The values of a brand platform usually lead a quiet life: they get written into a document, printed on a few posters, quoted in year-end speeches. "Transparency", "innovation", "people first". Nobody objects to them, because nobody feels bound by them.
In the INPS rebranding we tried to use that mechanism in reverse. One of the six values of the new identity platform — Continuous improvement — was, quite deliberately, a Trojan horse for a good cause. Behind an apparently obvious idea, "improving every day", we smuggled into the Institute's identity the iterative approach typical of people who design digital services. Nobody votes against "improving every day". But once inside the walls, the value produces very concrete consequences.
Why a value, and not a directive
A directive can be ignored, rewritten, or quietly left to expire along with the leadership that signed it. A value of identity is harder to dismantle: whoever denies it denies what the institution declares itself to be. For a large service provider like INPS, coding iteration into the identity meant a precise cultural shift: moving from compliance — "the procedure is online as the law requires" — to evidence-based management — "how do I know whether the service works for users? How do I improve it?".
This is the cultural bomb hidden inside the horse:
Quality stops being a requirement declared upfront and becomes a result to be demonstrated afterwards with evidence.
The question changes nature: from "did we publish the procedure as the law requires?" to "what data tells us where we did well, where the weak points are, and how we prioritise fixes?" The centre of gravity moves from self-reference to listening, from "we've always done it this way" to "what actually works for people". On why measurement matters so much in an institution that has no market to correct it, see Making Things That Work; here I'm interested in the trick used to install that approach.
The obligations that follow from the value
Continuous improvement, understood this way, has nothing to do with occasional cosmetic touch-ups: it means structuring an endless cycle of learning. Translated into practice, the value imposed six operational consequences:
Prototype before you ship. Anything involving interaction with users must be verified in a real context — real people, real content, real channels — while it is still a prototype. Prototyping is an act of design humility: accepting that no rule is good in the abstract, and that whatever you want to standardise must first prove it creates value in use.
Test with real people. Validation doesn't happen at a desk. Task success, completion times, recurring errors, drop-off points: direct contact with the people who use the services is what separates what is clear from what is merely well drawn.
Measure outcomes, not just outputs. Knowing a service is online isn't enough: you need to know whether it works. Completion rates, reduction in repeat contacts to the call centre, readability indices improving. The goal is producing better before producing more.
Read behavioural data. Navigation funnels, drop-offs, internal searches, heatmaps: signals that tell you what people are looking for, where they stop, what they can't find. A way of listening to citizens continuously and systematically.
Plan iterations. Every release opens a path rather than closing a project: an improvement backlog, a regular release cadence, retrospectives with traceable decisions. Quality comes from the consistency of improvement more than from the brilliance of the initial idea.
Make the rules public. Clear, shared acceptance criteria, with a definition of "done" that includes quality for users alongside formal compliance with the process. Transparency and accountability travel together.
Note the tone: these are obligations, not aspirations. That's the difference between a decorative value and an operational one. And one misunderstanding is worth clearing up: continuous improvement in this sense is a discipline that reduces risk — small steps, frequent checks, targeted corrections — rather than a licence to change your mind every week. It teaches an organization to distrust first-time perfection, which in practice almost always turns into paralysis.
If identity says who we are, continuous improvement establishes how we become, each day, a little more consistent with what we claim to be.
The other horse: Inclusivity
The same trick worked with a second value of the platform: Inclusivity. For an institution serving every citizen without exception — pensioners, workers, families, businesses, people with disabilities, young people, foreign residents — inclusivity is a structural necessity: if someone is left out, the system has failed. Here too, immediate and verifiable practical consequences followed from the identity principle:
- guidelines that ban acronyms in the names of citizen-facing benefits;
- a long programme of text simplification to cut legalese and technical jargon;
- for the first time, the systematic measurement of the readability index, so that content is accessible to people with lower levels of formal education too;
- a plain-writing guideline inspired by plain language principles, with practical, verifiable criteria for every editor.
A value that produces a measurable ban — no acronyms in the names of benefits — is worth more than ten manifestos.
The moral of the horse
The trick works for a precise reason: the brand platform is one of the few documents an organization approves at the highest level and then defends as its own public image. Whatever gets in there acquires a legitimacy no technical document will ever have. So if you want a practice to survive projects, reorganizations and changes of season, don't leave it in the manuals: put it in the identity. And the brand itself gains from it: it stops being a photograph and becomes a living system, correcting itself as it learns.