Breaking the Cylinders of Excellence (in the Australian Government)

YOW-Nights_Logo_stackedAt the recent YOW! Night in Brisbane (as well and Sydney and Melbourne), Lindsay Holmwood (the Head of Technology at the DTA) presented “Breaking the Cylinders of Excellence”. It was a rare experience to hear the story of how the DTA is using cutting edge development practices to help the government catch up with, and even exceed, the public sector. 

 

  • DTA – aid transformation in government, small agency
  • Delivery hubs in Sydney and Canberra – help identify and plug capability gaps in teams
  • Prototype of how government services could work  gov.au/alpha
  • Digital Service Standard – 13 characteristics on what good looks like in government, useful in organisations as well
  • Cloud.gov.au  – government cloud service, usage growing, continuous delivery pipeline (which is a major change for government who are used to 2 changes per year)
  • The unit of delivery is the team – not about individuals, but the team – borrowed from GDS
  • Government is slow, but government is designed to be stable, they cannot fail, they have characteristics that are resistant to change
  • Myth that organisations must choose between speed and reliability, high performing organisations deploy more frequently, have shorter lead times, fewer failures and recover faster, but they also have a greater profit
  • Want to deliver like a startup but be stable like a government
  • Not a lot of cross pollination between departments currently
  • Read the policy! – quite often the process is not mandated
  • Document what works and doesn’t so it becomes a repeatable pattern – ie. running a meetup inhouse, don’t tell me I can’t do it, tell me how I can run it without being thrown in jail!
  • Stick with technologies the government is comfortable with if you are changing the delivery engine
  • Security matters – prevention is a battle you will always lose, detection is your best defence – aggregate and log in one place, identify threat signatures, etc
  • Embed security people on big services so it is part of the architecture
  • Proactive testing between different governments around the world on similar platforms
  • Simplest security breaches make the most mess – infected excel macros, leaving free USB keys in the foyer that are malware infected
  • Need to put user needs first – alpha mockup using tools like Jeckyll, then beta then live
  • Lots of people strictly interpret the design and delivery guides – they are guides not rules!
  • Create a longer runway by pulling tech forward – turn down the volume of design, turn up the volume of tech
  • If it hurts, do it more often!
  • Fixed cost delivery with agile is a thing, agile is a way to de-risk in the government
  • Don’t put manual testing on the critical deployment path – have special skills on hand for accessibility, performance and security
Advertisement

Tech Connect 2013 Brisbane Review

Tech Connect 2013With thanks to my very good friends at SlatteryIT, I headed off to Tech Connect 2013 in Brisbane this week to network with the Brisbane startup community.

Brisbane Lord Mayor Graham Quirk kicked off the day by reaffirming the Digital Brisbane strategy which already has seen the appointment of a Chief Digital Officer, a visiting entrepreneurs program and a strategy to attract more startups in Brisbane. He was keen to see more success stories in Brisbane, following on from start ups like Halfbrick, the inventors of Fruit Ninja.

Here are my notes from the sessions.

Enabling Entrepreneurship

Tyler Crowley is the first speaker to be brought to Brisbane as part of the visiting entrepreneurs program. Tyler is a well known and respected entrepreneur, best known to me as the co-host of the This Week in Startups podcast. I am not sure if there was meant to be any direction to this session, but it turned out to be a rambling question and answer session.

  • bring awareness to the community by getting people to tweet for being at an event – you will be surprised how quickly the city will get recognised
  • startups tend to congregate in the same area – in San Francisco it is around the Twitter office, in Palo Alto it is around University Avenue and in Los Angeles it is around Santa Monica, places that have startup centres benefit due to the cluster effect, cities like Stockholm are suffering because of a lack of this
  • River City Labs is probably currently the nexus of the community in Brisbane
  • the crowd funding model is still on the horizon and should be awesome for startups outside Silicon Valley, AngelList just got approval to have a crowd funding model from the SEC in the last week
  • another company getting acquired in the next 6 months will really put Brisbane on the map
  • attracting VCs – LA does content very well, San Francisco does social networking very well, so attract the kind of startups to the industry that you do well and get that message out
  • Brisbane has incubators like River City Labs (private) and ilab (government / university)
  • documentarians are important – TechZulu in LA and Scobleizer in San Francisco – provides a window to everyone outside, TechZulu is a great model, it took two years as a labour of love until it became profitable, the mainstream media follows when it becomes popular and it will explode
  • Brisbane is in the frustration phase of funding and media coverage – they will take notice when they become embarrassed by the success of startups and documentarians – currently at the tipping point, it is inevitable!
  • if your city had a blank canvas – getting a nest is important, support the documentarian, support angel events, support local meetups and events!, hold regular monthly events, calendars and job boards, strategy to attract outsiders to get the “Apple Store effect”
  • This Week in Startups – is the global meeting place to inspire people
  • build a startup map of Brisbane, one exists of Australia, maintain a database like CrunchBase
  • it is time for the banks to wake up and support startups like Silicon Valley Bank who are now spreading as the major banks are asleep at the wheel, there are opportunities for supporting industries to step up

