Release like there is no tomorrow
Software is never ready, never finished, never perfect and never done. The interesting thing with that statement is that the opposite is also always true, its always ready, its always finished, always as perfect as it can be and always done. This is the mindset of the generations of startups that are leading the way. They know that they do not have time to be perfect or to be “finished”. For them competition is fierce and speed has scientifically proven to be better for both the quality of the products as well as their security. It is supposed to always be release ready and always live with the users.
The economics of the printing press
The economics behind this thinking is that software basically comes at no production, material or distribution cost. A student in a dorm room can change the world as long as she has a computer and a internet connection. She can do that because software has no production, material or distribution cost, only a design cost. This means in theory that she can ship her product every day, every minute. If she wants to change the design you just click backspace and rewrite the code. The traditional model of progress that still often is ruling our organisations is that products/services lasts long and you rather do not want to change their design because that requires new production lines and new material. If you want to communicate a new product/service, you needed to be sure what to say and how it works before going to print. Once in print you reach the point of no return.
Therefore we needed to be sure, we needed to put in a lot of work up front, cause we could not stop the printing press and change without big costs as a consequence. With software that is not the case. This also means we can work and design our organisations and also our humans and our cultures in an completely different way. If changes of the design is cheap and there is no material costs to it, we change all the time, we release all the time.
Changed without changing at all
During the last few months here at Malmö Stad we have had an amazing group of students helping out to turn pdf forms into e-services using an e-service platform the city uses. The main case we wanted to make was that the workflow of the citizens today needing to download → print → fill out → sign → and scan in or send in could be done in a different way. We wanted to use the built-in online signing tool and let users be able to fill it out online. I do not want to cover all the benefits we see in filling it out online but rather just focus on what we did. We knew changing these pdf would be somewhat scary, change always is. So we choose a way to change as little as possible and just copy and paste the information fields. By doing so we had hoped for a easy roll out, we had not changed “anything”, just giving the citizens the option to fill out and sign the exact same pdf as of today. We still let the pdf be there for those who still wanted to fill out by hand.
However we quickly learned that this was not the way either. We thought we had not changed anything. We thought this would be a minor incremental release that a lot of the departments would accept, after all we did not change the information they were collecting or their workflow. There was no introduction of new systems or changes in how they needed to work. We also learned that a lot of departments had “started” a lot of their processes to redo their e-services. They wanted to restructure the pdf, change the language, split up the service, integrate the usage of the platform for the administrators etc. All with the right mindset of that there is major efficiency gains for the city, done right with e-services directly connected to internal systems, a lot of gains could be made. But those are years away with the established procedures the city has today. Here was a shortcut at least making it easier for the citizens.
Utopia can not be reached
Our learning is that working in very complex organisations with complex problems we need to break it down. Even we released to late, we also did the mistake of wanting to fix all departments PDFs at once and not releasing fast enough.
Innovation is not done in one big release, but by everyday taking incremental steps towards a better whole. We all dream of that perfect solution, the one that solves it all. But we will never get there, it is a utopia. When we are reaching what we set out to do the challenges have already changed and we want to do more. To not be paralysed in this we need to release as often as possible, even incremental steps is important. As long as they are backed up by data of usage they can stay. The worst thing is releasing it all att once after years of hard work, and seeing the data for usage crumble. Then you would be happy you released early, even if you felt embarrassed at that point you at least learned that it was not the right thing to do at this point.
In the end of the day we saw it as that the only thing we tried to change was to make it easy for the user without a printer. Removing the printer in the user flow might seem trivial but start calculating the time it takes for users at scale to print and scan back in. If you can easen that flow for a lot of users do it. The qualitative full-blown utopia e-services can still be released and this move did not destroy any road-map of getting there. If you can release today you should. If you do not feel embarrassed you released to late.
