Leadership
People are leaders through status, or through their actions. At Haplo, we built a team of leaders who lead by their actions.
Leaders with status act through authority. When I look at the world, I see that giving people authority is often a very bad idea.
We believed that good leaders will lead through their actions, by setting an example and creating the conditions for others to succeed. I use “good” in both senses of the word. A good leader is effective, but they’re also kind and have an entirely positive impact on their colleagues.
As founders, we had authority through our status of starting the company, our legal obligations as Directors of the company, and our interests as sole shareholders. But we tried to limit our use of our status to setting the values of the company, enforcing them, and ensuring good governance.
These self-imposed limits gave our colleagues opportunities to lead, because there was little difference in what the founders and the rest of the team could do in terms of leadership. Everyone had to lead by their actions.
This strategy was surprisingly effective. Because we had chosen our colleagues carefully, our team was made up of ambitious and capable people. It was fairly typical for a new colleague, within a year of graduating, to be leading the development of one of our products, working with clients, and making big changes to the way we worked.
Every decision you make about your company is interrelated. Expecting colleagues to lead, without the authority of formal positions and responsibility, can only work in a respectful workplace where everyone is valued and supported. And perhaps you can only truly respect and support someone if you give them the freedom to lead?
The nature of leadership
There are many aspects to leadership. While an emerging leader may only do a few of them, an established leader will need to do most, if not all.
Strategy and values
An organisation exists to achieve a goal. A leader must set a strategy and values which provide a framework for deciding what to do, and how it’s done.
This might be quite vague. At Haplo, our goal was to create some software that helped worthwhile organisations do things of value to society, and an exemplar workplace.
Our strategy was to ensure that our software was truly beneficial to every user in every role at our customers, fund the company entirely with a predictable revenue stream from a software-as-a-service product, and build a high-functioning team by embracing the diversity of the society we worked in.
Our values were to do it in the right way, so our impact was entirely positive for everyone involved, and we helped as many people as possible along the way.
To reinforce the strategy and values amongst our team, on a day to day basis:
- We described Haplo as a values-led company, and made it clear these values were more important to us than anything else.
- We were open that these values were our personal values, and that we believed your work had to embody your personal values.
- We used formal communications to make sure everyone knew what we wanted to achieve.
- Whenever there was an opportunity, such as over lunch, we talked about our values and strategy, and why they were important to us, to reinforce the message informally.
Evidence-based decision making
The wider strategy and values are only a framework to make the right choices. The actual work is a series of small decisions that move towards the overall goals. A leader must gather evidence and understanding, and make decisions that are likely to succeed.
This was our starting point for enabling our colleagues to lead. As much as possible was delegated to those actually doing the work. Our role as founders was to create the culture that would guide people to make the right choices, and creating the processes which would identify and correct mistakes as quickly as possible.
The latter was important. Individuals will inevitably make mistakes, and if they’re not, it’s unlikely progress is being made. But a team as a whole shouldn’t make many mistakes, because an open culture enables everyone to see what’s going on, and light-weight processes ensure that everything is properly considered and reviewed.
On a day-to-day basis:
- We used a semi-formal process for managing change, where each change was adopted if an experiment with pre-defined success criteria showed that it worked.
- Anyone could propose a change, even (and especially) if they’d only been with the company a few weeks.
- Our processes included reviews before and after the work was performed.
Communication to build consensus
A good team is unified in goals and actions. A leader should build this consensus by communicating strategy, values and decisions.
It’s not enough to state the conclusions and impose your will. The reasoning behind them is important. Can you really believe something is right if you don’t understand why the decision has been made? And will you be able to do your work well if you don’t believe in it?
Leaders at Haplo communicated in several different ways, depending on the scale of the decision and what it affected:
- Decisions about projects would be through small group discussions, or written word in emails, chat and documents.
- The weekly developer meeting would introduce new tools and change the way we worked.
- Company meetings explained the bigger decisions and strategy.
- Conversations over lunch would reinforce values and discuss why we cared about doing things the way we did.
- We were always completely open about the commercial realities of the business, and how this influenced our decisions.
Embrace uncertainty
So much about business, the workplace, and the craft of writing software is uncertain. It is largely impossible to predict anything to any degree of accuracy, especially when it involves the decisions of people outside your organisation.
A leader must embrace this, and be comfortable with not being completely in control. Uncertainties must be communicated well and in advance, so that changes to the plan aren’t a surprise to anyone.
- We weren’t afraid to say we didn’t know something. (This is also important to creating a culture of learning.)
- We used strategies to make decisions even though our evidence was uncertain, for example,
- gaming things out, making decisions as “if this then X, otherwise Y”.
- taking a probabilistic risk mitigation approach, where we plan for the most likely outcomes and make sure that we know how we’ll cope if it doesn’t turn out that way.
- We turned uncertainty into an opportunity for emerging leaders, as it identified areas where they could take responsibility and lead the entire process from discovery to implementation.
Removal of obstacles
One of our most important roles at Haplo was to remove obstacles which prevented our colleagues from doing their best work. A leader’s role is to make others more effective, and the easiest way to do this is by eliminating things which prevent them from being effective. The effort we put into this was repaid many times over.
There are many potential obstacles, so there was lots to do, including:
- providing a quiet and comfortable office environment where our colleagues could think clearly without interruption,
- making sensible decisions and explaining them properly,
- ensuring they had the support and information they needed to do the job,
- thinking ahead so the resources they needed were available,
- dealing with external aggravations,
- and giving them the power to remove obstacles themselves.
Let go, so others can lead
If you’ve got everything else right, you can step back a little to let your colleagues become leaders themselves, because you know you can trust them to:
- ask for help when they need it,
- make decisions in a rational way,
- uphold the company’s values in everything they do,
- be bold and get things done, because they’re confident that making a mistake will be caught before it’s a problem.
We rarely asked anyone to lead. We just left space for them to do so, and encouraged them when they did. Our techniques included:
- letting colleagues take responsibility for solving problems when we didn’t know the answer,
- trusting them to communicate directly with clients, without needing to CC us into everything,
- asking them to do small things to build confidence, such as running a meeting,
- referring questions to colleagues who’d taken on a role, even if we knew the answer,
- and needing someone to help us, because were were a bit over-worked.
How this can go wrong
No organisation does everything perfectly, and Haplo was no exception. While we always upheld our values and maintained our standards, there were times when we didn’t execute as well as we’d have liked.
Pressure: This approach to leadership had the potential to put considerable pressure on our colleagues. Most of the time, it was a positive kind of pressure, as our colleagues could influence it and actually do something about it. However, when we were affected by external circumstances we weren’t able to control, this spilled over into short periods of stress for our team.
Supporting colleagues: Two way communication and support is essential. Having high expectations of a colleague’s performance and ability to get things done independently means that sometimes they won’t tell you when they need help. This is especially difficult where you’re working at a distance. There were a couple instances where I wish I had been much more proactive about digging into exactly what was going on and providing the support needed.
Scaling without hierarchy: We combined this leadership approach with an extremely flat hierarchy, because we thought it would make it easier for leaders to emerge naturally. This combination was hard to scale. When we were around 10 people, it worked extremely well. At 15, it was more difficult, and we started to put more effort into tools and process for communications. At 20 people, we reached the limit, probably because of the lack of hierarchy in the company more than the leadership approach. Our acquirer immediately put in place a small amount of hierarchy without interfering with the leaders, and things became much easier.