Think Big!

Matt Barrie is the creator of Freelancer, which is the worlds largest outsourcing marketplace. I was really looking forward to seeing this talk and it did not disappoint! His presentation is also available online.

  • software is eating the world – the biggest bookstore in the world is digital (Amazon), Scrapbooking (Pinterest), Evernote, maps, music, yellow pages, fashion, money, real estate, jobs, etc, etc, etc…
  • 66% of the worlds population are yet to join the Internet
  • demographics are changing and aging – lots of opportunities in this space as well
  • lots of online learning opportunities – you can design logos easily via Envato, Stanford University had 170,000 online students enrolled in an Artificial Intleligence course that normally attracted 250 people and the highest achievers were outside Harvard, there are also options like Coursera and Khan Academy amongst others
  • created Freelancer.com – we are now a service economy, the world is becoming very globalised, crowd sourcing (outsourcing has now turned to crowd sourcing eg. logo design)
  • Exposé the Freelancer.com logo! – crazy pushing the boundaries of crowd sourcing, also Kickstarter and the Pebble Watch – wanted $100,000 to build, raised $10 million!
  • Kickstarter funds more projects in arts than the US government, the next big thing is musc
  • The World is Flat
  • parabolic growth comes from distribution firehoses eg CityVille growth in Facebook, Viddy for reading news in Facebook, Google, Reddit
  • you need to strike early, before the idea gets crowded eg. games in the AppStore like Angry Birds who owned the market early
  • the new metric is growth – referrals are extremely important – see Startup Metrics for Pirates
  • all the software you need is free or cheap
  • Reddit.com ha a sub-site for everything
  • sites like RetailMeNot started with $30, Digg started with $60
  • the first dotcom bust was due to bad business models and a strong reliance on advertising, but in reality we are still in the original boom
  • Zynga – 96% of users don’t buy anything, the 4% that bought a cow to impress thir friends raised them $1.2 billion
  • many companies are not ready for growth, a mess internally and financially
  • there is lots of potential to replace people with algorithms and insights
  • in Australia we need to find a way to build technical businesses in the financial sector, like we do for mining
  • enrolments in engineering is down 60% in the middle of a technology boom, we need to start seeding interest back at school level
  • any job that can be described by an algorithm can be turned into software
  • Australia is a good place to base a startup, easier to hire people and start a network, but our exchange rate is crippling and if it rises it could become a major problem
  • you can build a big business targeted at the local market (eg Realestate.com.au), but always think global
  • Mary Meeker presentations should be used more in presentations

Money To Grow

This panel included John Hummelstad (Ephox and Concept Safety Systems), Sean Teahan (Nimble) and Doron Ben-Meir (Commercialisation Australia).

