HP being one of the top few leaders in the IT industry, ranked 9 out of fortune 500 in the year of 2009. However, like any other companies, it went through declivous period.
In 2005, there were 14,500 job cuts in order to align their company cost with sales. In extent of this, HP numerous systems and applications had made things worse due to data management problems. There's difficulty in obtaining information when there's more than 750 data marts and this caused data inconsistency which eventually dragged down sales and profits. Randy Mott, the current vice president and chief information officer of HP, was hired during this down period. His aim was to streamline the company information technology. To achieve this, he introduced Enterprise Data Warehouse by consolidating the existing data marts into a single data warehouse system which serve the whole HP enterprise.
To ensure the system works, precise designing and modelling the single enterprise wide database is a must. Data had to be up-to-date, consistent and complete for the entire enterprise. The enterprise data warehouse was rolled out in 2006. To date, HP had managed to cut down hundreds of data marts to a mere 200 containing hundreds terabytes of functional data. Now, all HP financial data can be accessed via their data warehouse system. Not only does this benefits HP, it also open up a business opportunity for the company.
Neoview was introduced in the year 2006 from the proprietary work that the company had done in its own data warehouse. Though it's a proprietary work piece, it's robust, simple and flexible functionalities differentiate itself from typical data warehouse, which are costly and tend to focus on one area of a business rather than an entire enterprise.
Both HP customers, Bon-Ton and Wal-Mart, which were the users of Neoview, complimented the product as it boosted their overall revenue till date.
Question 1. Identify the problem described in this case. What people, organisation, and technology factors were responsible for creating this problem?
Hewlett-Packard Company, also known as HP, is founded in Palo Alto, California, USA by 2 guys namely, Bill Hewlett and David Packard. No one will ever imagine that it started famous by making electronic test equipments such as oscilloscope, signal generators, etc. during the early years. With this, they went listed on November 6, 1957. However, through transformations for more than 40 years, it's now one of the few leaders in the Information Technology companies. It is located in almost every country and provides all ranges of IT products and services. With products like printers, personal computers, enterprise servers imaging and storage devices as its main product lines, HP is able to market them from small to large scale businesses and even enterprises. Moreover, HP is able to target into households with their innovative products design. In the year 2007, HP topped the chart to be the first IT Company in history with a posting revenue exceeding $100 billion. With about $119 billion of net revenue in the year of 2009, it's expected to increase further more in the next 5 years. With this huge posting revenue, HP had beaten its rivals, IBM and DELL, down the road. Not only does it excel in products sales, HP also became number 2 in IT services globally as reported by IDC and Gartner. In the year 2000, HP announced its merger with Compaq and this pushes HP even further up to the summit. In barely 1 year from 2008-2009, it had acquired Electronic Data Systems (EDS) and 3Com. HP was ranked 9 in the fortune 500 in the year 2009. With the incursion into enterprise networking gear market as its latest move, it's expected to be a strong competitor to Cisco.
However, in the year 2005, it went through declivous period. During this period, there were 14,500 job cuts in order to align their company cost with sales. In extent of this, HP numerous systems and applications had made things worse due to data management problems. At one time, HP had 5000 information system applications, 85 computer centres, approximately 19,000-22,000 servers, 17 different database technologies and 14,000 different production databases. Though it had enormous storage capacities, there were difficulties in getting consistent and timely data as all information were located at different part of business. Without consistent and timely data, more time would be wasted on compiling them, especially when information was obtained from various systems, and this eventually dragged down the sales and profits. On top of this, some of their systems tracked price and sales by products while others tracked this information based on locations which worsen the inconsistency condition. Not only this, commonly used financial information were calculated and evaluated differently across the board which indirectly affect the profitability of the company. With all these problems in place, a simple question such as amount of money spent in marketing across the business was difficult to answer. Without this consistent view of business, chief-level executives were having serious problems in making decisions such as assigning a bunch of specially trained IT personnel to operate certain sophisticated systems. Without a clear view of business, any wrong decisions could inadvertently inflict loss.
There were several factors responsible for creating this problem. We'll discuss them in terms of People and Politics, Organisation and Technology.
People and Politics
Like any companies around the world, people and political issues would always be present. Although an integrated system sound appealing to the top managements, the users across the board might not think so. A centralised database would never be appealing across different department in the same company as they cross organisational treat lines. It would also mean data ownership and access would be changed which will affect the work practices of highly independent and powerful user communities in the departments. This was a serious fear factor as divisions, departments and business units which had their own devious ways of keeping things precise and simple were understandably doubtful about exposing these out as people may have a vested interest or historical reason to point out the data quality issues. The other reason why most employees resisted was because of difference in work flow process. A new system could mean that all process would be revamped and not a lot of people can adapt to it. That's why many of them preferred smaller data marts designed for their own usage which cut down complexity.
