How we spin the Feedback Loop

Yes, I promised that this week, I’d share tips on raising capital in Silicon Valley. But I’ve postponed that post.


This blog is to firm up in real time the lessons I’m learning, which could help other entrepreneurs facing similar challenges. I’ve thought a lot about the concept of feedback loops and how start-ups manage the evolution process.


I’ll share how we manage the evolution of our product, strategy and processes – areas I think in which we are improving. I’ll write more about Silicon Valley later, but if you’re keen to discover how I pitch, check out this previous post on pitching techniques.


I first heard of the concept of a ‘closed loop management system’ in a Harvard Business Review article I read last summer. It’s a process that big companies use which involves first developing a strategy, then translating it into measurable targets, planning operations, implementation, monitoring and learning before testing and adapting the strategy so that the team can develop the next iteration of the strategy.


It’s a solid process and HBR suggests that companies go through it annually. That would work if your company was the Commonwealth Bank; not so much if you are a start-up with limited time, resources and opportunity to win a new market before the competition.


The trick for a start-up is to get into a manageable heartbeat, enabling the company to iterate its strategy fast without being scattered, and learning from each evolution. A start-up’s management cycle requires the same steps and rigour as a big company but needs to spin faster. The number of spins your start-up receives depends on two finite resources – money and the enthusiasm of the team.


We’ve been lucky enough to have more spins than most since, even though Posse isn’t big yet, our investors and team see us making rapid progress and are impressed by our approach to evolution.


I introduced the concept of the ‘Closed Loop Management System’ at our team offsite in March and since then we’ve spun through a strategy iteration cycle each quarter. We start the quarter with a full team offsite. We all go to a beach house for a few days and discuss what we’ve learnt from the past few months, what concerns people have, and any ideas on improving our strategy.


We look at our competitors, what ideas they have that work, and what distinguishes our product from theirs. At the end of the time away, we don’t have a strategy, just a lot of feedback, and ideas that float around for the next couple of weeks.


After more thought and discussion, that’s when the strategy comes together. Within each quarter, we have two six-weekly development sprints, although we release new code to production daily.


Now as I write it down, the system sounds more formal than it feels. The benefits are that there’s a clear pathway from strategy to plan to execution, everyone agrees and can focus on this plan for a set period knowing that the plan may change in the next cycle.


They also recognise that success will ultimately come from learning through the process of the cycles and not necessarily having perfect strategy every time.


It’s the only way to sustain team enthusiasm when people are working hard, often late at night and on weekends to release a product they’ll probably have to redo repeatedly until it’s right.


Followers of Lean Start-up methodology will know this principle as the ‘Feedback Loop’. It makes sense in principle but is very hard to execute well. As you’ll know from reading my blogs, I’m no expert and I’m still learning all this stuff as I go, but here are a few things I’ve figured out that are important to making the feedback loop spin effectively.


1. Listen to everyone


As a founder, you’re naturally optimistic and you believe you’re on track to build a hugely successful company. If you didn’t have a high level of self-belief then you’d never be able to convince others to invest in your idea or join your team.


Balance this optimism with a passion to constantly test your plans and product with users. Not just the initial model hypothesises but every single thing you do should be thoroughly tested before, during and after it’s implemented.


Trust me; this is a hard practice to sustain. It’s easy to fall in love with the latest version of your strategy and go into a vortex of hardcore implementation without talking to anyone, only to find that you’re going in the wrong direction.


I find that, although painful, every time I set aside an hour or two to meet with users I discover an inspiring insight, which leads to better product strategy.


2. Take time out to think


After I’ve listened to as many people as I can, including users, team members, board directors and my friends, I’ll sneak away and spend time by myself to think.


It doesn’t have to be a beach holiday (though this is nice); often it’s an overseas work trip or a series of long walks. I’m not sure how this works, but it feels that everyone’s ideas float around in my subconscious for a while and the change of scenery triggers them into combining in a logical way.


Through this, I can build a strategy. Once I have the strategy worked out, I take it back out to the team and test the ideas. If the strategy is good, everyone gets excited; if it’s bad, everyone argues. I know it needs more work.


3. Face big problems head on


One of my mentors in music gave me some great business advice once. ‘If you think there’s a problem – there’s a problem.’ As the entrepreneur, you live and breathe the company, you can sense when something isn’t working.


At the beginning of this year, I sensed that the retail recommendation app we were building wouldn’t get traction with customers and so I immediately started a process to redesign the strategy.


Right now, our biggest problem is that our user experience design sucks. People don’t understand what the site is about or why they should use it and when they do join, they find it hard to navigate. It’s a major problem that we’re confronting head on by completely redesigning the user experience, and will launch at the end of the year.


Many start-ups that I’ve seen fail, knew they had the problem that killed them long before they died. I think you have to tackle every major problem head on in the next spin of your management loop.


4. Move fast


It’s obvious that with limited time and resources you have a certain number of spins to get the product – market fit right.


We still don’t have ours 100% there, although our user engagement metrics suggest we’re coming much closer. The faster you can create good new strategy, implement and test, the better your chances are of success.


5. Be disciplined


It’s so easy to want to react to every small problem, fix every bug and make lots of incremental improvements. Earlier this year we were sucked into addressing all of these as they came up. Weeks passed by without any major advances in the implementation of a strategy, so we’ve learnt as a team to focus on getting out the big important things that we can learn from rather than making sure the site is perfect.


Right now, we’re at the start of another loop. The strategy came together during my overseas trip a couple of weeks ago, it was formed after a month of feedback from a lot of people and the team are all excited so I know it must be good! It’ll be a major upgrade to the user experience, site interface and game design and everyone is working long hours to get it out before the new year.


As always, these are just my ideas and are from the perspective of someone who is learning on the job and is struggling with these concepts every day. I’m no expert – I’m just sharing what’s been on my mind.


(I was prompted to write this after reading an article in Forbes about Foursquare this week. The journalist thinks that ‘Foursquare’s bubble is about to burst’ because they’ve failed to spin the feedback wheel fast enough and adapt to the changing market and needs of their user base. It’s scary to think that a company this big could still fail because they didn’t evolve quickly enough but I guess we’ve seen this happen before with MySpace.)


Notify of
Inline Feedbacks
View all comments