  • options include borrowing from friends, earning a buck first and investing your own money, earning from the bank (usually $2 million and up), funding by credit card, selling something, grants and R&D concessions, pay in 60 / collect in 30 days, angel network, information memorandum by getting close to OEM’s particularly those who you might potentially sell to, Venture Capital but this is decling, crowdsourcing
  • fund like minded companies to work closely with you
  • relationships at a strategic level are hard, but can bring rewards in the long term
  • CFIMITYM – “cash flow is more important than your mother”, business owners are awake at 2am worrying about cash
  • lean startups work very well for technology types of businesses, cash hungrier models need to exist because we still need to build products like iPads
  • figure out who in the supply chain cares about your type of business – they are good areas to look for investors
  • the large amount of successful businesses do not have venture capital funding – should only consider this if it brings value
  • essential that you break every rule in private funding that you can’t break in the public market
  • Australian Government has uncapped 45c / dollar R&D tax incentives (quarterly in arrears), then you move to incentives from Commercialisation Australia
  • Enterprise Connect is great way to take your business through the washing machine – your business needs to be able to stand up to audit
  • getting a commercialisation grant – it is your onus to prove that your invention works, once that is proved the commercialisation then needs to be tested
  • make sure when meeting potential funders that you can answer how you are going to solve their problem, also make sure you know all about them, use tools like LinkedIn and be educated, it is just courtesy
  • important to have front foot sales to fuel the fire but most importantly to get validation from the customer base, it also builds credibility by reinvesting your profits, biggest issue with technical companies in Australia is there an aversion to being a saleman, it is all about selling and if you are not prepared to do this then don’t start!, you are selling yourself not the product
  • selling is learning – you need to learn what your customer wants and your value proposition

Accelerating Growth

This panel included Natasha Rawlings (StreetHawk), Ric Richardson (inventor of software activation amongst many other thing) and Steve  Baxter (River City Labs). Ric mentioned that he gets lots of attention from his appearances on Australian Story (The Big Deal and A Done Deal).

  • your first role is not to be the CEO but the Chairman of the Board (looks at the business to ensure they have the ideal CEO and are delivering to plan)
  • find the right people and partner with people who have done his before and hold them to their agreements
  • the right investors can bring you the introductions to partners, even if they decide not to invest in you
  • always deliver a good product and don’t piss the customer off – service is still important
  • raising equity starts with a plan – do this only when you need the funds, it should be the last resort, ensure you have a capitalisation (cap) table so you can understand what will happen to your equity – start with 30% for founders, 30% for management and 30% for investors
  • always start a business by looking at what it will look like when it is finished
  • Stanford University has a useful entrepreneurial course, be disciplined when following the Lean Startup model and have a plan
  • build a better prototype and often it will sell itself, other people will tell you quickly what it is worth
  • network deeply, meet people twice, ask questions even the ridiculous ones
  • investors are there to support and provide leadership, when you start pulling out agreements you know things are going wrong, like to know that they have listened

Setting Up for Global Success

This panel included Brendan O’Kane (OtherLevels), Jeremy Colless (Artesian Venture Partners, which was spun out of ANZ) and David Israel (UniQuest)

  •  base yourself close to your prospects
  • important to think global from day one, particularly in the technical field, be worlds best rather than Australia’s best
  • take advantage of the Australian talent spread across the world, utilise international students particularly those from north of Australia

Building Value

This final panel included Anne-Marie Birkill (OneVentures), Bob Waldie (Opengear) and Steve Baxter.

  •  people are cheap – premium prices do not work, people care about price
  • service is key – give good service
  • businesses are not charities, lifestyle is not a sustainable currency, you need to make money
  • ideas are not traction – get off your arse and learn what you don’t know
  • when looking for value, you are looking to triple an investment
  • success is building a valuable, non-charitable business, a second round of investment is not success
  • don’t think you can take a great technology idea to a crappy service industry eg taxis – they are not interested
  • make sure you formalise arrangements, in case things go bad, write down the exit conditions and make sure you are aligned

Overall a great day of presentations, panels and meeting new and interesting people in Brisbane technical and startup community.

Agile Australia 2011 Day 1 Review

Agile Australia 2011Agile Australia 2011 was held for its third year last week at the Hilton in Sydney. Once again I was honoured to be offered an opportunity to present, be an MC for speaker sessions on both days, moderate a panel discussion and run the end of conference retrospective. The conference attracted 675 attendees and the buzz over the two days indicated to me that the conference was a huge success.

For the second year, it was a great pleasure to be one of the conference advisors. As the conference was brought forward to June, there was only six months to prepare between conferences and lots of suggestions and improvements to implement from previous years. A lot of review, debate and discussion went into putting the program together and ensuing there was a good mix of speakers, variety of topics and sessions for different levels of expertise. More effort was also put into shepherding speakers. A huge thank you needs to go to Rachel Slattery and Zhien-U Teoh from Slattery IT for their commitment to the conference as well as my fellow conference advisors Phil Abernathy, Adam Boas, Keith Dodds, Martin Kearns, Dave Thomas and Nigel Dalton.

