Double loop learning

I used to work at a digital agency where the director was concerned that a lot of the work taking place was reactive in response to a seemingly endless stream of problems with customer websites or projects that we had taken on.

I didn't really understand what I was seeing at the time but I remember feeling frustrated that the scope of things to do seemed very limited in the face of the problems we were having:

Offer overtime to developers to try and get through the work in fewer days.
Encourage employees to get better at estimating work so it would be easier to plan.
Tell developers to inflate estimates so small pieces of work used up more retainer hours and we would be obliged to do less work.
Tell account managers to inflate the inflated estimates further so there was even less work to do.
Tell account managers to plan the developers calendars into 20 minute chunks so the business knew that all the work they needed to do could be finished.

It doesn't take an expert to realise that this approach led to more problems than it solved. Unhappy customers frustrated at the poor value they were getting, limited growth on the accounts a constant need to manage expectations with the customers and a long list of reasons why things weren't progressing (often these were considered upsell opportunities, like suggesting if they pay for a Magento upgrade maybe the work will happen quicker).

At the time the problem seemed obvious to me and I'm sure a handful of people reading this have arrived at the same conclusion. There was simply a capacity problem, we needed to hire more people to get through all this work.

I had a sense that we were also making poor use of the time we had and that the calendar micromanagement was probably causing capacity issues as well but at the time I had a strong assumption that creating the needed capacity would put and end to this practice.

My perceptions of the problem and the solutions I saw are what I now understand as single loop thinking.

Taking a step back and thinking about some system conditions under which the business ran lets us ask some powerful questions.

The business ran a two step process which I'll outline below and highlight the assumptions behind the thinking.

Sell a new website or redesign to an SME
On completion of the project set up a retainer contract where the SME pays a fixed amount monthly for a number of hours work per month.

Assumption

Action

Outcome

Selling retainers is standard practice for an agency so must be a good way of running the business.

Sell retainers to customers who don't need them as the only way they can get us to do work for them.

Customers are frustrated that they have to commit to a long term spend to get a small piece of work done.

Having reliable retainer income is good for business and cash flow.

Selling and increasing retainers is an objective for account managers, they are rewarded with bonuses renumeration and promotions for selling these.

The business has committed to take on more work than it is able to do. There is a considerable management overhead to owing customers this time.

Retainers are a good value option for customers as they get access to website changes at a cheaper hourly rate.

Promote retainer sales to customers.

Customers have a monthly budget they must use and asked the business to do work that wasn't valuable to them to 'use-up' their retainer hours. This led to issues with their core offering, extra work for them managing the contract, and frustration that if they needed something bigger doing they couldn't group together hours from multiple months.

So what is double loop learning and where does it fit in?

There is a lot of preamble here but in essence the power of double loop learning is to shift the focus away from customers and individuals and instead look at the system and understand it's impact on the ability of the business to meet the customers demands.

Instead of creating committees to 'plan' work in chunks, why not look at the wider system and understand how it is creates conditions that shift our focus away from meeting value for the customers and instead on internal planning and scheduling?

By starting to think clearly about the assumptions we hold and how they lead to actions and outcomes we can focus our learning on challenging our assumptions and begin making changes to the systems we work in. This removes the need for us to be reactive and lets us focus on doing valuable work for our customers.