Sunday, February 24, 2013

Zara Case Preparation


My fellow executives,

                Our business has consistently been a leader in our industry for net margins having last year surpassed 11% as we grew our net income to 438€ million from 341€ a year ago.   Our company sustains rapid growth as we’ve expanded our footprint across 45 countries and unveil a new store location every day.  This success is a byproduct of our business model, which as you’re aware hinges on our ability to respond immediately to changes in consumer demand and is bolstered by the independence and empowerment of decision making we entrust in our employees.   This business model works because it allows us to respond the quickest to consumers while minimizing our investments in marketing, inventories and IT.   Our IT systems have become a hot topic of discussion as they are now seriously being called into question.   Can we continue to rely on our incredibly dated yet reliable DOS operating system?  Or does the risk our hardware supplier may someday quit producing DOS compatible devices justify a full scale modernization of our operating systems and POS terminals?  I believe that the answer to both of these questions, especially the later, is no.  Rather we should find a position that lies somewhere comfortably in between these alternatives.  It would be prudent for us to continue utilizing our affordable and operationally sound DOS system that continues to meet our essential needs.  We need to simultaneously educate and prepare ourselves for the day we could be forced to transition off DOS.

From a technological perspective our current DOS operating system is prehistoric.  Microsoft, the makers of DOS replaced it 18 years ago—the technology time equivalent of a century—with Windows OS.  DOS does not support most modern functionalities or networking capabilities.  But DOS is however a simple, reliable and inexpensive system.   And for those very reasons DOS is an excellent fit for our business.  The current DOS and POS systems require no IT involvement.  We do not need to send a specialist to new store locations for set up.  We rarely ever receive calls from new or current stores managers related to IT assistance.  The systems are simple enough that our store managers can easily install and reboot them on their own.  DOS and our current POS terminals are fully capable of running our internally built customized software applications supporting our key ordering, fulfillment and manufacturing processes.  The operational stability of DOS has allowed us to keep a minimally staffed IT support function consisting of only 50 people.  By avoiding sophisticated technology upgrades and reducing IT support costs we’ve managed to spend only 0.5% of our total revenues on IT.  On average our competitors invest 2.5%.  This translates to an 80€ million per year cost advantage over our competition generated by our current IT structure.  So why would we consider changing anything with the current system when it’s yielding significant and steady contributions to our bottom line?  The reason is risk.  There is reasonable risk that POS hardware vendors could upgrade their machines to no longer support DOS operating systems.  This would effectively halt our ability to open new stores due to the fact the available POS hardware would not be able to port our customized DOS operated software.  If blind sighted by such a scenario, our minimally staffed IT group would likely not have the capacity in terms of availability nor training to handle an immediate operating system transition.  If we hired a team of 15 specialized professionals at a rate 450€ per person per day it would cost us 844€ thousand and take 6 months to complete the 15,000 hours of required programming time to port the existing POS application to a new OS.   All the meanwhile we would be unable to open new stores. Estimating at one new store opening per day and an average store contribution of $2.5€ million in annual sales our risk exposure could be 450€ to 912€ in lost revenues.  It would be reckless for us to simply ignore such a risk.

The alternative that’s gained the most attention among members in our company involves upgrading to a modern operating system, enhancing the POS application itself and possibly building networks between stores.  Before I touch on the potential benefits, let me first outline the costs associated with this option.  The total estimated cost for completing upgrades to each of our current 1558 stores would be around 23.4€ million.  Going forward the annual recurring costs would be 0.5€ million for all stores.  The breakdown of this would include (in millions) 1.3€ for operating system licenses and maintenance fees, 8.5€ for hardware and connectivity, 1.1€ for programming and 12.5€ for installation and training.  For a company doing nearly 4€ billion in annual sales 23.4€ million may not sound like much, but it’s a serious investment that outweighs the 20€ million we currently spend on IT in a given year.    And each new store would cost approximately 13.5€ thousand to integrate.  These costs are not even the most considerable associated with an upgrade.  A sophisticated system would be more complex and would require far greater internal IT support to maintain.  Store managers would not be proficient in performing the installation and maintenance they currently do with the simple DOS systems.  We would have to further develop and greatly expand our IT function to support the new platform.  While we wouldn’t move into online ordering, it wouldn’t be shocking to estimate an increase of IT costs from 0.5% to 1.5% of our net sales.  This amounts to an additional 40€ million per year.  There are other potential costs that are significant yet difficult to quantify.  These could include time and process inefficiencies related to assimilating store managers and employees to a new sophisticated system.  There is a greater likelihood that a new operating system, while current and powerful would be more susceptible to bugs and crashes than DOS.  As for the benefits to upgrading, the only justifiable reason includes avoiding exposure to a loss of revenues in the event hardware producers discontinue DOS compatibility.   Further benefits are little else but “nice to haves.”  Store managers suggest they would be able to more accurately and efficiently manage their inventories.  I don’t see how this applies to our business model that focuses on maintaining near-zero inventories.  One of our competitive advantages is that our 1,146€ million invested in total current assets—a metric that captures inventory— is considerably lower than that of each of our core competitors.  Why would we place an emphasis on managing inventory when our emphasis is to erase inventories altogether?  I just don’t see the business justification for wireless network supported inventory management tools.