The following are my notes from the sessions I attended on the first day.

Keynote: On Beyond Agile – The New Face of Software Engineering

Alistair Cockburn delivered this keynote, the slides are available here.

From Agile Australia 2011
  • agile software development is for wimps
  • cooperative game – invention, communication and decision making
  • projects are in a position, look for strategies to move our position, no fixed formula for winning the game (competitors and the economy are some of the enemies), only three moves to invent, communicate and decide
  • communication – whiteboard discussion provides stickiness over time (can just point back to conversation) as well as proximity
  • need about 3 minutes of video to enhance the distributed conversation, becomes archived documentation to remember user point of view or architect design decisions
  • craft – a lot has changed, software development changes every 5-10 years and you need to keep up
  • people learn skills in 3 stages – shu, ha and ri
  • work in progress is decisions that have been made but have not been shipped and delivered
  • like lean we we want multiple deliveries per day – continuous integration has evolved to continuous delivery
  • in decision making, look for the bottlenecks, the person with the full inbox is the person limiting the work in progress of the whole organisation
  • knowledge acquisition – real moment of learning often happens at the end when the surprise factor occurs when we deliver work, suggest that at the beginning of a project deliver a knowledge curve ahead of the cost curve (a number of small experiments)
  • agile says deliver highest level of business value first but projects tend to always deal with the risks first – learn about your business risk (should we build it), social risk (do we have the right people), technical risk (API’s, performance, architecture), cost/schedule risk (gain knowledge about the solidity of the estimates) – you need to decide whether to deliver business values or knock off some of these risks
  • need to identify the tail to determine whether we deliver business value early or add more features later (Apple are good at this, for example shipping an iPad without 3G initially)
  • self awareness – the team is self aware when the team can talk about the team and where they are

Keynote – Is Business Ready for Agile

Rob Thomsett delivered this keynote, his slides are available here. He advised that he was going to run his talk in two sprints and check the heart beat halfway

From Agile Australia 2011
  • agile is a church of all people – the newbies through to the true believers and a few drooling old people
  • agile is not new – been going for 40 years
  • is business ready for agile? – yes and no – every company in the world is ready for agile, they just don’t know it – how do we develop agile into a broader organisational paradigm
  • we work on a set of models that were developed when the world was relatively stable
  • the average window of stability for an organisation is about 3 months, change is normal and everything changes
  • old business techniques have reached their use by date (Gary Hamel – Management 2.0)
  • 100% of C-level executives believe that project management is too bureaucratic, projects take too long, business cases are poorly developed, transparency is adequate, they expect to be ambushed, steering committees are a waste of time, reports are not accurate
  • in 1968/69 NATO held a conference to address a perceived crisis in software engineering – sound familiar?
  • software is not engineering, wrong paradigm, it is a craft
  • the closest to what we do is not making buildings but making movies
  • summary: we took the wrong model, flogged it to death so let’s throw it out
  • agile is a cultural and disruptive journey – first question to ask is are you up for the cultural challenge, for every company that says no there is one that says yes
  • business approach needs to pass the simple and transparent test – most powerful test to clean up broken processes
  • go back to work and annoy people by asking people if we can make it simpler
  • cultural values are openness, honesty, courage, trust and money
  • the people at the top are the easiest folks to get on board with agile
  • most people link budgeting to estimating – agile demands we move money around more quickly
  • sponsors must get dirty – they must be part of the process because they own it
  • PMO should exist for resourcing, not reporting

Panel – The Changing Role of the CIO

Beverley Head moderated this panel with Jeff Smith from Suncorp, Steve Coles from Allianz, Daniel Oertli from REA Group and John Sullivan from Jetstar.

