How to promote the Quality Best Practices

Two ways to promote the quality best practices are: Influencing the teams or to enforce the best practices. A periodical team reality check about the current practices is the beginning of the process. A team can easily live in their own Ivory Tower or to be biased by this statement: “We always did it that way so why changing it?”.

Influence or evangelize

Discussions: gather information about current pain points (retrospectives and lessons learned)
Involve QA Specialists in all SDLC (or Software Development Life Cycle)
1:1 meetings to get information and expose point of views, share ideas and explain the best practices: it may be easier to convince on person than a group
Talk to the teams
Provide existing articles and blog posts
Tech talk to expose the vision
Create training material: guidelines, processes, videos (e.g.: Pluralsight, etc.)
Involve people in quality solutions:
- Organize Workshops (on short term) so that the targeted public becomes active participants
- Organize Work Groups (on long term) so that the concerned people become active participants
To use data as leverage; more information is available in the data-driven decision section

Data-Driven decisions

To show metrics of the current situation (or challenge the “no metrics” or lack of metrics). It is the also called “Shift-Right” approach in the Software Quality world.

End-users or Customers survey can be used (e.g.: use a post-transaction experience survey on the mobile app or web site)
White-Label partner surveys
Use production dashboards (examples: Datadog or SignalFX metrics) and performance tools (e.g.: WebVitals, etc.) to monitor APIs, SQLs, etc.
- Daily performance emails can be sent to the engineering teams
NPS and CSAT surveys: NPS vs CSAT definitions
Real Customers Journey (traces of the end-users through the applications)
Use support CIM dashboards: customer tickets, production tickets, ticket aging/bug fix time, tickets by services or teams, RCA (find the issues caused by quality process)
- CIM stands for Customer Interaction Management
- A way to generate CIM data is to use a tool like eazy bi
Correlation between test coverage (static code) and production issues
- It is also possible to get test coverage from a live service using a tool like Sealights
Compile the functional testers (manual tests) created Tickets/JIRA vs current team or service test coverage
Create Success Stories: start with one team and use the before/after statistics
GIT statistics (e.g. with a tool like GitPrime (now Pluralsight Flow)) in order to correlate bugs/tickets and developers stats (e.g.: number of commits, active days, etc.)
- I.e.: low stats may be caused by developers not working on features or test automations because they are working on manual tests or production support
Correlate the release cycle (how often the team releases or time to market) and test coverage; better/reliable test coverage and test regression suite will shorten the release cycle.
Gather statistic about how much time it takes to refactor an existing feature? Better test coverage will result in faster refactoring

Enforce

Contact the best practices stakeholders and ask for support (I.e.: to push the best practices downward)
To fix deadlines (must be adopted by specific dates)
Enforce with Gates, checklists, DoD, etc.
Set Quality specific OKRs
Assign Quality trainings