Sign in / up

back to article John Lewis unzips mega Oracle ERP package

Flagship British retailer John Lewis is starting a massive ERP rollout on Oracle as part of an ongoing business transformation project. John Lewis has picked Oracle over SAP and unnamed “other” candidates in a four-year project to modernise its inventory and customer systems, The Reg has learned. Oracle will replace 50 legacy …

COMMENTS

House rules Send corrections
This topic is closed for new posts.
  1. JetSetJim

    Parallels

    If it's anything like the "upgrades" of the Waitrose Online over the years, expect a few weeks of fail before they adjust for new load.

    On the other hand, stick some orders in during that time as, when they inevitably fail, they'll probably give you some decent presents to compensate just like Waitrose did

    0 1
  2. Frankee Llonnygog

    Oracle ERP?

    If John Lewis had shareholders, now would be a good time to sell

    4 3
  3. JeffTravis

    Oracle in Retail

    Following the model pioneered by Morrisons?

    1 0
  4. Anonymous Coward
    Alert

    Between a rock and a hard place

    I'd imagine nobody would envy them trying to do a horrendously complex "big bang" implementation of one of the monolithic ERP systems like Oracle's, with the very real risk that it can screw up your entire business. OTOH trying to integrate and maintain of hotch-potch of 50 separate systems almost guarantees that some part of it will fail at some point. How do you prefer your pain, acute and agonising or chronic and bone-aching?

    Still, betting against Gartner and the tea-leaf readers seems like a good starting point.

    5 2
    1. Roland6 Silver badge
      Reply Icon

      Re: Between a rock and a hard place

      The article gives no indication that the system will be a "big bang" implementation. Given the magnitude of the task, I expect a phased deployment - certainly the reported timeframe indicates a strong awareness of the scale and the risks involved.

      A real challenge will be managing the shifting sands of the IT environment, as we can expect new platform software and hardware to be released in the coming years and some will always be wanting to run with the latest and greatest products.

      0 0

  5. This post has been deleted by its author

  6. Anonymous Coward
    Anonymous Coward

    Money pit

    Oracle can be a complete money pit (although SAP is the same). It seems that often companies do not know just how much work needs to be done on top of the standard Oracle products in order to get it to do what they want.

    Over the past 5 years or so we have spent an unbelievable amount of money to migrate to Oracle from various other systems. There are probably another 5 years to go. And the logic would suggest that consolidating down to one product will be more efficient, in practice we have not seen any efficiency gains overall and it is likely to be some time until we do.

    Sure, in a decade's time it might all be running smoothly and efficiently.. and we may well wonder what the fuss was about. But you really do have to think about these as very long term changes with no quick or medium-term gains.

    5 2
    1. Anonymous Coward
      Reply Icon
      Anonymous Coward

      Re: Money pit

      In a decade you'll have amortized all the investment and unless you've invested a fairly big amount of money, equivalent if not larger than the original implementation budget, you'll find yourself with a hopelessly outdated system that has to be replaced by... you guess it, another big bang style implementation.

      Seems that not only nobody learns from their mistakes, it is even worse: nobody learns even from mistakes of others.

      5 0
      1. Tony S
        Reply Icon

        Re: Money pit

        "In a decade you'll have amortized all the investment "

        Probably not; most of these bigger ERP implementations will only be cost effective if it's calculated over a longer period of time such as 20 years. Based upon my own experience, if they are saying 4 years, then it probably means that they will still have consultants on site doing various bits of work in 6 years time.

        " unless you've invested a fairly big amount of money,"

        I think that is an absolute guarantee. Certainly, this project will be costing many 10s of millions of pounds (possibly into the 100s of millions). In most cases, they go way over the original budget as the sales people work on the basis that once you are locked in, you won't want to back out and they deliberately under quote to hook the customer. I'd not be surprised to hear that it was 3 or 4 times over budget by the end.

        "Seems that not only nobody learns from their mistakes, it is even worse: nobody learns even from mistakes of others."

        That is very true; there are more papers, studies, investigations into failed projects than you would want to read unless you were an academic researcher. Ultimately, they all boil down to the same failings, each and every time. But hey, what do any of us know about anything; we're only the silly buggers that have to try and make these things work.

        4 0
      2. Anonymous Coward
        Reply Icon
        Anonymous Coward

        Re: Money pit

        If you're implementing a large ERP system simply because it's apparently cheaper than running the legacy, then that's almost certainly going to end in tears. It's almost always cheaper in the medium term to just keep things as they are.

        Big ERP systems can pay for themselves many times over IF they offer significant business advantages over the hotchpotch of legacy. That could be enabling a company to have better financial management, reduced error and fraud, enabling customer self-service, reducing time-to-market, real-time stock management, earlier closing of financial books etc. etc.

        Also agree that big bang projects are highly risky and almost impossible to estimate. It's incremental change, and gradual evolution every time for me.

