Share
Embed
Video
Transcript
27 slides
-
why we cannot increase code quality and work in legacy
? why we write tests, do pair programming and collaborate ? why the features we created are not used as expected and thrown away ? why I cannot improve myself and do the same things over and over again ? ALWAYS LEGACY NO TESTS & SOLO WASTE FEATURES LOSING SKILLS -
the team is rushing to catch the deadline get-it-done mindset
overemphasis on efford estimations primary goal is to complete all tasks no time for testing in development custom frameworks are created to narrow focus and understanding new teammates join to go faster technical dept is accepted -
individual speed and productivity are crucial no time for helping
peers and working in pairs hard to onboard juniors to team critical tasks go to seniors micro management happens -
metrics are collected to manage productivity story points per person
is calculated tons of metrics with no real meaning are collected, like team productivity, velocity, code coverage performance appriasals based on metrics destroys justice & team spirit -
plans should be obeyed scope is fixed we know customers
will like it product owners as as analysts, no need to know the product internals feature prioritization is for catching the deadline poker planning for hours only few people touch and hear real customers -
lords of the house are the decision makers know-how is
kept at seniors and architects, who form a separate team everyone has separate duty, has limited visibility on the big picture lack of skill development for others high turover rate with many outsources team mates -
all the problems you face are normal if what you
build is a project it does not matter if everyone say you build “a prouct” what you build quacks like a duck, and walks like a duck, then it is a duck (a project) that is the bitter truth -
software product is not a software written for a customer
by a team a green-field project with latest technologies used an output of a Scrum team running sprints something developed iteratively something you demo at the end of every sprint an output of the team using a board and tasks a product if it is been built as if it is a project -
building a project is like taking care of someone else’s
child what you really care is getting the given tasks done and get paid for it -
you follow long term vision and a purpose learn from
every failure aim is sustainable satisfaction & growth there is no end for growth and development look for continuous improvement seek for good people and processes building a product is like growing your own child -
Product KPIs Time https://medium.com/ctonun-el-defteri/startup-dinamikleri-ve-cto-rolleri-18af101649cc Hakan Erdoğan, Startup Dinamikleri ve CTO
Rolleri EARLY STAGE SCALE STAGE LATE STAGE Main concern: Speed Quality Business Continuity Main Focus: Validation Growth Customer Satisfaction Customer Development Sustainability Customer Loyalty Developed: Product Projects MVP products start to die when it loses capability for innovation and adaptability SIGMOID CURVE OF BUSINESS -
Product KPIs Time SIGMOID CURVE OF BUSINESS SECOND CURVE THEORY
AT RE-SCALE STAGE Main concern: Quality Main Focus: Growth Customer Satisfaction Developed: Product first curve second curve -
from building products LESSONS LEARNED 10 to show you how
building a product is different than building a project -
estimations considered waste 1 efford/complexity/time estimations are never accurate due
to high degree of uncertainty and variability, diverse perspectives, insufficient knowledge and unreliable assumptions care customer priorities, quality and purpose more -
seniority levels are toxic, get rid of them product success
needs collective ownership and decision making, and collaboration of the whole team. expect craftsmanship and professionalism from everyone, and let everyone initiate leadership. 2 you don’t have time of bureaucratic bullsh*t! -
be an expert of your product & domain without full
knowledge of the product, you are not a part of the solution, you are a part of the problem recently developed with latest technologies. 3 good products are built by “product” developers -
program in pairs and with team do together, practice together,
learn together, improve together 4 slow down to go faster -
cultivate a quality-oriented collaborative neighbourhood pressure 5 hire like-minded, disciplined,
motivated, disciplined talents, give them the purpose, principles and values and let them shape the culture quality is an output of principled & disciplined teams -
get the best feedback you ever get touch/support your customers
directly know how customers feel, what they want, what they don’t like. developers not touching the users develop projects, not products 6 -
use technologies you are good at 7 or select the
ones you don’t know, but you can expertise fast how can you claim you can deliver quality when you are not good at the technologies used ? be pragmatic and wise -
always be ready to deploy to prod release changes early
& frequently 8 invest in CI/CD pipeline from day one shorten feedback loops with efficient pipelines and make changes available to customers fast -
quality & productivity cannot be measured don’t fool yourself, stop
measuring it when a measure becomes a target, it ceases to be a good measure - Goodhart’s Law 9 -
focus on building something in a way that you feed
proud forget about agile {placeholder} stuff 10 -
1 cultivate neighbourhood pressure be an expert of your product
& domain program in pairs and with team touch/support your customers directly use technologies you are good at release changes early and frequently don’t fool yourself, stop measuring productivity 2 3 4 5 6 7 8 9 estimations considered waste seniority levels are toxic, get rid of them focus on building in a way that you feed proud 10 10 lessons learned from building a product -
One-Stop-Shop Payment Orchestration talent @ craftgate.io we are hiring! LEMİ
ORHAN ERGİN co-founder, craftgate speakerdeck.com/lemiorhan