4 Simple Suggestions that Make a Better Testing Strategy

Aug 22, 2017 by Anna Heiermann

As a QA Lead, I’ve had a few challenging instances where I’m delivered a new application, or existing enhancement, and without any real direction I’m told to “test everything”.  If no one is dedicated to spelling out some requirements, much of the assessment and understanding will fall on QA.  I strongly feel that the worst thing you can do is nothing, and just start testing.  Following these simple suggestions can provide full exposure to your team on gaps early on. 

First, get acquainted with the change from a technical side.

Ask for a demo session with the developer(s) making the enhancements. Have them walk you through the code.  Ask them questions like, "Have changes/updates impacted other applications due to shared code/apis/systems/?”, etc.  Ask about impacts to tables/wherever the data is stored.  Escalate immediately as anyone impacted by this needs made aware so they can plan. 

Next, conduct some basic risk assessments. 

Start with an Excel sheet listing the functions of the application at a high level, (example:  registration page, credit card & information page, site security, etc).  The key here is to make it as “simple” as possible so that your team can easily identify the risk level of that function.  If needed meet with business and technical teams to make sure all of the functions are documented. 

Set up a meeting with the team and stakeholders. 

Using the list from above, have the team and stakeholders assess each function by priority.  If they feel everything is critical, challenge them with questions such as, “If the whole application was to stop working, what is the first thing you need as soon as possible?”, to help them really nail down what functions are the highest to lowest priority.  After the meeting estimate the number of tests by each function and priority assigned, (if you don’t know, write out the scenarios you need and go with that number as a start).  Your sheet would read, for example:

Credit card transaction, High, 30 tests;                                               

Username creation and registration, High, 50 tests;

Top Navigation Banner, Low, 10 tests

Finally, forecast all manual and automated tests against each business day by how many you plan on completing each day just for a baseline.

Start escalating immediately if the timing is compromised or the estimate for testing is very high.  Now that you have a risk analysis sheet everyone can identify where the focus of testing should be first, and from there you can start working with the scenarios you need at a minimum based on technical and business concern.

A wise Test Manager once told me “Numbers Don’t Lie”.  I’ve found that when you can provide numbers in some way, whether it be a template, or a chart, it’s easier to help your stakeholders understand your concern and have what they need to make decisions on the solution.  The sooner this is escalated, the better for everyone!  

Load more comments