All areas of the financial sector feel the pressure following the global credit crunch. Most customers, both in the public and private sector, look through their budgets for savings and as a consequence the insurance companies are pushed to offer lower premiums just to keep the big customers in their portfolio. Jacob Austad tells the story of an insurance company experiencing great difficulties in delivering services according to customer expectations, and how it tried to changed the status quo.

In the face of high levels of stress and only little employee trust in management, the company featured in this article had to achieve a substantial transformation in the way it worked.

The demand on the insurance broker department comes directly from the insurance broker and should only consist of two types:

  1. Request For Quotation (RFQ) related to possible new customers;
  2. Changes to existing customers’ current insurance coverage.

Management described the department burning platform like this:

  • The externally conducted insurance broker rating of the department performance has been declining for three consecutive years and is now the lowest score on all parameters for the top six national insurance companies;
  • Customer satisfaction survey reveals lowest score ever;
  • Employee satisfaction survey show low engagement, high stress, lack of trust in management;
  • The backlog of cases yet to be handled is growing;
  • There is no overview of the amount of on-going cases;
  • Amount of new RFQ has been declining;
  • Hit-rate for winning new customers is lowest ever and still declining;
  • Employee overtime has been growing – affecting costs – but no visible results on the backlog have been seen.

In terms of organisation, daily operations were divided into silos with each employee only responsible for selected customers (insurance broker houses) and no overlap or support from colleagues.

To follow the development of the department’s performance three overall measures were put in place:

  • Overall portfolio measured quarterly;
  • Adherence to lead time Service Level Agreements target – two weeks to respond to any question, case or RFQ (not to solve, only to respond);
  • Backlog of not started cases counted bi-weekly.

Senior management designed a better daily operational management system and tried to fix the current internal issues before focusing on increasing the overall portfolio. There was a notion that the surfaced problems were merely the result of things happening along the process and that the root causes were still hidden, even though employee performance and behaviour were seen as unacceptable.


Facts beat feelings, and in order to clearly define and document the problems and find the true root causes, management and employees were asked to study the demand put on the system. Four main channels were defined (e-mail, phone, mail and management). The demand per channel was studied, and the research revealed that:

  • 83% of all e-mails were Failure Demand:
  1. Customers asking for status
  2. Asking for more information or providing (unnecessary) additional information;
  •  96% of all phone calls were Failure Demand.
  • Mail primarily was wrong delivery or advertising brochures (3% of total demand).

Traditionally, management is seen to state both the problem(s) and the solution as well as ask an external consultant to go implement as ordered.

Especially in transactional environments, a lot can be learned from the old Deming quote:

Cause and Effect are often separated in time and location.

… and management need to learn to understand the general system complexity as well as the fact that employee behaviour is a result of the system design. Therefore any journey should start with the management and employees studying the current system in 4 dimensions:


The study will reveal whether the current system design is capable of solving the customers problems and whether there is alignment between demand placed at the system by the customers, the purpose of the system, what the processes were designed to deliver and how the measures and targets are used for planning, operations and follow-up, etc. All this information will be a valuable input to the forthcoming re-design.

During the study, management and employees started to pay interest in understanding the root causes to the failure demand and they were eager to solve the problems.

Knowing about the power of intervention as well as the need to ensure that the activities selected for implementation will be likely to solve the issues at hand, it should be obvious that the focus should be on building trust and competence in the employees and for management to actively support this.

A brainstorming workshop was conducted with the aim to define what “perfect looks like”. According to the team, perfect performance is achieved when:

  • There is no need for Service Level Agreements;
  • Failure Demand is eliminated;
  • Backlog is eliminated in six months;
  • Daily “planning” is done by employees for employees: full visibility, share the work, personal planning model (for those who need it);
  • Targets are eliminated and measures are plotted in a time-series-plot.

The idea was that delivering these elements would increase customer satisfaction and contribute to winning more quotes because of good service and fast response (for example, keeping promises and first-time-right) instead of sole price focus. As everyone would contribute and actively change the system, employee satisfaction should increase and the system should be able to adapt to changes in demand.

The team invited senior management to see the learning’s and to give their input as to the way forward, but they challenged the findings and would not accept the fact that the current targets were part of the problem. However, as the team explained the connection as well as its belief in the fact that it could turn the business around, it was given the chance to deliver.

It took some serious support from the CEO of the insurance company to help employees to see the glass half full rather than half empty.

A plan was established to deal with the findings from the diagnosis. Each employee signed up for the areas they had special interest in and from that three working groups were created. Each working group became responsible for designing the system in a way that solved the problems and to include other colleagues, also from other departments. There was only one rule: that all solutions should prove sustainability before any IT solutions could be developed. Of course this was seen as something of a set-back and was only accepted when the working groups saw that the solutions did not need IT to work (actually they worked better using a whiteboard or pen and paper).