As a huge enterprise company, HP had far too many computer centres which were running with different information system application. Not only these, it had a large variety of database technologies running on numerous databases. As discussed earlier, this was a serious problem as data would be located everywhere and inaccurate information would surface. Due to old culture system, like any other organisations, departments were allowed creating, using and managed their own databases without sharing the data with other departments. This posed a problem as functional silos would be jeopardising the company. On top of this, the legacy systems existing were unable to support the real time accessing of data due to the fact that the hardware were unable to support.
An integrated data warehouse require massive amount of work to get it establish. Just integrating and organising the data require unique knowledge of database technology and design principles. Talents, from the pool of potential employees in the enterprise, were difficult to find too, be it technical or non-technical. Apart from this, HP lacked of the technologies, regardless of hardware or software, to build such a sophisticated database that's easily and quickly available to more than 50,000users.
Established in 1939, Hewlett-Packard started small. However, as years pass by, its Research & Design were producing more electronic products and the recent years HP even provide complete IT-support solutions for companies located around the world. All these require numerous gigantic data centers. As their businesses increased, their data marts increased linearly. Not only does it expand locally in USA, HP was spreading its businesses internationally. In the early years, these might not pose a problem to them. Instead, to them, the more data marts they have, the better. This was because they could house a lot more data than usual. However, down the road, these pose a serious problem. All sub-branches around the world or even individual departments were maintaining their own databases. No one had a clear overview of the business as a whole.
Question 2. What solution has HP chosen to fix this problem? Did management select the best solution alternative?
Randy Mott, the current vice president and chief information officer of HP, was hired during the year 2005, the period when economy was not doing well. His aim was to streamline the company information technology. To achieve this, he introduced Enterprise Data Warehouse by consolidating the existing data marts into a single data warehouse system which serve the whole HP enterprise. In order to streamline numerous data marts from all over the world, Randy Mott gathered 300 technically trained personnel to form a team which will be running the whole data warehouse project. They'll be involved in the whole software development architecture design. Like any other projects, data warehouse will require strong and sturdy foundation. To have a well-established foundation, the data had to be constantly up-to-date, consistent for entire enterprise and complete. These three goals may seem difficult to achieve but to HP, they were both important and the key factors to success. This was because; the initial proposal for deploying data warehouse was to integrate their data marts to lesser quantities so that information can be managed with control. However, if information were not accurate, they would have fallen back to square one and this data warehouse project will be deeming as failure.
As discussed earlier, there were three points that's hindering HP from gaining success. The key to success would be to eliminate these three problems.
To solve both people and politics problem were never easy. Management had to put in a lot of effort in educating and changing the mindset of their employees so that they can accept the new system and new way of working. However, this requires frequent interaction between the managements and employees. It was also during this period when top managements could find out from their middle managements and operational employees whether existing processes, which were lengthy and time consuming, could be shorten. This could greatly benefits the company while integrating with the data warehouse system as it could further reduce the man and time efforts to get things done faster. Nevertheless, it was a good chance to clean up or wipe out all underhand methods used by employees or even middle managements during the period before implementing data warehouse system.
In the previous HP organisation structure, functional silos which consist of both vertical and horizontal existed. They were responsible of isolating every operating unit's environments from one another. This was undesirable as it affect co-efficiency. However, with the emphasis in deploying data warehouse system, it helps in breaking up the silos. The organisation had changed to cross-functional structure which smashes the functional silos by opening and allowing informational flows from departments to departments which greatly improve efficiency and teamwork. Inadvertently, it helps to boost team spirits between departments which was an important factor towards company growth. There were implementations of middleware tools in order to integrate their legacy systems and new systems. In the long run, HP would be phasing out legacy systems as they'll upgrade their hardware and software, as part of standardisation policy, in order to handle the sophisticated applications and most importantly storage of gigantic data and prevention of vulnerabilities attacks by hackers.
Last but not least in order to tackle the technology problems, HP hired specialists that were technically trained in Enterprise Resource Planning system to provide in-house training for HP selected personnel. This could significantly boost the knowledge of those personnel involved in this training and they'll be much more competent in implementing the data warehouse system. In addition, a Research and Design laboratory was specially setup for this project so that this pool of implementers was able to put their knowledge into valuable uses.
After a year of development, HP put its first data warehouse system into use within its own enterprise. With all the initial effort in place, the new system proves to be effective. It had reduce HP overall data marts to just over 200 containing 180 terabytes of raw data and 75 terabytes of functional data. The data marts were expected to reduce further with double or even triple increment of both raw and functional data.
Everyone had their own perception on whether this solution was successful. However, from the business point of view, the solution indeed had managed to cut down on HP average annual expenses and increase their annual revenue. Not only does it achieve their sales figures, they were also able to manage and control their overall information flow with a decreasing number of data marts. This greatly reduces man efforts and increase efficiency. With all the facts in hand, HP did the right choice to deploy data warehouse system especially when it's during the period they were not performing well and data management was in a mess. If HP was able to roll out the new systems to external companies confidently, it would only prove that it was a total success internally. Not only does it help HP enterprise, it was a successful business product.