Products compete in a landscape that is constantly changing, and changing unpredictably.
Covid happens. Brexit happens. Your competitor changes direction unexpectedly. Log4stash happens and security is at risk. AI takes a huge leap forward with GPT-3. Google changes its algorithm. Consumer trends change. New technologies emerge. A key employee leaves. The job market changes. And so on.
There is also uncertainty such as… Will new feature X have an impact? Will it be stable?
Under these conditions, it doesn’t work to plan everything up front (like “waterfall” projects do). It’s better to plan just enough.
To navigate change and uncertainty, to seize opportunities and reduce risks, the best bet is short feedback loops. This is often called “iterative development”.
The Hacker Way [Facebook’s culture] is an approach to building that involved continuous improvement and iteration
Short feedback loops is the core of Agile.
How Columbus discovered the New World
Navigating uncertainty is not a linear process like most people are used to. Most people are used to a designer doing a mockup, then an engineer building the thing, then the thing gets launched to customers. That’s linear.
Short feedback loops work differently.
Development of the iPhone was more of a squiggle than a straight line. The iPhone wasn’t a matter of design it, then build it, then launch it. The same for the success of Amazon, Google and Facebook.
Here are three ways that Apple shortened their feedback loops when developed the iPhone.
3 iPhone examples
1. The iPhone manufacturing cycle
Apple iterated not just with their software but with their hardware.
Apple iterates the design throughout manufacturing. The product is built, it’s tested and reviewed, then the design team improves on it and it’s built all over again. These cycles take 4-6 weeks.
This is a very costly approach but it’s one of the reasons that Apple… build incredible market changing products. It’s the process that the iPod, the iPhone and the iPad went through.
Apple’s Product Development Process
Note that it says, “This is a very costly approach” (for hardware). Feedback loops are so valuable that it’s worth the cost.
2. Divergence
Apple prototyped two different approaches simultaneously. The front runner was an iPod-like clickwheel interface, and the underdog was a multi-touch interface. These were working prototypes, not just designs.
It was this “divergent” process that gave Apple the feedback to pursue the iPhone we know today (with web browsing, messaging, and apps), not just “an iPod that can make calls”.
3. The Apple culture
Developers and designers constantly sought feedback. Here’s Ken Kocienda, the developer of the iPhone’s virtual keyboard.
Since Richard Williamson’s office was right next to mine, I often poked my head around his door, called him into my office, and invited him to pick up the Wallaby [iPhone simulator] so he could try my latest ideas. He always gave specific feedback…
This kind of collaboration was common. The programmers and designers on the Purple project were in and out of each other’s offices all the time. We exchanged frequent feedback on our work…
Inside Apple’s Design Process by Ken Kocienda
Power loops
There is one particular type of feedback loop which is very powerful: experiments.
Our success at Amazon is a function of how many experiments we do per year, per month, per week, per day
Jeff Bezos
Netflix… used experimentation to generate sustained levels of growth seldom seen
Netflix: Lessons in Experimentation
To get good at experiments you also need to get good at making hypotheses.
It’s complicated
These feedback loops and experiments need to be happening at different levels simultaneously, from high-level strategy to low-level implementation. And the different levels need to feed into each other.
This the core of agile.
(NOTE: For a team to continually adapt, it’s important to minimise dependencies outside the team. These are sometimes called cross-functional teams, or multi-disciplinary teams).