Working directly with activities focused at improving the relationship with customers involved some counterintuitive thinking and a new systems design that was quite different from normal practice.

No need for SLAs to control the business “Perfect” was described as all customer contacts being handled and solved within two days. The working group recognised that the two weeks response target had become an excuse to reply to the customer after two weeks rather than solving the query as quickly as possible – the target stated in the SLA proved to have the opposite effect than it was intended to.

The SLAs were part of the contract with the insurance brokers, and as such part of the framework for how business should be conducted. In daily life and running business the standards for response time did not describe reality as brokers would call and mail with rush-orders in need of immediate handling. The study of demand revealed that rushorders (status inquiries) and changes to previously placed orders counted for approximately 30% of the failure demand. Instead of referring to the SLAs and upset the brokers, the idea was to design the system to be able to deal with actual demand and not a theoretical contract.

By studying department capacity over time it became apparent that there was enough capacity to handle all demand (the average size of the backlog was almost stable) – so why not do it immediately instead of waiting two weeks?

The work group developed a principle stating that demand coming in day 1 would be handled on day 2 and that, in case of extra high demand one day, colleagues handling RFQs would help handling demand related to changes to existing customers current insurance coverage.

Handling of RFQs was not a problem as these cases always included a deadline for sending the offer and on a daily basis employees would decide if it was a RFQ worth giving an offer and if the capacity was available within the period. The only change was that the employees now had the authority to make the decisions.

Using this approach there was no need for hiding behind the SLAs as the lead time was now less than two days, which was never seen before in the financial sector. As a consequence of this, failure demand dramatically decreased, by 75% and still improving.

Customers (brokers) began recognising this change within two to three weeks and asked “What have you done? Did you hire more people?”

Failure demand eliminated. This focused on actions common to all workgroups and on measures to follow the development and improvement in systems performance.

The table below shows part of the demand analysis as well as some of the actions related to eliminate failure demand.

The study also revealed a few “grey zone inquiries” which might be needed or were unavoidable but could be minimised with a focused approach. The team had a firm belief that the elimination of failure demand could be achieved through implementing the actions in the table and as a result time to deliver better customer service would be freed up.

To follow the change in type and frequency of demand the team developed a simple Tally-chart method to register the amount of failure demand at first-point-of-contact and show it on a daily/weekly plotted graph.

Once a week the daily briefings also included reporting on action status and the team decided that support activities based on the actions impact on process performance.

Backlog eliminated in six months. The backlog was counted and isolated and a plan was created to eliminate it. The principle used was that customers placed in the backlog were already dissatisfied and that handling the cases after four or eight weeks would not make a big difference.

The team had an overview of the demand to be handled each day and was therefore able to decide how many cases from the backlog that could be solved. Furthermore, special efforts could be decided, all planned by the employees (only when special decisions were needed was management involved).

After four months and a half, the backlog was eliminated and as a consequence capacity was freed up and for a period used for training and competence building. Later, capacity will be used to provide better service for existing customers and thereby win more quotes.


This work group was focused on establishing the internal routines primarily surrounding daily planning and how to create the visibility which will ensure problems to be surfaced and act as triggers for corrective and preventive actions.

The team began with a workshop to design the “rules of engagement” or a common agreement on what to expect from each other. In reality, the aim was to have open and honest discussion.

In addition to several attempts to design the boards to include the relevant information, the team also devised a process aimed to help the employees to plan what they should do individually week after week. This process became a part of individual coaching, both employee to employee and manager to employee. The idea behind this was that employees expressing the need to have high visibility would be able to help each other as well as creating the legitimacy to ask for support.

As a result relationships were strengthened and general engagement increased.


This group worked on how to design the department structure supporting competence building. A personal development plan was created for all employees and as a result a new internal structure was implemented leading to everyone in the department being capable of handling all types of (value) demand and not only parts.

Furthermore, targets were eliminated and all measures were designed in a way that made it possible to show the progress and status in a timeseries- plot. This allowed for long-term sustainability as the system became capable of handling changes in demand (type, amount and frequency) through a substantial knowledge and use of Deming’s PDCA, because every variation created the desire to find the root cause and eliminate it by adjusting the system.

The work group put a big effort in involving management, directly through invitations to participate in the workshops to observe the daily routines and indirectly through a monthly report showing the new measures and how the work actually improved performance beyond the former targets.


The trick in transactional lean is to start by understanding the problem to solve, and by always listening to the employees and having them and the managers study demand. This also includes not seeing tools as the solution. Too many lean initiatives are implemented just as a list of tools used in a given sequence. John Seddon calls people using this approach tool-heads: this often gives lean a bad name and these projects are never sustainable because they follow a consultant’s approach.

The benefits will only impact the “L” in the business P&L, because they don’t recognise an important teaching from Deming: cause and effect are often separated in time and location.