What happened when the ideas met reality?

Evidence

Cralgo observes what changes when questions are explored, systems are introduced and real environments respond.

Situations are shared; client detail stays private. What is useful to read is the observation, the outcome and the question that survived it.

Read the observed work

Signal · response · observation · what is still open

The premise

Outcomes have to be observed.

A question can sound convincing. A system can be well designed. An intervention can appear sensible.

The real environment decides what happens next.

• what changed
• what did not
• what people did
• how the system behaved
• how the organisation responded
• what outcome emerged

What counts as evidence

Four ways reality answers.

Observation

What behaviour or condition became visible?

Measurement

What changed quantitatively, and over what period?

Experiment result

What happened when something was altered or introduced?

Field finding

What became apparent over time inside a working environment?

Application and outcome are not evidence.

Application

Where the question entered reality.

Evidence

What we observed or measured.

Outcome

What emerged.

Outcome is what happened. Evidence is what lets us understand it.

Observed

• adoption increased
• users hesitated
• decisions took less time
• a team bypassed the intended process

Outcome

• better adoption
• a worse experience
• faster execution
• no meaningful change

Evidence explains the outcome. It is never a substitute for it, and an outcome is never a fourth lens.

Observed in practice

Applied evidence from real environments.

Each entry records the question, the context, what was observed, what emerged and what is still open. Where the evidence is early, mixed or inconclusive, it says so.

Omnichannel

Observed in practice

When online and offline needed to start behaving like one business.

Question

Could one customer promise be kept across channels that were operating as separate businesses?

Context

Stores, ecommerce and the app were each performing against their own view of stock, promise and priority. Customers experienced one brand; the systems underneath still worked as separate businesses.

Lenses involved

Technology × Organisations

Intervention

Cralgo worked alongside leadership, internal teams and existing partners to read the situation as one operating picture rather than a set of channel problems.

Observation

The difficulty was less about any single platform and more about how availability, order routing and fulfilment decisions were held across commerce, ERP, order management and operations — and who was accountable when two of them disagreed.

Outcome

Channel decisions began to be taken against one promise rather than three separate views of it.

Learning

Before choosing an order management or inventory approach, it helps to be clear on the customer promise the business intends to keep, and which decisions the systems must be able to take on their own.

Still open

Whether the improvement held once seasonal volume returned was not measured in this period.

Do we need an OMS?

Product & Technology operating model

Observed in practice

When too many Product & Technology decisions kept coming back to the founder.

Question

Where were Product & Technology decisions actually being taken?

Context

Capable teams were in place, work was moving, and the founder was still the point where product direction, technology judgement and trade-offs were resolved.

Lenses involved

Psychology × Organisations

Intervention

Cralgo participated close to leadership, separating which decisions belonged where and giving the existing teams a shape to hold them in.

Observation

The boundaries of ownership were still unclear. Product decisions and technology decisions were treated as one conversation, so ownership remained unclear and leadership visibility depended on the founder being in the room.

Outcome

Fewer decisions travelled back to the founder; ownership sat closer to the work.

Learning

Founder dependency is rarely a capability problem. It is usually an ownership problem waiting to be made explicit.

Still open

Whether the new ownership survives leadership change is not yet observable.

The founder technology bottleneck

Platform & architecture

Ongoing

The platform had started setting the pace of the business.

Question

Was the platform limiting what the business could attempt next?

Context

Every commercial initiative required a deeper change than it appeared to, and each one left the platform heavier than before.

Lenses involved

Technology × Organisations

Intervention

Cralgo worked on sequencing — reading architecture against commercial dependency rather than technical preference, with building and delivery staying with the organisation's own teams and partners.

Observation

What the platform could carry had quietly begun to decide what the business could attempt next. The useful separation was between what the platform had become and what the business actually required.

Outcome

Sequencing changed. Some platform work was deferred; some was brought forward.

Learning

Architecture questions are commercial questions first. What the business intends to do next decides which parts of the platform deserve attention.

Still open

The commercial effect of that resequencing is still being measured.

Headless commerce, or not

Field observations

Conversion

Inconclusive

Conversion needed the whole organisation to learn together.

Why did teams reach different explanations for the same drop?

Each team was learning on its own. Discovery, product pages, search, trust, checkout friction and payment behaviour sat with different owners, and a settled rhythm for testing a view and moving on was missing.

No single cause emerged. The learning rhythm mattered more than any one fix.

The more useful question is often where customers dropped, and whether the organisation can learn quickly enough to find out.

