Why the mantra “the strategy is delivery” is not enough
Today we published California’s statewide data strategy. (View the announcement for more details). But I’m not going to talk about the strategy here. Instead, I’m going to talk about why I wrote one and why I think strategy matters.
The UK’s Government Digital Service (GDS) championed this mantra: The strategy is delivery. This mantra comes up over and over again in the civic tech and data space.
Well, based on nearly a decade of working in government transformation, I don’t think it’s enough.
Delivery is necessary but not sufficient
One of my mantras is “show, don’t tell.” I like to demonstrate what’s possible as opposed to pleading for it (though to be perfectly honest, I’ll totally plead if I have to :-). At the same time, if you only focus on delivery, you eventually end up with a series of projects that will be orphaned then absorbed by the infinite complexity of a large organization.
That’s why you need a bit more. The concept of a strategy is overused and often trite. But if used wisely, it can lead to durable change and it starts with goals.
Goals help you answer “are we there yet?”
I find delivering a series of disconnected and ad hoc projects depressing. A well-written strategy serves not only as glue and alignment but also as a filter. If a project isn’t well-aligned with the strategy, you know to avoid or minimize it. (FYI: sometimes you’re stuck with projects you have to do — here are my tips on managing what I call Minimum Political Products (MPPs).
To some extent, a set of project selection criteria fills this function. But criteria don’t establish a roadmap or connective tissue between projects. That’s what strategy is for.
And projects aren’t enough. Bureaucracies need more intervention tools for long term change, aka “the g word”, Governance.
Balancing delivery with governance
When I was CDO of San Francisco, we had an ongoing conversation about the balance between delivery and governance. Governance is the mix of policy, institutional process, and standards that rule an organization. Large (and small) organizations rely on institutional processes and rules in order to process inputs from diverse cultures, conflicting priorities, and reconcile strategic approaches. And once they are in place, they are often sticky. Our operating model at DataSF was implement, learn, then codify. If you want to affect institutional change, you need to insert yourself into those institutional processes and make your approaches sticky. A great example is the budget request process in government. The process rarely changes, including how new budget requests are framed and defined.
By attaching your change to institutional processes or creating new ones, you also reduce dependence on individual champions and reduce the risk that the change evaporates when they (often inevitably) move on.
Our ratio in SF was about 75% delivery and services and 25% governance. I don’t know what my ratio in California will be, but I expect governance to be higher. The state is HUGE, like really, really big. And so I’ll be leveraging sticky processes to spread the data goodness.
Modeling transparency and accountability
As change agents, we should exemplify transparency and accountability in our work. In the data world, we encourage ongoing measurement and evaluation. A public, declared strategy is about modeling how we plan to measure and be held accountable for our work. It’s saying I’m taking a stand: here’s what I’m committing to. With the California strategy, I’m now on the hook for doing what I said I was going to do or explaining why I didn’t get there and what I learned.
Managing expectations for those new to government
I’ll close on one final thought. There is a lurking risk in messaging to would-be government change agents that delivery is the (only) strategy. You risk disappointment when inevitably they learn that government change is hard, takes time, and requires new skills, including coalition building, change management, influence, and capacity building.
Happy change making!