                I believe we should continue to use our DOS operating system for as long as we are able to.  DOS is consistently stable, supports our core competencies and allows us maintain the lowest IT costs in our industry.  We are poised hold an 80€ million annual cost advantage over our competition for every year that we continue riding out the DOS system. Our hardware supplier has not made any indication that they intend to discontinue supporting DOS.  On the other hand, they also refuse to commit by contract that they will continue to support DOS.  Our current exposure to a sudden discontinuation could be upwards to 900€ million in lost revenues resulting from an abrupt halt in store openings.  We should not make a full upgrade at this point.  We should instead take small but meaningful steps to prepare ourselves for a potential scenario where we are forced to move away from DOS.  This can be done by hiring a small team of 5 to 10 IT employees who will focus on testing out new operating systems at a strategic sample of 5 current and 10 new store locations.  I estimate these measures would increase our annual IT spend by at most 5€ million.  This team would be tasked over the next 18 months with completing the necessary 15,000 hours of programing required to port POS applications with the new operating system.  By completing the programing time and creating a specialized IT team capable of leading an upgrade we can alleviate our 900€ million exposure.  We’d also educate ourselves with the sample store test upgrades to be better equipped to handle obstacles in the event we needed to upgrade all our stores.  In this strategic position we’ve now prepared ourselves for a transition, reduced our risk exposure and we can maintain an annual 75€ million in IT cost savings advantage.  This way we get to have our cake and eat it too.      

Junk Van Reflective


Fellow executives,

Having the opportunity to reflect after the consultant presentation, I feel the best path in front of us is quite clear.  The presentation has bolstered my confidence in PaaS is the best fit for our company given the available options.  During our preparation I evaluated the six potential candidate options using an evaluation matrix and scoring system.  While PaaS scored the highest among these options, a few others including Microsoft Access and Custom Application were not far behind.  Their power and reliability as well as remote access may have made them hard for some of us to rule out.  But as the presentation carried on something became quite clear to me.  We do not have time to fiddle on our own with an implementation of a new database.  We have a rapidly growing business that we must devote all of our time and resources to.  We need to know we can count on a team of database professionals to handle upgrade.  We don’t have the time or knowledge to take that task on ourselves.  We want an affordable and dependable option that fits for us without having to get our hands dirty in anything but our business of junk.  While Microsoft Access is a fine product it would require us to learn how to put together relational databases and handle much of the setup ourselves.  Custom Application would be designed for us by a third party company.  However, due to the fact that we wouldn’t be able to test the product until completion we would need to allocate considerable time and energy in first understanding our detailed requirements then specifically mapping out and communicating those blueprints for the designer.  This takes too much involvement on our end.

PaaS on the other hand will allow us to get up and running without the need for us to get heavily involved.  In as little as three days we will have a system in place that we can begin using.  It’s flexible to our needs and user friendly.  If we need to modify the applications we can rely on a separate support team to handle this.  While I understand $180 per hour to do so isn’t cheap, this option gives us plenty of comfort with the knowledge that we can cancel service with only a month’s notice.  We pay $300 to $600 per month so we are not on the hook with long term payments.  IIn the event we are not pleased with PaaS or feel we are not getting what we’re paying for we can cancel, reevaluate our options and move on without sacrificing significant time or cash resources.  I don’t see another option that fulfills our requirements, minimizes risk and allows us to keep our focus on our business.

Bobby

Sunday, February 17, 2013

1-888-Junk-Van Case Proposal


Fellow executives,

                The growth and viability of our small business has become plagued by errors related to data processing coupled with operational inefficiencies.  These symptoms are byproducts of the severely constrained data infrastructure our operations utilize.   A key weakness of the current system is that it requires all customer and even payroll related data to be received, interpreted, updated and transferred in an entirely manual fashion.  This convention readily invites informational errors affecting our job orders, scheduling, staffing, invoicing and billing and even employee compensation.  This problem is further compounded by the limitation that information cannot be stored in or accessed from a central database.   From origin to final destination, data slowly flows sequentially along a cumbersome chain of employees.  In most cases the information is manually handled by call multiple center operators, a data clerk and drivers more than once by each in the process.  This virtual game of telephone is causing such an array of errors it is driving the quality of our customer service and our reputation straight into the ground.  This is completely unacceptable for the standards to which we claim to hold our company!  We market and model our business on the premise of providing quality, timely and professional waste removal services.  Our inadequate information systems are spawning operational inefficiencies.   We are not processing data in a way that would allow us to optimize our delivery loads to recycling centers.  We are paying centers to accept small drop-offs of recyclable waste that if disposed of in bulk we would be paid for.  The excessive time and manual effort drivers must spend sorting and transmitting receipts and invoices.  To complete a cash order our drivers must double-back to the locations they already finished at to collect payments.  In this competitive commodity business we cannot afford to waste money on gas and driver hours for the purposes of billing transactions.  We must improve our information systems. Otherwise, we not only won’t grow—we won’t last.
                I’ve scrutinized six available options for improving our informational infrastructure for the purposes recommending a selected course of action.  The options consisted of Microsoft Access Database (either installed locally or on a centralized server), Custom Application, Google Docs, Platform as a Service, and ERP.  I rigorously evaluated each option according to the following eight key criteria; power and reliability of the central database, automation, remote access, ease of use, flexibility, customer support, time to implement, and price.  To organize this array of feedback and help streamline my results I constructed an option evaluation matrix.  The matrix presents the scores each option was assigned for each criterion (1 being the lowest to 5 the highest).  I included this visual aid for you.  As you can see, none of the options attained high marks across all categories.   Before I discuss these options and provide my recommendation I want you to be keenly aware that each alternative carries a degree of risk, uncertainty and cost that may offset and even potentially outweigh their expected benefits.


