As a service provider, it happens all the time that some companies request our services or in need to have immediate solution for their current problem regarding SQL Server, Once we provide them our proposal and associate cost, they usually put a side our proposal and mention that ‘We are charging beyond market and other competitors are providing same service with far below cost’.
There are few areas that causing to have above respond:
- Client would like to compare everything ‘Apples to Apples’.
- Client would like to get the cheapest in the market without thinking about consequences.
- Competitor is in the market longer than us.
Based on mentioned areas, I would like to write up specifically.
Client would like to compare everything ‘Apples to Apples’
Not everything can be compared with Apple to Apple technique, as a service provider, lots of factors affect the cost, for example the consultant expertise and knowledge. We might be able to compare tangibles with such technique but not services. For instance, our competitor in Malaysia also does provide such service as ‘Performance Tuning’, at first you might think we do provide both services, but based on our collected data, we found the following differences:
Fard Solutions Competitor
- Workload Analysis
- Database IO Balancing
- Table Re-Structuring (Transparent to Application)
- Index Analysis and Tuning
- Query Design Analysis and Tuning
- Execution Plan Analysis
- Windows Configuration
- SQL Server Configuration
- Using Profiler or Perfmon
- Rebuild Indexes
- Using Database Tuning Advisor
We bring our depth SQL Server knowledge and expertise to provide highest quality service in the market. We understood that our service cost is quiet high in the market but it is one-time-off, as we did for Bursa Malaysia Berhad and GHL Systems Berhad. Competitor service result only can sustains for 4-8 weeks only, but our solution sustains for years.
Client would like to get the cheapest in the market
You get what you pay for!, our competitors are providing services with very low cost, but what is the consequence costs from their service. Let me explain to you! For example for ‘Performance Tuning’ service, they will just do some very basic tasks, then once the performance decreased again, they will suggest you to scale up the hardware. Scaling Up the hardware comes with a cost more than we usually propose due to software licenses. In some cases client needs to scaled up the server memory based on competitor advice from example, 8GB to 40GB and increase number of processor cores (This dedicated server only holds a database with a size of 7GB), at the end they still needs to restart their server every day due to not tuned queries and bad database design.
Competitor is in the market longer than us
This is a common excuse that most of client will give, the competitor is in the market longer than us. It does not matter how long they are in the market if they do not have sufficient knowledge and expertise to accomplish the project successfully with no consequence hidden costs. Our company name is listed on Top 191 SQL Server Consultants and Service Providers in Microsoft website. We are expert in the field and trusted by Microsoft.
Case Study – Bursa Malaysia Berhad
Bursa Malaysia Berhad is an exchange holding company. It operates a fully integrated exchange, offering the complete range of exchange-related services including trading, clearing, settlement and depository services.
Automated Risk Management and Decision Making Analysis (ARMADA) System is the centralized system for BURSA MALAYSIA BERHAD to collect and process all the stock market data from broking houses and investment banks.
Problem: When multiple brooking houses were submitting their stock market data into the ARMADA system, then database was encountering blocking and deadlocks and the victim broking house should submit the data one more time. Total data processing was about 22 minutes for each broking house submission.
Solution: By changing some of the table structures (Transparent to Application), Creating new effective indexes, changing 150 queries (Static and Dynamic), we managed to reduce the processing time from 22 minutes to about 5 minutes with concurrent data processing and once we done our performance tuning project, multiple broking house could submit their data without encountering blocking or deadlock.
By providing such service, we managed to reduce the Memory and CPU consumption significantly. The following figures illustrates the performance before and after ‘Performance Tuning’ service.
Figure 1: CARCompute (Main SP) -> Full Scan (Blue Line) – Before Service
Figure 2: CARCompute (Main SP) -> Full Scan (White Line) – After Service
Dear reader, not every proposal can be compare via Apple to Apple comparison as service quality matters.