        Re: Don't shoot the messenger

        You mean the test lab that metrics say shouldn't be served by the now defunct technet or the rapidly evaporating Small Business Server?

        Good luck with that. I think it's time to move on.

  15. Marco van Beek

    You have to laugh when...

    Some virtual server solution salesman gets you on the phone and tries to explain the benefits of virtualising all a client's servers after finding out they only have one!

    1. Anonymous Coward
      Anonymous Coward

      Re: You have to laugh when...

      equally funny when the cost of the comms required to make it work as well as the old dell box sat in the corner is twice as much as upgrading that box - every year!

  16. Anonymous Coward
    Anonymous Coward

    You simply occupy a different space to these people

    I agree with much of what Mr. Pott has to say, but the provider to SMEs inhabits a very different world to those who sell into large businesses.

    If you actually work for an SME, and find the Dell laptop you've been given is crap, or the mobile phone service is terrible value for money, you're probably sitting very close to someone who can and will listen/act. In my experience, your typical SME owner (and we're usually talking family businesses) is tighter than a duck's arse and has a passing interest in running a sustainable business.

    Add a few layers of bureaucracy and a few tens of thousands of employees, and the bonkers purchasing decisions creep in, usually over several rounds of golf. Looking at my own experience of working for a large business (just the one - we grew through being acquired a few times), I still don't know why we don't eat our own dog food. It's good, tasty dog food, proven in the field, and scalable from SME to carrier. If you can't get a business to eat good dog food, what chance do you stand if the dog food is off cuts and offal?

  17. Huw D

    FWIW, Trevor...

    I managed to get the ear of someone (very senior) at HP during Spiceworld London and pressed them about how they interact with the SMB/SME market.

    One of the things that vital to me is the ability to try kit out to see if it's fit for purpose. I have no budget to buy something, find out it's no good and scrap it - even if it's just a £500 laptop.

    If I'm in big business I can get stuff thrown at me. I can go to the likes of SCC/ComputaCenter and play with every toy under the sun. If I'm SMB I just have to hope I've made the right decision.

    An email conversation has been started. Hopefully something good will come of it.

    I think we all need to pressure the vendors for fair treatment.

    1. Trevor_Pott Gold badge

      Re: FWIW, Trevor...

      Well you did better than I at Spiceworld Austin, good sir. Mind you, the HP guys at Austin were printer dudes and the only other major was Dell...and Dell gives none of the fucks. Ichann believe they have bigger things to worry about.

    2. Don Jefe

      Re: FWIW, Trevor...

      The HP people should work with you, especially if you're wanting to trial off the shelf kit. If you want customized stuff it will be difficult for your sales guy to get it approved.

      That being said, make sure you have a solid test regime in place before you get too far down the road to a trial. Once you get a deal sorted that stuff will be at your door and the trial period will be over before you blink and you'll be on the hook for it and still have no answers. We have acquired more than a few things from different vendors that way...

    3. Anonymous Coward
      Anonymous Coward

      Re: FWIW, Trevor...

      I am in big business and I'm buggered if I can get free kit to play with off HP / SCC and co!

      1. Trevor_Pott Gold badge

        Re: FWIW, Trevor...

        :) I think you'll notice a distinct lack of HP in my lab too.

      2. Anonymous Coward
        Anonymous Coward

        Re: FWIW, Trevor...

        AC: I think thats cos cockroach PMs like where I used to work, would hassle SCC for "loan kit" and "temp licenses" on the understanding of a PO being raised, start developing and then cancel the PO; several times in fact.

  18. msg

    agreed. what about disaster recovery on a budget?

    good article and comments here. I've always felt pretty left out of the "white paper" class builds due to budget constraints, and yes, even a bit inferior whenever I have discussions with vendors who work with the larger businesses, and I have to talk about my affordable work-arounds on stuff. most are gracious, but yeah, can't help feeling like a bit like a side-yard mechanic sometimes.

    how do you guys address disaster recovery properly without budgets? I'm fairly impressed with AppAssure. it's $$$, but not as untouchable as this sort of stuff was several years ago. automated server and machine imaging is what draws me to it most.

    1. Trevor_Pott Gold badge

      Re: agreed. what about disaster recovery on a budget?

