Most Organisations Are Not Yet Designed to Learn From Them 

For decades large organisations were promised: Digital transformation, customer-centricity, data-driven decisions and measurable ROI. Yet most enterprises still experience the same unsatisfying pattern:

​

• Projects launch.

• Technologies change.

• Consultancies and agencies rotate.

• The cycle repeats. 

And organisations still struggle to answer the simplest questions:

​​

• What actually improved?

• What did we learn?

• What should we do next?

​

Because most organisations were never designed to remember what actually happens inside their digital systems.

​

Digital programmes generate enormous activity, but very little institutional memory.

​

The enterprise never existed to support endless IT programmes.

​

IT exists to reduce friction and solve real problems for customers.

​​

​

But There’s a Structural Problem

Most delivery models were designed for a different era. Software vendors sell platforms.​ Consultancies implement them.​ Internal teams try to operate them.

​

And, responsibility fractures in the spaces between. Learning fractures there too. 

​

When outcomes fail to materialise, the reaction is all too familiar. Finger-pointing. 

​​

Consultancies blame adoption. The business blames IT. IT blames third parties. 

​

All have a point, but that’s not the point.

​

The underlying issue is systemic.

And the root cause is what we now refer to as institutional memory, or the absence of it.

​​​

​

The Importance of Memory

Institutional memory is the reason air travel has evolved from the riskiest form of travel to the safest. 

Industries like aviation and civil engineering - where our safety is paramount, solved the problem long ago.

They record incidents.

​

They record decisions.

They build on what they learn.

​

They rely on governance systems designed to accumulate knowledge over time.

​

For decades that level of operational discipline was only feasible in heavy industry and major public infrastructure - where the cost of failure is too high. 

​​

Today, it’s possible everywhere.

​​​

​

The Consequence of Forgetting 

When organisations forget their own operational history, every initiative becomes a fresh start.

Campaigns repeat the same mistakes.

Incidents recur.

Teams rebuild knowledge that had already once existed.

​

The technology advances. But the organisation’s memory resets - cyclically. 

​

AI placed into that environment does not necessarily solve any deep problems.

In many cases it simply detonates fragmentation.

​​

New Architectural Thinking Now Needed  

The next era of digital systems will not be defined by platforms alone. It will be defined by institutional memory.

Systems that:

• Observe what actually happens

• Record decisions and outcomes

• Connect signals across delivery, experimentation and operations

When systems remember, organisations can focus their energy on progress instead of stability. 

​​

Evidence of Conviction 

First Three Things was founded on a simple belief. 

Organisations cannot become learning systems if their tech partners do not operate that way themselves.

​

So we must lead by example.

​

That’s why our systems of governance are machine-legible. 

Decisions, incidents, releases and outcomes are recorded continuously and made available to both people and intelligent systems.

Because when governance becomes machine-legible, digital systems begin to improve at the pace they were always promised to.

We are not theorising this shift. We are building it.

Our clients operate complex digital platforms such as Optimizely across enterprise environments where delivery, experimentation, customer data, finance systems and operations constantly interact.

To support this reality we built Ovie.

An AI interface that transforms how those systems are understood and supported. 

Ovie allows organisations to:

• Interrogate the history of their systems

• Understand incidents, releases and experiments in context

• Learn from operational behaviour

• Improve systems continuously

Not as a dashboard.

As a living institutional memory layer. 

​

What This Means 

Systems that remember improve quickly.

Systems that forget repeat work.

As AI capabilities expand, the constraint will not be the models themselves.

It will be the environments in which they operate.

And that changes where long-term value sits in the technology ecosystem.

Not only in platforms, such as Optimizely. 

But in the operating layers that allow organisations to learn from them.

​

And, What Happens Next? 

For years the enterprise brief sounded familiar.

“We need a better website.

Better UX.

Better CRO.

Better SEO.”

All very important work. But surface-level work.

​

The new era will sound different.

​

“Help us build systems that learn.

Systems that remember what happened.

Systems that tell us what to do next.”

​

Systems that accumulate knowledge, so that each improvement builds on the last.

​

Industries like aviation learned this lesson decades ago.

​

Digital systems are only just beginning to.

​

That shift changes how software is built.

It changes how organisations learn.

And it changes how technology partners must operate.

​

We are designing for that reality.

​

​

First Three Things

Architecting digital systems that learn.​​​​