Coach, lean advocate, and author Jeff Gothelf talked about scaling Lean principles at this year’s London #mtpcon. Lean methodologies work well for a single team observes Gothelf: “A lean startup reduces risk by regular and continuous experimentation. When you take the concept of lean and combine it with agile you start to build a practice of continuous learning, and at the core of continuous learning is experimentation and humility. For lean and agile thinking to work we need to have a humble mindset… we work in software, which is complex and unpredictable, and it requires the organization from the top down to embrace a position of humility.”
Experiment Or Die
For start-ups it’s “experiment or die”, says Gothelf. You know your choices, their consequences, who you have and don’t have to do the work. “Most importantly you know how much runway you have and you know that if you make the wrong decisions a couple of times in a row you’re out of business.”
In bigger companies this all changes. With a larger scale comes more responsibility, access to more money, larger teams and a seemingly endless runway. There’s the ability to keep spinning out products and services and extending their lives without any humility or experimentation and learning. “And yet, with my clients it’s all agile and lean – everyone wants to be like a start-up,” Gothelf says. But “it turns out when you take the concept of humility, of continuous learning, to larger organizations it starts to break.” This happens for any number of reasons – business case, short-term business needs, process, approvals, silos, concerns about the impact on the brand, deadlines, targets, and so on.
A Framework For Big Companies
So how can big companies move past these hurdles? Gothelf suggests looking at lean at three different levels in a large organization – project, program, and portfolio – and says there are specific challenges to scaling lean at each level.
- At a Project level: most teams are incentivised to get product out the door as fast as possible. They struggle with making product discovery and delivery work together, and they have no autonomy to make decisions. They have an incremental mindset, and the perceived risk to brand and current is always front and centre of their thinking.
- At a Program level: here the problems get a little bigger. How do you coordinate teams and manage their knowledge? What about offshoring and distributed teams? It becomes even more difficult to build a continuous conversation about what is being learned and how to change course.
- At a Portfolio level: it becomes even more complicated. You start looking at discipline and business unit silos, IT as a service provider, short term business value, annual planning cycles, incentive structures which are focused on delivery, and governance.
There is no shortage of frameworks available for lean startup at scale, says Gothelf. “As a place to start they’re fine. But these frameworks are not end points. In order to evolve them into something that makes sense where you work you should start to think about principles, not the processes you want to scale. In other words, how do you want to work? Not what are the rituals you want to practise every day.”
Four Principles That Scale
Gothelf sets out four principles that he believes can help product managers apply lean at scale.
1 – Customer Value And Business Value Are The Same
If you build a product that makes your customers successful, that respects their time, is easy to use and solves their problems, then they will reward your business with loyalty and longevity, says Gothelf. And a tactic to achieve this, he adds, is to manage to OKRs. OKRs (objectives and key results) are outcomes, so they are measures of customer behaviour. The objectives are qualitative, inspirational, time-bound and actionable by the team. The key results are quantifiable and cascade down throughout the organization.
For example, a portfolio level objective might be to become the dominant platform for car dealer inventory management by Q4 2016. Key measurable results are as follows: 50% of all independent dealers are customers, six out of 10 car manufacturers mandate the platform for their dealerships and 40% of car auction houses use the inventory API. These OKRs are all customer measures of success, says Gothelf, and they can be used to reduce planning horizons. This means that decisions can be made based on evidence from a short time frame and also based on customer success metrics.
2 – Value Learning Over Delivery
One tactic to ensure the success of this principle is to build momentum with pilot teams, says Gothelf. A pilot team’s goal is to learn not necessarily to ship product or make money. The team’s velocity naturally will be slower, he says, but they are working out how to operate within the confines of the organization. Give the pilot team sandboxes, both organizational and digital, says Gothelf, and encourage creative experimentation.
3 – Radical Transparency
People get scared whenever there is change in an organization, so the more transparent you are the less fear there is. Gothelf suggests that rituals are an effective way to introduce transparency. “Get people to practise things like stand-ups, demo days, progress reports, anything that exposes the work you’re doing. People don’t think they’re learning anything, and it’s this transparency that gets you to a level of trust and a continuous learning mindset.”
Access to customers also brings transparency, comments Gothelf, as customers are the ultimate test of whether a product is valid or not. “The more we can give product teams exposure hours to customers, the more transparent we are about why we make certain product decisions.” Access to data is also fundamental to transparency, he adds, “don’t make analytics difficult to get to”.
4 – Humility In All Things
Just because we’ve always done things one way doesn’t mean that it should continue, says Gothelf, “approach your work with strong opinions but be willing to change your mind in the face of overwhelming evidence”. Product discovery is a great way to put humility into your work, he adds, it regularly challenges your assumptions and makes you think about what you’re trying to learn.
As we scale continuous learning we have to start looking at staffing with a new lens, Gothelf adds. There may be jobs that we carry with us simply because they’ve always been with us. He says: “As you build autonomous, empowered, self-organizing teams there may be some jobs that may need to be transitioned into new roles. Are they roles we truly need with continuous learning teams or do they add process for process’ sake?”
He also advises that organizations look at whether they provide a tech stack that is modern enough to allow teams to build continuous learning into their work.
Carve out a Safe Space
It’s a top-down effort, summarises Gothelf. “You already want to work this way. We need executive support. We need humility at the executive layer to say we’re running a software-based business and it’s different. If you’re not that executive, then you need to find them. Seek out the executive with the clout to carve out a safe space for you and you teams to try out continuous learning, lean startup and product discovery.”