AGILE is mainly a mindset than a set of rules.
AGILE is not reading the AGILE manifesto is understanding it.
The steps are:
- Understanding what means value for your customer.
- Removing all the bureaucracy that “you need 5 signature” to create a software release, instead having a full automated process that checks the product, almost, without human interaction.
- Creating value for your customer.
- Having processes in place that will be overridden when needed – people first.
- Implementing an AGILE practice that work for your organization, yes nobody uses SAFE in all its steps.
- Nurture and coach people before everything else, people are the engine of AGILE.
- The optimum is enemy of the good, better having good things fast than perfect things later.
- Explore, is better failing fast than “we have always done this way”.
- Analyze specifications, is not coding but coding without analysis is like inserting bugs in a textpad.
Let me explain every step.
Understanding what means value for your customer.
Customers like fast results that will solve their needs, I.E: talk and ask, then ask again, then write down, then share your way with the customer, then start to build the software.
The bureaucracy dilemma
In the software industry, like in war or in marketing, the process must be lean to react faster, release faster, give value faster; get rid of all the five levels of authorization , empower your people and watch the results.
Creating value for your customer
Customers need answers and help, I.E: having a great UX is more important that having a system that is fault tolerant from the day one.
Are rules that clarify and simplify the flow from point A to Point D, sometime you have to use shortcuts to avoid shortcomings.
The only thing that you have to avoid is an exception-based development.
SAFe could not work for your company
Yes it could be overkill, too organized, too perfect.
Many organizations thrive in uncertainty, embrace it or better handle it.
The people do the work, create value and could destroy value if not valued or never heard well.
Yes it could break with a probability of one in a million
Cannot cost 10 times the cost and to be perfect must pass the quality processes that your company have in place.
Don’t be afraid of failing
POC and prototypes have a cost, but starting straight to invest 100.000 € in your next product is not the solution.
Don’t start to code, analyze first
A Good analysis must not be 200 pages but cannot be done on a tissue of the pub.