From Agile Australia 2011
  • key is to turn decision making over to employees and leaders need to become coaches and create a great environment
  • The Corner Office by Adam Bryant gives advice for success – passionate curiosity, battle hardened confidence, team smarts, simplicity, fearlessness
  • role is to understand where the company is going and deliver things for them to succeed
  • relax the old techniques like governance that gave a veneer of confidence
  • need to understand and remove the barriers by becoming an active listener
  • fundamentals like attract and retain the best have not changed
  • learn the business in which you operate and realise the definition of leadership has changed (don’t be afraid to higher smarter people than you)
  • need clarity of purpose, avoid constraint of thought and don’t filter based on your experience
  • resistance at the frozen layer – middle managers are typically the blockers so need to change the communication structure (for example, using Yammer for communication to give everyone an equal voice)
  • distributed teams always need a local decision point like an iteration manager
  • leaders need to eliminate information handlers
  • offshoring value proposition – you need to decide if your assets are a strategic advantage, do not offshore things that are volatile or if the project is too big to handle yourself (which essentially means you can’t explain it to someone else), offshoring is good because it keeps us on our toes to be competitive and continuously improve
  • need to look at outsourcing from a productivity point of view and not just a cost point of view (we are not buying pencils)
  • life long learning for developers – people have to follow their own course, inject talent and different thinking, look back each year and think about what you added to your bag of tricks
  • most people are capable of learning new skills, that’s the beauty of human beings
  • what does quality mean – quality is something that is fit for purpose and testable and maintainable, quality is everything
  • pushing agile into the business – need to agree on one way of working, once you are successful people want to jump on the bandwagon

Agile Architecture & Design

I had the privilege to introduce Neal Ford for this presentation, and his slides are available here. As I had seen many parts of this presentation previously, I did not take many notes as they can be found across other posts on this blog.

From Agile Australia 2011
  • in the software world we deal with known unknowns
  • spikes are your friends, purely informational gathering
  • ckjm – tool for reporting complexity and coupling
  • don’t pay for technical debt that you may never justify

Key Metrics for an Agile Organisation

It was my pleasure to introduce Craig Langenfeld from Rally Software to deliver this presentation (originally scheduled to be presented by Mark Ortega). The slides are available here.

From Agile Australia 2011
  • cumulative flow – look at the top of the line to see what is the scope and how has it changed (total features), then ask if the team limits their work in progress by looking at the time between the boundary of in progress swim lanes, finally look at the lead times and how long it will take to deliver a feature
  • work in progress limits allow the team to move through work more effectively
  • lead and cycle time report – allows you to see where your bottlenecks are
  • stop focussing on the workers and focus on the work product – so rather than lines of code look at the the value delivered
  • productivity – understand your teams velocity, throughout mapping stories that were completed and carried over
  • earned value – useful to measure how much value we are delivering (the difference in agile is we are actually delivering the value)
  • predictability – answering the question of when we will be done – throughput chart can show you if a team is getting more predictable over time, burn up is used to show predictability of meeting scope, release burndown to show meeting a date and demonstrate additional scope being added
  • use cumulative flow to track the cone of uncertainty
  • quality – defect trends and counts, most code altered, number of changes, etc…
  • net promoter score for tracking customer satisfaction and if it is increasing
  • get customers to vote on what aspects of the product they like and don’t like
  • for cloud computing track the features that are actually used

Leading by Serving

Simon Bristow delivered this presentation, his slides are available here.

From Agile Australia 2011
  • a perfect team is one that can do anything, in control, do anything thrown at them
  • lots of teams act in an agile manner, the leader makes the difference
  • Robert Greenleaf – The Servant As Leader – others highest priority needs are being served first
  • bridge the gap – gaps when making a decision which is the unknowns, bridges the gap to the future by seeing the unseeable
  • one action at a time
  • forcing a decision on someone will engender resistance, some you must persuade – need to listen to connect at the grass roots level
  • withdraw and acceptance – step back from the team to allow them to look after themselves and accept that the team know best as they are the subject matter experts and will get the job done
  • facilitate community – need to build
  • lead using art not science – if you turn to science in agile you will turn to process

Soldering Irons, Consumer Devices and Hardware Manufacturing in the world of Agile Software

Dean Netherton and Neil Brydon for DiUS delivered this talk which was one of the highlights of the conference for me. The slides are available here.