        1 0
      3. qwertyuiop
        Reply Icon

        Re: Money pit

        Sadly all too true. On the plus side, however, having had the pain of moving from multiple systems to a single one means that the next move will be a little easier as it's just from one system to another. Hopefully..

        0 0
  7. koma

    Umbrella management

    Nobody ever got fired for buying Oracle !

    1 2
  8. Anonymous Coward
    Anonymous Coward

    So...did John Lewis

    Get a two year guarantee and the promise that if they could buy the same ERP product cheaper elsewhere, Oracle would refund the difference?

    Sad to see another British company risk being hung out to dry from one of the masts of Ellison's latest super yacht.

    3 0
  9. Anonymous Coward
    Anonymous Coward

    Having worked at Waitrose I can tell you they have never been particularly good with technology. They got some new checkouts put in just before I left that were actually slower than the ones they replaced as they put in a pretty touchscreen that didn't just let you type in the 4 digit product codes. so instead of typing 1083 for broccoli (yes 8 years later i still remember this) you had to browse through the menu to click on it instead with massive lag between each screen - little wonder it was so bd though as they were crippled running a Java application on 512mb RAM.

    1 0
  10. JLV

    2018?

    IMHO if you get an ERP system, a big success factor is implementing straight out of the box, with limited customizations. Be cautious and frugal - configure, don't develop.

    (After doing due diligence that it will support your business volumes - large retailers and volume businesses have struggled with general purpose ERP systems that do not take kindly to a firehose of transactions)

    An excess of time, $ and development consultants is not always a good thing.

    Customizing ERPs is almost always a problem - possibly straightaway, because your corporate IT and consultants aren't really up to re-designing best practices on top of a complex software solution. Or later, because you can't track patches and updates coming in from your vendor. Especially problematic if the software is buggy to start with - many patches to be expected. But, hey, it looks attractive when you have access to the development tools, does it not?

    Good requirements management, proper C-level sponsorship, user training and change management are better places to spend effort on and cost way less. Do pony up for data integration feeds.

    If you are planning for a 4 year rollout, it could mean one of two things.

    a) you are doing a cautious bit-by-bit rollout, but each module is put online in a relatively quick fashion. What should be easy wins (but the devil is in the implementation), spread out over time.

    b) you're budgeting a lot of time and $ to customize, rather than configure, your ERP to your requirements. Big bucks, big bang. Much risk.

    2 0
    1. Roland6 Silver badge
      Reply Icon

      Re: 2018?

      >If you are planning for a 4 year rollout, it could mean one of two things.

      No it not just an IT systems rollout, it is part of a large enterprise business transformation project (I suspect being run by Mokum) involving both business change and IT systems enabled change. From my experience, the number of legacy systems, stores and business units, combined with the change in the business that the new system will enable, largely explains the timescale; which may seem to be generous but in fact will be demanding.

      1 0
    2. Trixr
      Reply Icon

      Re: 2018?

      Have yet to see ONE instance where an ERP system can do a significant amount of what is required "out of the box". They may well exist, but not in any organisation I've worked for.

      Because yes, in that instance, it would make sense. Implement your ERP to manage core processes, then incrementally bring in all those zillions of peripheral systems are used for one measly obscure function.

      Personally, I still think a well-defined suite of nicely interlocking parts is usually easier to implement, even in the long run. You evaluate your processes and then implement a "best of breed" solution for each part, one by one, with an eye on which parts need to talk to each other. No-one's ever managed to explain to me why CRM and HR systems have to be tightly-integrated, for example (not that they are in SAP anyway). Then put in a nice simple data interchange process for the bits that do need to talk to each other.

      0 0
  11. John Smith 19 Gold badge
    Unhappy

    Likely won or lost on the spadework.

    And there will be plenty of that, starting with getting the existing system to dump out their data models and working out how that's going to map to Oracle's view.

    Then looking at how Oracle does it's thing and deciding a)re-train staff or b)Re-write Oracle to look more like it's predecessor, using configuration options Bad idea.

    Then finding out the clever stuff you in house systems can do that (presumably the things that give you a business advantage in the first place, which is why you wrote them),and implementing that in Oracle, or deciding the world has moved on and they are unnecessary.

    Then doing the data cleaning before you extract/translate/load the legacy data onto Oracle

    And that's before a single line of Oracle (which should be the last option behind configuring Oracle or retraining staff to do it the Oracle way) is written.

    On the upside SAP financials were installed at NASA and they found $500Bn (yes that's more than 1/2 a trillion $) unaccounted for over the 11 centres over about a 35 year period. (See CFO Magazine).

    It'll take a much better person than I am to pull that off. I wish them good luck.

    They will need it.

    0 0
    1. Roland6 Silver badge
      Reply Icon

      Re: Likely won or lost on the spadework.

      >Then looking at how Oracle does it's thing and deciding

      c) Change the business processes (and re-train staff)

      This is most probably what is going on, as one of the reasons JL give for going with Oracle is the implementation of a single view of a customer.

      0 0
This topic is closed for new posts.

Other stories you might like