[Media reproduced, please indicate that from China Business Intelligence Network By] No matter what kind of business and organization, large, medium, or small, how to save business intelligence how to save budgets is very interested. On the one hand, many large companies desperately compressed $ 10 million of commercial intelligence projects to $ 5 million, which is obviously trying to compress costs, and may be extreme caution for previous project failures. On the other hand, there are many small and medium-sized enterprises to pay $ 1 million in investment, but this does not mean that small and medium-sized businesses cannot implement business intelligence solutions. Whether it is a large enterprise attempt to compress budget, or small and medium-sized enterprises face complex issues but not enough financial resources, when a successful business intelligence solution is proposed, you must find a way to make budgets are elastic. This article will give some strategies that can make your budget to optimize your budget to adapt to business intelligence solutions.
Unhappy voice
For a saving budget, the best business intelligence architecture may not be a product you invited by the vendor. Of course, if you tell them that products can't meet your needs, they will recommend the list of needs of the same company, indicating that these needs are in line with their product characteristics. If you tell them that there is no need so many features, the sales chamber tips the limitations of your business intelligence architecture. If you tell them that these products are too expensive, they will tell you that the return on investment is much more important than the initial investment. Briefly: Only include all products of the vendor will be a complete business intelligence solution.
But before you open a check, you will first ask yourself two important questions:
1. Is the function of the product too powerful? If you can use all the features and features they provide, the well-known brands are naturally very good. If it is not, such a product will become bloated tools, low efficiency, implementation and maintenance is too expensive.
2, are you attracted to the low price of the product? The price is low, but it must meet the needs. When you tell the manager you choose the extraction, conversion, load (ETL) tool saves a big fund for the company, and he expresses you to your congratulations. But just a few months, there is a problem, you need more programming, because you need to write a lot of code to support your ETL (you bought the ETL tool unable to win some work). The first thing that the manager wants to ask is: You didn't have an ETL tool to do this? Purchasing the right technology is the key to successful budget. Some companies often buy unwanted or expensive products, and then feel a successful guarantee. But often, traditional well-known brands of product lines are not most effective. Unfortunately, choosing a cheap solution may play a greater role. So what is the best choice? In the following part, I will exemplify some of the product portfolios of non-traditional combinations to save budgets. They are real solutions, in any case, they should be able to stimulate your investigation of non-traditional combinations to meet your needs and comply with budget.
Choose Microsoft
Microsoft's Microsoft Analysis Services is an excellent online analysis processing (OLAP) server. It is a low price relative to traditional OLAP servers. For example: products relative to Hyperion and MicroStrategy. But Microsoft's products are often ignored, because the project manager thinks it can only run on Microsoft's own SQL Server. In fact, when the source data comes from Oracle, IBM's DB2, or NCR's Teradata, you can achieve and excavate the value of Microsoft Analysis Services. Microsoft's database Combined with Microsoft Analysis Services will give you an excellent OLAP solution. For large company compression budgets, or small and medium-sized companies want to get enterprise OLAP, this non-traditional system is effective and low in price.
Lightweight ETL with SYNCSORT
The main database manufacturers intensely debate how their own products are different, but they remember that they have done some of the same things. For example, they all offer an ETL tool that can be said to be a lightweight ETL tool. Database vendors are very concerned with ETL, such as those dedicated ETL manufacturers ascential Software or Informatica tools, they value more integrity, database manufacturers seize market share through price advantages and product bundles . For example, Oracle bind its Warehouse Builder V. 9.2 (OWB) to the network development kit, if your company purchases Oracle Enterprise Edition, then you may already have an OWB of License, or even you haven't realized that you have It has it. Similarly, Microsoft users can utilize data conversion services (DTS), which are bound to Microsoft's SQL Server. Lightweight ETL tools If you don't meet your ETL needs, do they have value? It is sure that the answer is sure. A large number of ETLs are standard behaviors: simple sorting, integration, aggregation and conversion. Many people have invested in full-featured ETL tools to achieve simple features that the inexpensive ON OR DTS can achieve, and of course, sometimes the lightweight ETL needs to purchase high-performance sorting tools, such as Syncsort. By obtaining patented sorting algorithms, I / O optimization, and dynamic monitoring, Syncsort accelerates the processing of ETL transactions, data mining, and clicks, 90% of data warehouses, data mining, CRM, ERP and decision support. It can be handled with it. The combination of lightweight ETL and SYNCSORT allows you to use the price advantage and do not increase the workload. Lightweight ETL tools provide a formal framework to create, scheduling, and maintaining your ETL processing. Products like Syncsort can handle larger loads, that is, when sorting, integration, conversion, and aggregation cannot be processed in time, it is easy to use Syncsort. Companies purchase Syncsort to improve the performance of ETL products, this kind of thing is not common. The combination of these kinds of products should be considered before making decisions to select the final ETL product.
Combination of TERADATA and Microsoft
Most commercial intelligent project planners have never considered Teradata's combination, from NCR's data warehouse section and Microsoft Microsoft Analysis Services. But in fact, it is not a simple system combination work. It may be the most effective way for certain business intelligence applications. I mentioned TERADATA and Microsoft Analysis Services earlier, I am willing to focus on data mining and advanced analysis.