      I'm working on an eBook to this regard. It's too long for an article, much less a comment. ;)

      1. msg

        Re: agreed. what about disaster recovery on a budget?

        cool, I'd be interested to hear what you have to say on robust disaster recovery on a budget, Trevor. how can I stay in the loop with your eBook?

        1. Trevor_Pott Gold badge

          Re: agreed. what about disaster recovery on a budget?

          Follow me on Twitter? Cakeis_not_alie. I am sure I'll mention it there. It will also be talked about on The Register and WeBreakTech as well as I'm sure when I have finally finished it that you will hear the gnashing of teeth from the whitepaper brigade so loudly that it will haunt your dreams and prevent sleep.

          Then you will know it has been released.

          (Actually, If I publish an eBook on how to do the MacGyver-like paranoid systems design that I do, does it suddenly become a "whitepapered approach?" Will CCNA's heads actually explode? Things to find out!)

  19. Anonymous Coward
    Anonymous Coward

    A fer days ago I got a few comments about the fact that we are using OS/2 on our back end servers because it just works. What I didn't mention there was the fact that a lot of our hardware was obtained second hand when larger enterprises went tits up because they overreached themselves buying 'enterprise' solutions they couldn't afford as Trevor said.

    In fact we got so much equipment from one business that I was able to set up a community project at the cost of one of my technicians for two days and much of that time was spent putting in the network cables - the village supplied the ADSL, so no cost there. The servers (2) run Linux - we didn't have any spare OS/2 licences - as do most of the workstations. There are a couple of people that bring in their windows laptops but from what I've heard we might be changing them to Linux soon.

    The above paragraph is just to highlight Trevor's point about enterprise solutions being sold to SMEs with the emphasis on the S.

    1. keithpeter Silver badge

      "The servers (2) run Linux - we didn't have any spare OS/2 licences - as do most of the workstations. There are a couple of people that bring in their windows laptops but from what I've heard we might be changing them to Linux soon."

      Stick that Web page up. The more small companies/charities/communities document their frugal solutions, the quicker we can get out of the 'IT means expensive and complex' mind set.

      PS: Have a look at Stella Linux for a conservative client system with years of support (based on CentOS) and codecs/multimedia.

  20. RonWheeler

    Money is very flexible in SMB

    Have worked in and out of the SMB field for a long time. The thing that often ticked me off royally was a complete lack of budget for (say) backup software renewal, or a copy of Sage for finance that isn't 25 years out of date. Yet the money was always magically there when the boss was speccing up his own laptop. Same crap, smaller scale.

    1. Anonymous Coward
      Anonymous Coward

      Re: Money is very flexible in SMB

      Has to be said, I see that sort of behaviour all the time. On more than one occasion, letting the IT kit degrade while buying 'more important' items (signwritten cars, new chairs for the managers etc) has ended up costing the company far, far more money than the kit was worth through lost business.

      After all, what customer is going to trust a company with thier needs, if they can't look after their own core infrastructure?

    2. keithpeter Silver badge

      Re: Money is very flexible in SMB

      Sage accounts: students in all the Colleges/Centres I teach in are working on reasonably up to date versions. I don't teach Sage but I do run maths workshops which Sage students drop into. Some have noticed differences between work and College systems.

      1. Steven Raith

        Re: Money is very flexible in SMB

        Semi-related, but I worked with an old boy recently who was using a version of Sage Line 50 so old, his HP drivers didn't have the 16bit code Sage needed to set up his margins. (his old Canon printer he had being using physically crapped out on him, natch)

        Worked around it by setting an old (Adobe Pro 6 I think?) PDF printer as his default, previewing the reports (which use the old PDF printer to do it's page layouts), then changing the printer to the HP. It was a Photosmart C4100, for reference.

        If anyone has weirdy beardy problems with printing from flaky old versions of sage, try an old PDF printer of some ilk as the default, preview in that, then change the printer in the print dialogue - worked a treat...