Attribution across discovery, trust and checkout remained unresolved.

Delivery

Mixed

Delivery kept moving, while predictability kept slipping.

Why did delivery keep moving while predictability slipped?

The decision surface was fragmented across teams. Every team had a view of progress; the whole picture sat outside any single team, so coordination cost more than the work itself.

Predictability improved where decisions were held; it did not improve where they were not.

Predictability comes from where decisions are taken and held, more than from how the work is tracked.

Mixed — the effect varied by team, and the reason for that variation is still open.

Why ecommerce breaks at scale

Vendors & partners

Early evidence

Several partners were delivering. The whole picture needed one view.

Who was carrying the judgement across several delivering partners?

The organisation was carrying integration and judgement work it had assumed someone else owned. Dependency had formed where capability was intended.

Integration and judgement work became visible as work someone had to own.

Partner arrangements work when the organisation retains the judgement, and knows which capability it intends to keep.

Whether the retained capability stays retained is a longer observation.

In-house, agency, or Centre of Excellence

AI & innovation

Field observation

The real question was where judgement belonged.

Which decisions genuinely required judgement, and which did not?

Separating routine decisions from those genuinely requiring judgement mattered more than tool selection — and made clear which experiments were worth turning into operating capability.

Escalations reduced where routine decisions were separated out. Tooling changed little on its own.

Innovation usually starts with a better question rather than a new tool. What changes is where decisions are taken, more than how much technology is present.

No claim is made that the tools themselves changed the outcome.

Three ways to read it

One record of evidence, read in three ways.

Observed Situations

Individual environments where a question met reality and something became observable.

Read the observed situations

Field Notes

Smaller bounded observations from real environments, recorded before they have been generalised into a broader claim.

Read the field notes

Change Over Time

A time-based reading of how some of those observations persisted, regressed or adapted after the first response.

Read the observations across time

Reusable guidance and reference structures are not evidence — playbooks and architectures are Knowledge. Change Over Time holds no separate dataset. It reads the same records held in Observed Situations along time. Evidence can strengthen a belief, weaken it or change the question entirely.

Reading the evidence

One observation, read through three lenses.

Evidence is not filed by lens. A single situation usually reveals something through each of them at once.

Psychology

What people actually did — not what the plan assumed they would do.

Technology

How the system behaved, and how it influenced the action taken around it.

Organisations

How the environment enabled, constrained or quietly altered what happened.

Outcome

What emerged from the interaction between the three. Never a lens of its own.

Signal · response · observation

Research and evidence

Research asks the question. Evidence shows what reality returned.

Research

What are we trying to understand? Questions, concepts and active explorations.

See what we're investigating

Evidence

What did reality show us? Observations, measures and what was learned.

Question → Exploration → Application → Evidence → Question.

Evidence does not close the loop. It changes the question that re-enters it.

Question → Exploration → Application → Evidence → Question.

Where the evidence comes from

Applications give the question a context. Evidence shows what the context returned.

Consumer and commerce, organisations and work, wellbeing and learning, emerging research — each environment returns a different answer to a similar question. Technology sometimes appears as the intervention, sometimes as the thing being studied, sometimes only as the means of observing.

Building something is not evidence. What changed after it entered the environment is.

See where the questions are applied

Across the situations

What recurs, whatever the domain.

The domains differ. What sits underneath them is usually the same.

Ownership before tooling

Most situations resolve into a question of who decides, rather than what to buy.

One operating picture

Channels, systems, teams and partners read together rather than as separate relationships.

Decisions that stay made

Where commitments keep moving, the missing piece is usually the decision surface, more than the effort itself.

Commercial sequence

Technology work earns its place against what the business intends to do next.

Capability that remains

What has to exist internally is built to stay there once Cralgo's attention moves on.

These recurring situations are worked through in full in the Knowledge pages.

Back to the Signals

What emerged from the exploration?

Evidence can strengthen a Signal, complicate it or surface a better question. The connection remains visible as the thinking evolves.

Signal 001

An organisation can begin to reflect on itself as one connected whole.

Explore Signal

Signal 006

AI is opening a new way for organisations to observe continuously.

Explore Signal

Signal 010

Customer understanding now spans more than the customer.

Explore Signal

Where this leads

Evidence does not end the question. It changes the next one.

Some of the observations above are settled. Several are early, mixed or unresolved, and are written that way on purpose. The evidence here reflects where the work has actually reached.

Explore the research See where the questions are applied