From Agile Australia 2011
  • DiUS working on a fire danger smart meter and technology for charging electric cars
  • how do you demonstrate value when haven’t built the device?
  • had to work around the vendor for the smart meter because they had a traditional process for building the device – aligned project plans around hardware drops and had to simulate the hardware in many cases
  • used wireframes to drive design but had to spend longer on design to ensure it was right – for example, can’t add a bigger battery later
  • first drop was an off the shelf component board to kick start the software
  • second drop was a bare board that was the basic board without the LCD
  • third drop was the plastics without the screen as the component had not arrived so replaced with paper
  • challenge on how to articulate stories – had to break stories own technically
  • used Cucumber to test as there was an embedded USB port on the board – helped embedded engineers assist the Ruby engineers understand how the device worked
  • did continuous integration by plugging a device into the build server, had an issue about flashing the device when the code changes
  • hardware engineers slightly change the design with each revision which had affects on software design as well as having hardware for continuous integration
  • built own hardware prototypes and used local suppliers to cut down lead times (China cheaper but added 6-7 weeks to the lead time)
  • used mocking to show users the idea ahead of hardware being available
  • planned for multiple hardware revisions to allow for late decisions
  • these days you can send a 3D model to a design house and they can pop out a prototype, design exercise ensures that screws line up, etc
  • no excuses for automated testing, in the past it was not embraced in hardware, can test the integration layer without the need for hardware
  • no benefit in running tests for the hardware design as you only get a handful of drops
  • use automated test to ensure buttons and light work, good when you get new hardware and good for checking faults on the production line
  • had to learn about the hardware stack early on which challenged whether value was being added
  • firmware development not integrated onto the story wall
  • technical tasks are OK but really understand what done is
  • luckily the stakeholders were quite technical

Putting It All Together – Agile Transformation and Development Tooling

Philip Chan from IBM delivered this presentation, his slides are available here. I failed to see a lot of agile in this talk personally.

  • established teams – communication difficult across timezones but tools make it easy, different tools used in different teams,
  • IBM agile process – 2 week iterations, 2 day inter sprint break every 4 weeks, develop for first 6 days followed by 2.5 days for bug fixing, do acceptance testing for first 6 days and 2.5 days of exploratory testing, showcase on day 9
  • test management very waterfall for audit purposes
  • using automated tests and continuous integration to assist global team optimise processes

Panel – Continuous Delivery

Evan Bottcher, Neal Ford and Martin Fowler from ThoughtWorks were on this panel.

From Agile Australia 2011
  • continuous integration – everyone in the team integrates with the mainline at least once a day
  • continuous delivery is taking the same approach as continuous integration and apply it to the last mile – decision to deploy should be business only with no technical barriers
  • continuous deployment is continually delivering on a regular basis – continuous delivery enables this if you want it
  • rare to find a company that is pulling in the same direction, so you need to automate in pockets and add manual checkpoints and then you can look for ways to automate them
  • risks – need to bring pain forward, which was the tenet of XP, not doing it is much more risky, there is pain and effort to setup, you need to look for a leverage point in your production systems to justify
  • if you do something rarely you don’t get practice, by doing it more often you improve which actually results in a process that is more auditable and gives you more confidence
  • is a good approach to shorten feedback loops, also allows you to give confidence to the business on delivery timelines
  • packaged software makes continuous delivery hard, important to look at the automation of the configuration as well as automated tests, looking for fast feedback to give confidence in delivery
  • need to push software vendors to make things more deliverable (this was a rant by Martin Fowler, that I tend to agree with)
  • make database changes with the same fundamentals – break them down into small changes and combine schema changes with the data migration and string them together into a package, tools have got more specific like dbdeploy and Liquibase support this and Ruby on Rails just supports this out of the box
  • DBA’s are the final frontier because they like to fiddle with scripts, need to bring them in or deal with smaller changes
  • testers tied to manual processing – need to separate the low and high value testing work, fear that they will be replaced by a small shell script, will make their job vastly easier, need to get buy in by demonstration
  • difficulty is always the human element – testers are moved from the backend to the front end of the process, specification by example at the front now, need to look at incentives and make them common between developers and testers
  • key is a business decision of when to delay so you can deal with business change, training, etc…
  • people are now used to the fact that websites or apps on their phones are continuously changing
  • gives the option to deploy to different types of users when they need it
  • Go was built with continuous delivery in mind, version control systems are critical because everything needs to be in there, automated testing tools are also critical, continuous integration servers can help if they have an extra build pipelines
  • Puppet and Chef both allow you to script your environments
  • need to place people in teams who believe things are possible

Other Stuff

At the end of a very long day, it was good to network with attendees at the ThoughtWorks open office.

Also, I have to send congratulations to my colleague Adrian Smith from Ennova on his talk Agile for Startups which I hear was very well received (I have attended previous incarnations of this talk).