        Steven R

        (spreading the love)

  21. Anonymous Coward
    Anonymous Coward

    "If you can't get a business to eat good dog food, what chance do you stand if the dog food is off cuts and offal?"

    love that bit - an upvote just not enough :)

  22. Denarius Silver badge
    Thumb Up

    perceptive article and posts

    Trevor, well said. Is the root cause of this behaviour a cultural belief that everything can be planned, managed and controlled ? I see this in several fields, such as aviation which has an excellent historical analysis of incidents. This virtue seems is becoming pathological because it is seemingly practical to load up all aircraft with small nearly affordable electronics to track all movements so a few edge cases of collision might be avoided in remote areas. A simpler approach is to continue to encourage and if need be, enforce better radio habits in all pilots, especially in very rural airfields. There is not much that can be done about terminal stupidity, other than natural deselection. Even so, the number of checklists one can memorise is limited. The increasing process load is taking the fun out of recreational flying. Equivalent to firms going under as participation falls. A parallel may be the belief of governments that mass surveillance prevents bad guys doing things.

    In IT, all of this planning and managing activity seems to forbid any custom fit for purpose work because the process does not allow it. The problem then becomes preventing management recruitment from the bigger companies where the best practice disease is endemic.

  23. P. Lee

    I'm always curious about the desktop / server division. Does anyone have any stats on performance and failure rates for desktops in server duty? I often wonder if it isn't better to plan for failure with with some dynamic DNS and an LTM, and then skimp on the server-class hardware where the usage is low and the main thing is to have the facility, not to have massive performance.

    With a decent disk subsystem, is there a great deal of difference between a hex-core i7 and a quad-core xeon? More to the point, is the benefit enough to warrant the additional cost? Does anyone have stats on how often ECC RAM saves the day? How many SMB apps need that much reliability?

    That leads me on to being skeptical about VDI. More expensive hypervisors, double-licensing for desktops, and more expensive server hardware and disk hardware. Does no-one ever have a use case for piling up those old core2's in the corner and load-balancing desktop sessions across them? Has no-one done a script which sends WOL requests when the desktop farm begins to fill up and shuts machines down when they aren't used?

    I work firmly in the Enterprise space, but I'm constantly confronted cost weirdness. Is a proxy really with Bluecoat's prices; Checkpoint have VPN failover... but why not just use two VPNs as they are stateless? Stop threats at the network edge... with the same AV you have to have on your desktop!

    Few things dismay me more than seeing all those laptops in the office with tiny screens. Get a really cheap laptop a user can keep at home and hook up some desktops with decent screens - cheaper and more far more productive, far longer life-cycles as kit isn't being lugged to and fro.


    1. Trevor_Pott Gold badge

      I have had 10 ASUS P5E-VM-DO motherboards in service as file servers /with 8GB of RAM and Adaptec 3805 RAID quite literally since the veryfirst of these boards hit the distribution channel in Canada. That's what, 5 years now? 6? My home lab is largely composed of vPro ASUS motherboards and a significant chunk of my primary lab as well. (See here:

      In my experience higher end (read vPro or equivalent) profession/workstation-class desktop motherboards fail less than server boards. This is because they are significantly less complex than server boards but manufactured with the same high-quality components.

      Lower quality motherboards (Asrock, ECS, bottom-dollar ASUS/Gigabyte/MSI boards) are the worst of the lot. They are still made with liquid caps, have dirty volt regs and the traces are made out of what seems to be easily evaporated pot metal.

      I trust a quality manufactured Micro-ITX motherboard designed for a home NAS far more than I do an EATX monstrosity powering a 4U HP superserver. The Micro-ITX is probably made from the same top-bin parts, but is so much less complex that the failure rates are way, way lower. Give me 2x quality desktop boards in the same chassis with literally every conceivable component duplicated, wrapped into a cluster. I will take that before your "statistically unlikely to die, 4 hour service life" enterprise-class server.

      My design costs the same (or less) but offers complete redundancy and is statistically less likely to have even one of it's nodes die than the Enterprise version. If I can't shoot one of the nodes and have the overall system still working, then it just isn't good enough.

    2. Anonymous Coward
      Anonymous Coward

      You seem to be thinking that the modern corporate IT department exists to provide a cost-effective service to the rest of the organisation.

      Not sure that idea has been correct in recent years. My experience says it exists to maximise the IT department's budget (buying all the shiny you mention and more) while minimising the positive impact on the business e.g. by operating as inflexibly as they can get away with without terminally upsetting those outside IT, in particular those organisationally above IT who should be supervising the input to, and effects of, IT's activities..

