How Self-Management Boosted The Performance Of A British Aerospace Manufacturer
On a rainy day in February we find ourselves in Poole, a town in the county of Dorset, on the south coast of England. We are here to meet with Julian Wilson and Andrew Holm, the owners of Matt Black Systems. The UK-based company was founded in 1971 (by Julian's father) and manufactures parts for the aerospace and defense industries.
The company had low productivity and dangerously poor delivery performance when Julian inherited the company and surely would have failed if it were not operating in the slow to react defense market.
In search of a solution to their crisis, they started to experiment with their operating model and eventually transformed into a unique self-management model, resulting in (among others) a 300% productivity boost.
Julian: "When I inherited the company it was in a terrible shape. Our costs were painfully high and the quality of our products was unknown. Only 17% of our orders were delivered on time and there was a strong reliance on paid overtime. One of our clients even told me once that we were in their top 10 of worst suppliers! I knew it was time to change. So, I brought Andrew on board, as he had extensive experience in bigger corporations."
Establish clear indicators before experimentation
Andrew, a former commercial director within a large multinational group, joined the management team and soon they both realized that the poor performance was most likely linked to their organizational culture. In particular, they identified employee disengagement, lack of intrinsic motivation and control over the production process as systemic problems.
But before Julian and Andrew started their experimental journey they first established clear performance indicators. These indicators would help them to review objectively whether the organizational change they initiated was improving the current situation or not. Subsequently, they decided to run every experiment for at least six months in order to gather enough data to measure the organizational performance. They measured their performance against the following four indicators:
- Quality (customer rating)
- Delivery (on-time-and-in-full, or not)
- Profit (return on capital invested)
- Compliance (audits - both internal and external)
Julian: "This is not a ticking-the-box exercise- getting these things right is fundamental to the sustainability of our business. The measurements are based on relevant business data and form the core of what we do. The targets are relatively simple. For quality and delivery they must be 100%."
Micromanagement, Agile & Lean principles
In 2003, Julian and Andrew started their experimental journey in a rather conventional way with the hiring of a (highly expensive) MBA manager to command-and-control the 30 employees. After one year of unsatisfactory results (QDPC) they increased control and split the single operations manager role in into two posts. Each of them now managing 11 productive employees. But, despite the hiring of expensive managers they did not measure any satisfactory improvement in their monthly metrics.
In 2006, the owners decided to take their journey to the next level and split the company once again. The organization then consisted of 4 parts, each equipped with its own micromanager. This was broadly in line with the Lean approach of establishing “cells”. They hired a bunch of consultants in order to adopt the principles of Agile working and Lean manufacturing. For a long period of time they kept on experimenting in the same direction. Each time they tried something new, it worked for a little while. Results went up, but after some time all metrics went back again to the baseline. In Lean they refer to this as “the grass growing back”. Sadly, after 2 years of experimentation still no significant improvement was achieved and a considerable amount of money had been spent.
Julian: "I felt like shit, was overworked and was sometimes even sleeping in the office. Everything we did was counterproductive. It was the most soul destroying period of my life. We tried every conventional thing, blew £250,000 in the process but only learned one lesson: we didn’t know what we were doing. Questions popped up constantly: what do we do? Are we going to sell? Do we give it away?"
The last resort
As a last resort Julian decided to take one last look at all the data they had gathered. Then he found something surprising. He discovered that the entire backlog of work was always kept at a certain number of man-hours. This, despite the fact that every project had a different total of man-hours and implied productivity was being flexed to keep the backlog hours constant. meant that the fixed number of hours backlog represented. It didn't make sense to him.
Julian: "Soon it became clear to me that it was actually the backlog of work that was being managed by the employees. The salary of the employees was mainly reflected by their paid overtime, which in turn was dependent upon that very same backlog, so the backlog was in their best interests! I decided to take a closer look at the backlog data and discovered two distinct peaks. One in Easter and one in November. After some reflection, I realized that this would result in some extra money for the summer and Christmas holidays!"
Clearly there was no incentive for the employees to get rid of the backlog. In fact, it was the opposite! Managing the backlog provided them with precise control over their own salaries. It was a clever response to the structure Julian and Andrew had put in place: earning a constant bonus based on their overtime. Julian realized that there were two kinds of Management Teams at work at Matt Black Systems: the formal one operating from their offices and an informal one operating on the factory floor. Obviously, the informal one was winning big time. But what might happen if the interests of the two could be aligned?
Determined to explore this issue they decided to introduce a new experiment once again, having spent all their money on the previous conventional experiments, this time the experiment had to cost nothing. The goal was to see if the effect was real.
They decided not to offer paid overtime. Suddenly, the employees had lost this financial benefit, they lost their sweet spot.
Andrew: "When tackling the first issue of employees’ focus on paid overtime, we decided to remove overtime policy, while offering the employees an “outcome” bonus as compensation on top of their regular salary which was equivalent to the overtime pay they were previously getting.”
The experiment cost the business nothing, the employees were not worse off either. Even so, it took a while to gain the approval of the staff involved- for just a 6 month experiment. We agreed we would return to paid overtime if that was what they preferred after 6 months. There was no longer an incentive to manage the backlog, but there was an inherent incentive to increase productivity so that the employees could leave on time yet still receive their outcome bonus.
