After a month of blogging inactivity, I'm clearing the brain backlog, beginning with the big question: are we doing good work?

Every few months, I have an existential moment where I wonder whether what we are doing lives up to the rhetoric of social problem-solving and improves lives. I had one of those moments over the Christmas Holidays.

On a trip to Port Arthur, Tasmania I visited a 'solution' to a complex social problem: criminal recidivism. The separate prison, build in 1849, relied on psychological rather than physical punishment. Convicts were stripped of their name & possessions, sealed off from other prisoners 23 hours a day, and prohibited from making noise. The hunch was that social isolation and sensory deprivation would prompt reflection and rehabilitation. The reality was that social isolation and sensory deprivation prompted above-average mental illness rates. While Port Arthur's separate prison was only the third jurisdiction in the world to implement the model, they would not be the last. The model still exists today, in the form of high-security supermax units.

The separate prison model was a big idea. And, like the social ideas we co-create and prototype, started small. There were multiple iterations. Yet the iterated models yielded similarly negative effects. Negative effects for the end user group weren't enough to stop the idea from spreading. One might hazard to guess that the negative effects for the end user group were offset by the positive effects for other groups of people - e.g. prison officials.

That got me thinking about how we identify & balance the needs and wants of all the people implicated in social change.

We tend to talk about our projects in terms of two groups of people: the users and the system. Often, our starting point is what the system wants. In the case of Family by Family - the network we are currently prototyping in South Australia - the system wanted fewer families engaging and re-engaging with crisis services. In all of our projects, we find that success for the system is rarely success for users. Families want less stress, more fun times as a family, kids better behaved, and more time for self. Our methodology starts with user needs and wants, and tries to construct new kinds of services & systems around that.

One of the complexities of Family by Family, though, is that we've created two user groups. There are 'sharing families': families who have gotten through tough times and are willing to share their stories and strategies. And there are 'seeking families': families experiencing tough times and who want something to change. The theory of change we've been testing in our prototype is that by linking sharing and seeking families, we can shift mindsets, know-how, and ultimately, behaviour. Sharing families are conceptualised as the vehicle for change, and resourced as such through whole-family training camps, weekly coaching, peer-to-peer learning, and a cash grant.

Ten weeks of putting this theory of change into practice has led to a few surprises.

One. Linking sharing and seeking families has proved surprisingly easy. Families are choosing each other, using their own pictures and words, and hitting it off. They are building rapport and trust quickly, and swapping life histories and experiences without any prompting. If anything, our presence negatively affects the dynamic and turns a flowing, informal conversation into more nervous banter.

Two. Not every sharing family seems able to shift behaviour. While all of the sharing families have been able to shift mindsets and know-how , not all of our sharing families have had success enabling new behavioural patterns. It's hard, and quite new work.

We're asking sharing families to be more intentional and intensive than a friend, but less formal and focussed than a professional.That's proved to be an uncomfortable role - one that we've only just figured out how to talk about and train for.

Families who have pushed past the discomfort, used some of the tools and frameworks we've iterated, and touch-based with their link-up families several times a week, have seen some early indicators of behavioural change - e.g. a family starting a new 'family time' routine; a mum going back to school; a child making friends and behaving differently; a family engaging in community initiatives. But, as we know from the literature, behavioural change is rarely instant or consistent. It happens at it's own pace - and not always in the confines of a 10-week prototype! If nothing else, our sharing families seem able to jumpstart the process of change. What we do to support change after the initial jumpstart certainly requires ongoing testing and iteration.

Three. While not every sharing family can shift behaviour, all of our sharing families report pretty substative shifts in their own behaviour. Yesterday, I tried out a new way to capture & document the changes sharing and seeking families are seeing. Each family member, including kids, received a pack of cards and had to sort the statements into three piles: no change, some change, heaps of change. The kids of one of our sharing families highlighted some major changes in their own household - in what their parents said and how they acted, in their own sense of hope and value, and in the quality of their own family time. So all of the resource we are putting into sharing families seems to  be paying dividends - but more significantly for one user group than another.

It's time to recalibate the balance between user groups. We put a lot of resource into sharing families thinking they would transfer their new insights & skills to their link-up families. While that's happened in some cases - it's not been across the board.

So, this week we're trying something new. We'll putting more of our resource into sharing-seeking family pairs through joint coaching sessions. We're also holding a fun family day - with loads of developmental experiences - specifically for pairs of families. Our revised theory of change is that: by linking sharing and seeking families, and holding open a new space for both tough conversations and fun experiences, we can shift mindsets, know-how, and behaviours. Because the space is so different and uncomfortable, we need to take a more direct role in maintaining its boundaries.

If we can do that and find a way to balance investment in both user groups, then, good work might be near.