                  The first two options I evaluated were Microsoft Access Database installed locally on several computers versus installed on a centralized server to that could be accessed remotely via the internet and secure virtual private network.  Both of these options would provide a reliable and powerful centralized database.  The local installation could be installed in as little as few as two weeks and is relatively easy to use and inexpensive —at $179 per PC it would likely keep the total investment under $1000.  The drawbacks are that it’s neither remotely accessible nor automated, meaning the database would still need to be manually updated and transmitted every day.  This is a deal-breaker.  MS Access installed on a central server would provide automation and remote access.  This option would likely require the added expense of hired technical support and could take double the time to implement.  A more complex system, this would require extensive employee training.  It would also carry additional costs for hosting a shared server and user licenses.
                The custom application option involves having a third party build a completely customized system for our business.  Upsides include remote accessibility and a reasonable cost of $2,000 to implement.  This would hopefully meet most of our power, reliability, automation, ease of use and flexibility requirements.  I emphatically stress the word “hopefully” because this option requires upfront payment before we have the ability to see the finished product.  While we could do our best to communicate our requirements there is no guarantee those needs would be met.  I risk adjusted most of the criteria rankings for this option.  Another drawback is it would take approximately 4 weeks to implement.  Also, the $2000 price tag does not include maintenance costs at $60 per hour and there is no way to approximate the amount of maintenance.  Customer support is available—also at an hourly rate.  The high degree of risk associated with this option regarding both quality and price is very alarming.       
Google Docs is the most cost efficient option.  It is free of charge for the first 10 user accounts and carries only a $50 per year charge for additional accounts.  Google Docs would only take 2 weeks to implement and is very user friendly.  It supports basic automation and email distributions with links to forms.  Google Docs is severely limited in the power and reliability department.  It cannot utilize cross referencing between tables which is fundamental to relational databases.  All data is inefficiently displayed on single large form.  Google Docs provides no tangible hardware or software because everything is processed and stored by means of Cloud Computing.  This invokes risks related to confidentiality concerns and could leave us extremely vulnerable in the event Google discontinued its service.
Enterprise Resource Planning (ERP) systems such as SAP Business One and Sage ERP are best in class for the power and reliability, automation and remote access criteria.  They are also highly complex and inflexible platforms requiring considerable implementation time and extensive employee development.  They are also astronomically priced.  The Business One “starting pack” consisting of 4 licenses prices at $12,000.  And there are likely steep maintenance and customer service charges that follow.  Not to mention additional licenses for all users.   What bothers me about this option is not simply the substantial price tag, but that we might be investing in some services not applicable to our business like production and finance platforms.
The last option is Platform as a Service (PaaS), which utilizes cloud computing through the internet.  It allows users to access common applications and build their own custom applications using a shared computing platform provided and hosted by a third party.  PaaS is powerful, flexible and user friendly.  It allows users to build unique applications and link tables with relative ease.  I slightly downgraded the reliability rating for the fact that it requires cloud computing and carries the risks previously mentioned.  PaaS is best in class for time to implement, requiring only 3 days inclusive of data migration.  This quality product is not cheap as it can cost between $300 to $600 per month depending on storage and licensing requirements.  This investment is well balanced by low risk.  There’s no requirement for long-term contracts and the service can be canceled with 30 days’ notice.  The system features and capabilities are incredible flexible in that they can easily be scaled up or back depending on our requirements.  Customization costs $180 per hour, but may be worthwhile.
While there does not exists a clearly obvious choice, PaaS garnered the highest total score on the option evaluation matrix.  Receiving a 30 (out of a possible 40) PaaS was a full 5 points ahead of any of the next highest ranked options.  PaaS scored consistently high on all key criteria with the exception of price.   Again, this substantial investment is largely mitigated by the lower associated risk and commitment.  Considering the available options I am confident PaaS is the best course of action.  It provides the most balanced solution in terms of meeting our spectrum needs.  It is also devoid of deal-breaker characteristics.  As we make this decision I urge you all to examine the evaluation matrix.  The numbers do not lie.

Bobby