Over the course of the following 6 months, productivity jumped 20% and the backlog fell for the first time for years. Critically, this happened without a single instruction in favor of change.
Andrew: "Seizing the initiative, the second step was adding a bonus for delivery on time. This second experiment built on the first and maintained the momentum of change, these tactics – introduced in consultation with staff – dramatically improved productivity levels, but more importantly contributed to a gradual mindset shift, where operators were becoming increasingly in control of the value they were adding and consequently their income."
This second experiment used a bonus that was paid by the increased output as we worked off our backlog and saw our productivity climb further. Clearly, the informal management team were capable of introducing significant change and needed no detailed instruction; but this team were also capable of successfully blocking any imposed change.
Because we found this informal team so readily manipulable it was clearly not operating maliciously or perhaps even consciously. Of course, if the behavior isn’t conscious (perhaps simply an emergent property) this means you can’t instruct it directly, you have to create an environment that shapes behaviors, specifically the environment must be aligned with the needs of the organization and its customers.
After these successes became clear, Julian and Andrew were ready to take their experiment to the next phase; to increase the reliance on this informal management team and reduce the actual formal management team to achieve significant cost savings.
They decided to provide each cell with complete control over their own budget. Now, every cell became responsible and accountable for their own Profit & Loss and Balance Sheet. It was another step in increasing entrepreneurship.
As it emerged, automating this administrative part of the operation was relatively easy and produced instant savings.
8 self-managing cells of 3
Inspired by self-management through the works of Ricardo Semler at Semco and by the workings of nature, they decided to give this direction a try, pushing it to its inevitable conclusion.
For the first time, Julian and Andrew witnessed a significant improvement on the Quality, Delivery and Profit metrics. But they knew there was room for improvement and felt that the organization could do even better. One thing in particular was bothering them. They observed that the new recruits quickly adopted the ideas whereas the existing employees had major difficulties to move away from the old ideas.
They let go of their strong Lean and Agile focus, and formally allowed the employees to exercise much greater control and responsibility over the way they worked.
They divided the business once again, now in 8 self-managing cells, each consisting of 3 employees. So, instead of organizing the organization in silos, they now organized themselves in multidisciplinary cells featuring all general functions (HR, sales, manufacturing, etc.) inside each cell. The overall head count had dropped from 30 to 24, having lost the functional administrators.
The business was now firmly “on-the-up”.
13 self-managing cells of 1
In 2009, Julian and Andrew decided to make the final push to full decentralization. The owners divided the company once again, this time into cells consisting of simply one employee. Making each employee responsible for their own Profit & Loss and strategic decision making. Now all Matt Black Systems' employees were rewarded against their own personal performance.
Julian: "This was the big breakthrough! Suddenly we were able to gather the data of each individual, realizing that only 20% of our employees were responsible for 80% of the performance. The highly productive employees could now thrive without being held back by less productive staff (by choosing to only collaborate with like-minded people). The ones under-performing in the self-management environment eventually left the organization.
There were no managers left, as their functions had become obsolete and unnecessary, and now no under performers. Productivity quickly rose as the engaged teams integrated the work that the disengaged were previously doing and so increased their pay. Today, of the original 30 engineers only 12 are left. But now, the success of the employees, the success of the business as a whole and the success of their customers were all linked together."
Each employee is now completely in control of their own work. They are independent and fully responsible for all the aspects of their business: from R&D, product development, sales, manufacturing, delivering and finance to administration. They are not only fully responsible for improving their own performance, they also directly benefit from it (in pay, skills, self-confidence, self-esteem and security of employment). Effectively each employee operates as an entrepreneur, fully empowered over their own payroll and pay rises.
Andrew: "Eventually, the level of responsibility that each operator assumed for their work allowed them to develop relevant skills and take ownership of complete projects, from start to finish. Within the business they operate an internal marketplace, where all 12 are individual entrepreneurs, designing, manufacturing and marketing their own products. In this free market place they can distribute their work throughout the organization with multiple customer/supplier relationships, or subcontract it to external suppliers. They are all working to common KPIs and can trade between themselves to achieve the target performance. For example, if one operator wins a large contract, they can ‘negotiate a contract’ other colleagues to help them deliver the project for a fee (expressed in internal currency) that they agree. Equally, operators decide themselves when and where they work, as long as they’ve agreed with their colleagues how they are contributing to the overall performance target. Individual pay is determined as a reflection of their performance plus a percentage of the individual profit."
Performance against the four monthly metrics continued to improve. Within a year they were presented with an Exemplary Supplier award by the same customer who had previously rated them as in their top 10 WORST suppliers. From that point to the present day the performance on the metrics continue to improve.
There are only 12 engineers left, but the ones left thrived in their new environment: the cost reduced with 50%, the productivity increased by 300% and they enjoyed a significant jump in profits.
An ecosystem of startups
Julian and Andrew recognized that their organization is now successfully being self-managed. They decided to take a step back, and to barely visit the factory again. So what do Julian and Andrew do? They see themselves mostly as investors and owners of the business that acts as a supplier of everything the employees tell them they need. (i.e. the brand, the workplace, serious investments and their self-developed software).
As we've seen in other organizations we've visited, Julian and Andrew turned their organization into an ecosystem of start-ups and making themselves the venture capitalists of their own ecosystem.