      The idea of a rollout of tiny laptops with tiny keyboards and tiny screens is indeed ridiculous, but when the IT department has decided that is the "new policy", who is going to reconnect IT with the need for cost-effective productivity outside their organisation?

      I certainly agree that there is a *need* to "think different" than many IT departments (and suppliers) have been used to. Whether the technical details you mention briefly are the right answers is another discussion for another day.

  24. bobble

    Are there any cheaper alternative to networking ? I hear about Arista, Dell and others. But would like to hear from experiences whether these are really easier to use and cost effective.

  25. IGnatius T Foobar

    Ditch the big vendors

    The problem with the big vendors (Microsoft is particularly bad with this) is that they are only interested in large scale installations now. If you are a small organization, they want you to buy their service running in the cloud, thank you very much.

    Well, not everyone trusts their business lifeblood to someone else's hyperscale cloud, and with good reason. The solution, of course, is to ditch the big vendors and embrace open source.

  26. Anonymous Coward
    Anonymous Coward

    Foggy future

    < IT Manager / Sysadmin | UK SME <250 Users)

    Having recently worked with the latest batch of M$ offerings (Server 2012 / SharePoint 2013 / Win8) it seems to me completely obvious that the reason for much of the "f*ck the mid market" attitude is a deliberate strategy to push customers to "cloud" / managed solutions.

    (See decision to remove design view from SP "designer" so a mid-market SP admin needs to be or buy a proper developer just to do a data view)

    I can see this being my last "(almost) fully Microsoft" refresh.

    If the folks at Redmond want my budget they are going to have to work harder for it. My reaction to having my life made painful on their "on premises" solutions is not going to be to run willingly into 365. It's to give them the full FU and find a genuine alternative.

    Really disappointing that they have taken this position. I've always been rather the fan of M$, they put enough in their products that as a decent IT pro you could know enough to support a company (I'm the MSSQL admin, the SharePoint admin, the Exchange admin, the File/Print/Server admin etc...) without blowing the bank on consultant services or managed solutions.

    The other argument of course is that they are placating their partner network by trying to send them "consulting" business now that MS are Hoovering up much of the licence revenue directly through 365. Conspiracy ahoy!

  27. This post has been deleted by its author

  28. ecarlseen

    Chefs vs. Cooks

    Great points in the article and comments. Here's what I've found in this area:

    In the IT field there are two kinds of people, analogous to chefs and cooks. "Chefs" can create new recipes based on what the needs are, what's available in the kitchen, etc. - they know what flavors work well and never need to consult a cookbook (in fact, they could write one given the time and inclination). "Cooks" can only really prepare recipes created by chefs, with widely varying degrees of aptitude and limited ability to deviate from the recipe with any success.

    In the IT field, there are very, very few chefs. In fact, there are not nearly enough "chefs" to go around - so you wind up with a lot of systems "engineered" and implemented by cooks. Since they lack the skills and / or confidence to whip up a bespoke recipe, they (if everyone is lucky) "design to white paper." Think about it - it takes some guts to be able to look somebody in the eye and say "you can bet your business on my design." For those of us who are chefs, it doesn't really feel that way in the moment (because we've been around the blocks so many times we know every last millimeter of it) but it you take a step back and look at the situation it's kind of a Keanu Reeves in the Matrix "Whoa moment."

    And as bad as this "cooking to the white papers" situation is, the alternative is probably far worse - people with crap skills screwing organizations over with designs that self-destruct on impact with the real world. Most of us have been called in to fix these situations and it's just painful to behold. In many cases, if you can't get a properly-optimized solution, the next best thing is probably "design to white-paper" - even if it's inefficient it has a better chance of success than something pulled from clown's posterior. That being said, they'd usually be far better off paying quite a bit more for proper design consulting. We've often been able to underbid proposals make by "cooks" - we're offering a far better solution and walking away with more money for less overall cost to the customer.

    1. Trevor_Pott Gold badge

      Re: Chefs vs. Cooks

      The solution is that you bring in a chef to design the infrastructure - usually called a "datacenter architect" or something similar - and use cooks to keep it running. VARs and MSPs do this quite well in the SMB space.

This topic is closed for new posts.