The Innovation Framework I Built By Accident
- May 6
- 3 min read
By Carin Aharon
Founder of ITAKA
I didn't plan to build an innovation system. I set out to survive.
I'm Carin Aharon, founder of ITAKA — a three-person web design and digital strategy studio for founders and personal brands. We run seven projects simultaneously and once delivered project ownership to a client in six days. We never called what we do a framework until now. The story of how it came together is less about methodology and more about mistakes that forced us to get honest about where human judgment ends and where tools begin.
It started late one night, reviewing a homepage positioning while the team slept. The work looked strong — polished, intelligent, premium. The kind that passes every rational test. But sitting with it in the quiet, something became clear: it didn't feel like who the client actually is. Structurally convincing. Not true.
That's one of the most dangerous things about AI-assisted work: it can produce something that looks right without actually being right. That gap is invisible to the tool. It's only visible to people who have spent real time with the client — who have heard their contradictions and learned what they're actually trying to say.
The next morning, I told the team we were restarting from scratch. Nobody pushed back. Nobody pointed to the hours already spent. That moment became the foundation of our most important operating principle: We do not protect work just because time has gone into it. Nothing is precious. If it isn't true, it isn't finished.
In practice, this requires a specific kind of leadership — one willing to make uncomfortable calls, and a team that trusts those calls. Without both, it collapses into the sunk-cost thinking that kills innovation at every organizational size.
The harder question, once you accept the restart, is how to catch the problem earlier. Our answer came from a client who said something I've thought about since: “You’re not just saving us time — because you save us time, you save us much, much money.”
Before that conversation, I believed our value was speed. After it, I understood our real value is preventing wrong moves before they happen — eliminating the cost of building the wrong thing entirely.
That insight changed how we validate everything. Before we build, we stress-test strategic direction in conversation. The question we return to is always: if this launched tomorrow, what would a cold stranger decide in eight seconds? That question has killed more bad ideas faster than any formal testing process I've used. It makes validation a conversation, not a checklist.
Which brings me to the boundary that changed everything. We once let AI carry too much weight in producing a strategy document. The output looked coherent. It was a disaster — it misread what actually mattered, lacked real depth, and damaged the deliverable. It reinforced a rule we now treat as non-negotiable: AI can support thinking. It cannot replace it.
Shortly after, a client asked us to use AI to write their entire website content with no involvement from them. I refused. I cannot build accurate content without the client's participation. That refusal was the moment the boundary became cultural: AI assists, humans participate, and leadership means being willing to say so — even when it costs you a project.
The advantage larger organizations have had was never talent. It was capacity. AI has closed that gap.
But it cannot give you judgment about which work is worth doing, which direction is true, and when to restart. That judgment is still human. That judgment is still leadership.
We didn't design this system. We discovered it, one uncomfortable morning at a time.
Connect With Carin
Comments