For Teradata's users, I have a high evaluation of Teradata Warehouse Miner (TWM) regardless of whether they want high-performance forecasting tools or simple and effective demonstration tools. All of its data mining features can be in TERADATA data warehouse. Or any way you want to perform SQL. This feature can save a lot of funds because it can explore the value of the company's TERADATA data warehouse platform. If you want to implement TWM, I recommend you to build a Windows-based platform, not only for TWM client technology, but also to establish an active data warehouse (Living Data Warehouse). A active data warehouse is a system that has been created, and it is maintained as a hierarchical sample of the primary data warehouse, and it is regularly refreshed.
Consider this suggestion carefully. Your store has a huge TERADATA data warehouse running on the NCR server: an investment that can efficiently manage huge data sets. Now your company has a small but accurate active data warehouse, which describes the data in a large data warehouse, which is the Teradata database running on the Windows platform. It is like a mini-type data warehouse, the key uses the Windows version of the Teradata database, not TWM. This non-traditional combination expands the opportunity of the entire report, excavation, and analysis of applications, and do not continue to invest. You can effectively use the active data warehouse to reach the following purposes: 1. Establish a mining model
2. Run the mining application when the sample is sufficient
3. Develop and test SQL reports depending on the Teradata data warehouse
4. Establish prototype dependent on TERADATA data warehouse
Use Web Services
For many companies, business intelligence needs to be larger than the available budget or processing capabilities. Web Services is a promising solution that allows business intelligence to meet demand and do not invest in commercial intelligence itself. Web Services Relying on Simple Object Access Protocol (SOAP). SOAP allows Web Service to standardize, allows for programs written in different languages, and clients from different operating system platforms access the service provider's Web Service. SOAP is a standard Web protocol, including XML, HTTP, TCP / IP. Web services allows different applications to work together, loose coupling, sharing analysis services and data in the heterogeneous environment. Analysis of geospatial spaces. Your customer's location is closely related to your website, you can rent an analytical service in the transaction location. By measuring customers and your local distance, you can get a precise geographic concept, know where most of your customers live. Combining statistics in the same area will give you a predictable fees. If you evaluate websites, selling regional distribution rights, leasing negotiations, or conduct postal activities, then this information is very important.
Buy the packaged geospatial solutions and related statistics, it may cost from a few hundred dollars to a few thousand dollars. With web services, you can simply send your store and customer address to the service provider, such as ESRI, then return the map of the two trading areas and the small-spending report. Integration is also an ideal way for Web Services, which is also an ideal way to buy a lot of data and tools. Especially for geographic space analysis, space data, transactions and other location data is almost like daily necessities, you can access the Web Services, Web Services is responsible for maintaining data real-time, regular updates. The result of doing this is saving time and expensive overhead, and there are more additional benefits, which is always the latest content. Web Services allows you to pay according to the number of transactions or discounts in advanced trading volumes. The necessary investment you pay is avoided to avoid a heavy burden of maintaining the latest technology and human cost. If you want to get greater value from your business intelligence investment, consider the satisfactory dedicated analytical services brought by Web Services.
Database's built-in function
The seller's clumsy repair from several data sources has always been traditional business intelligence. Reasons are very simple: database manufacturers have previously provided very little business intelligence. Therefore, many commercial intelligent applications use SQL syntax from the database from the database, put them on the application layer, then convert the data to the required granularity, and then the commercial smart app is another thing. Now the traditional method is no longer necessary, just because most project planners have become accustomed to it. Today's main relational database vendors, including IBM, Racle, Microsoft, and NCR / Teradata databases have built-in business intelligence features. Currently on the road to built-in business intelligence functionality, advanced statistical analysis functions such as data mining, OLAP, ETL, spatial geographic analysis, and regression, variance, sampling, queue. Discover your inner business intelligence function in your relationship database: 1. Enable database to send accurate data to business intelligent applications
2. Transmit the scan, sort, connection, and aggregate data to the data warehouse server
3. Reduce the total amount of data from network transmission
4. Expose less data in a smaller security area outside the firewall
5. Reduce the risk of inconsistent data
The business intelligence features built into the database not only meet the entire enterprise-class real single version, which does not have to care about the end user's tools and applications, and reduce expenses. You don't have to build infrastructure (CPU, memory, network bandwidth, and software) to support traditional business intelligence mode. Instead, the database will be ready to optimize and accurate data.
Pay attention to what you need
As a strategist and data architect of the solution, I often have to balance between the company's prior technology, available budget and new demand, and my more than half of customers tend to invest in powerful, well-known brands. Unfortunately, they just used a small part of these product features, which is an expensive investment. And other customers are very sensitive to prices. They try to implement the technology of minimal investment, but later discovering that this method requires a lot of workload, and finally leads to greater investment. In the process of realizing business intelligence applications, when you need to trade off when you re-combined tools, you can use their features and fill their respective defects. When the budget is more tense, the appended tools need to be low. Re-combining and weighing the experience of business intelligence, let me always check three key architectural methods: non-traditional technologies, using Web Services, extended business intelligence features in database built-in database. Of course, there are other ways to save business intelligence budgets, but first investigate those methods to determine how to reduce investment can also provide a robust business intelligence solution to implement.
