When deploying a new CRM, we must consider how the system performance between the client and servers will factor into the CRM’s performance for users. Response time plays a key role in users' ability to adopt the system. Before the CRM goes live, it’s important to determine the CRM response time. This data creates a baseline to draw from in the future.
Why is it Important to Have this Data to Reference?
At the outset of the CRM deployment, the response time for the system is usually good. However, when new users are added, or a new version of the software comes out, it could slow that response time. System users may not recognize these slowdowns until they start affecting their ability to use the CRM. On the other hand, users could also perceive the slowdowns to be much more significant than they are. In both cases, having a baseline of data to draw from helps the organization determine if the slowdown is in fact affecting performance, or is just a minor blip.
How Do we Create a Baseline for CRM Response Time?
Baseline response time analysis is old school technique, but it works.
It involves taking the most common 10 to 20 transactions and using a stop watch to measure the time from when you hit Enter to when the results show up on the screen. Perform this test over multiple days at different times on each day, and you’ll get a fairly accurate average response time. A template for this testing is available at the bottom of this post.
If possible, it’s best to also know how many people are logged on during the measurement, or the number of concurrent users in the CRM when measuring. This can be discovered using either log files or via the Administrator tools in your CRM.
What Do we Do with This Baseline Response Time Information?
Our goal is to measure the effect of change on user response time. Change to the CRM comes in many forms:
- Increase or decrease in the number of users
- Patches or updates to the system
- New code
- Poorly written algorithms
By redoing the evaluation after you’ve made changes to the system, you can determine if there is an impact. First, look at the results and compare them to the baseline. Did performance change? If the answer is yes, the next step is to determine what caused it. This will affect how to move forward.
You can determine if the performance change is based upon user workload, which is common. Divide the number of concurrent users at the time you did the baseline tests to come up with a ratio performance by users. Then repeat this for the new tests. If it is just a matter of additional user workload, then the ratios should be close. If not, you have a problem with an algorithm
If the response times of the newest evaluation is significantly larger (a multiple larger) then you have some algorithm problem. If the growth is in uniform across all user functions, then it is a system problem. It if is limited to a single function, it is most likely a programming issues.
Having a proper snapshot of the system performance just after going live will be invaluable for analyzing problems in the future.
Use the template below to help you get started on CRM response time analysis, or contact our team of skilled experts.