Dynamic Narrative Analytics: The DNA of Data

Over the years, I’ve looked at a lot of numbers. Far too many really. Like many of us, I’ve struggled to keep my head above the surface of massive pools of data, desperately trying to understand what the floating numbers are trying to tell me. Excel straining, my sense of self dissolving, the what is often staring me in the face, but what I’m missing is the so what?

The fundamental challenge of the modern workplace is that data, in its raw form, is mute. A number on a screen has no context, no history, and no motivation. Yet we project onto it the weight of being an “answer.” We present these silent numbers in meetings and expect them to drive brilliant decisions, but they often fall flat, creating more confusion than clarity.

The DNA Philosophy: Every Dataset Has a Narrative

It took me far too long to realise that the answer was not in the numbers, but in the story they told. All data, when given context, can tell a story and at heart, I’ve always been a story teller.

The Dynamic Narrative Analytics (DNA) philosophy is built on this single, foundational idea. Whether it’s a quarterly financial report, a user feedback survey, or a complex A/B test, there are hidden forces at play—positive and negative, connecting all of that data into a story. Our job is to uncover it.

The philosophy proposes that to do this, we must interrogate our data with a disciplined, holistic, and consistently human-centric set of questions. Instead of just asking, “What happened?”, we must act as skeptical investigators.

It is not an algorithm for analysing data, rather a narrative layer on top of the data that follows four key rules or questions to present the story to the end user.

Because I love a slightly contrived framework, the DNA philosophy has a very on brand acronym to cover the four questions. AGCT: Assurance, Gain, Clarity, and Threat.

The Question of Trust (Assurance): Can I even trust this information? This is the bedrock of any credible story. Before we believe the narrative, we must validate the source. Is the data clean? Was the collection method sound? Is there a hidden bias? There is no story without trust.

The Question of Opportunity (Gain): What is the good news here? Every story needs a positive force. What in this data represents growth, success, or opportunity? Where is the upward trend? This gives us the measure of the total potential upside.

The Question of Focus (Clarity): Did we achieve our primary objective? In the midst of all this opportunity and risk, what was the one central plot point we must pay attention to? Every good story has a focal point, and it’s our job to find it in the data.

The Question of Risk (Threat): What is the bad news here? A story without conflict is a fairy tale. We must actively seek out the villain, the friction, or the challenge. What represents a loss, a cost, or a danger? Being hyper-aware of risk is the basis of mature, defensible decision-making.

As I say, this is a layer, a mental model for critical thinking. It is a way of seeing. And its power is best demonstrated by applying it to completely different worlds.

Putting the Philosophy to the Test

Let’s see how this might work in some practical applications.

Application 1: The A/B Test (A Formal Analysis)

Imagine AGCT is built into an automatic DNA engine for presenting analytics for A/B tests.

Assurance: The engine asks, “Was traffic split fairly? Did the test run long enough? Was the volume sufficient?” If not, the story is fiction, and the analysis stops.

Gain: It looks at all metrics that moved positively, weights them by business importance, and compiles a single score for the total upside.

Clarity: It isolates the primary goal and asks with laser focus, “Did we achieve this specific thing with high statistical certainty?”

Threat: It pessimistically hunts for any metric that moved in a negative direction, especially critical “guardrails” that protect the business. It is hyper-sensitive to harm. Did we see a drop in AOV? Did bounce rate go up?

From this structured inquiry, a rich narrative verdict is born: “Risky Winner”, “Net Loser”, “Solid Opportunity”. And from there a complete story can be born, not just a single green arrow.

You are not looking at just the goal metric, you are looking at the full picture, weighing up the overall risk and reward of the test. “Whilst we saw an increase in the goal metric of conversion rate, we saw a significant drop on revenue and aov. Stop the test and re-evaluate the hypothesis.

Application 2: The Annual Report (A Philosophical Analysis)

Now, let’s leave the world of p-values and apply the exact same thinking to a glossy, 80-page corporate annual report. This is not a formal engine, but a way of reading and interpreting.

Assurance: Before reading a single headline, we ask about the source. “Has this been audited by an independent firm? Are there footnotes detailing changes in accounting methods? Are they relying on vague ‘adjusted’ metrics?” This is the intellectual equivalent of the Assurance score; it establishes the report’s credibility.

Gain: This is the story the report wants to tell. “Headline revenue is up 20%. Net profit is at a record high.” We gather all these positive points to understand the scope of the stated success.

Clarity: Last year’s report stated the #1 priority was “sustainable, organic growth.” We must now ask, “How much of that 20% growth came from an acquisition versus organic sales? Is that growth sustainable if it required cutting R&D spending?” This tests the headline claim against the stated strategic goal.

Threat: Now we become a skeptical analyst. We hunt for the counter-narrative. “Why is our market share down in a key region? Why is employee turnover up by 25%? Why has our debt-to-equity ratio worsened?” Actively seeking this information is crucial for a balanced view.

One Philosophy, A Universe of Understanding

And look at that, with one framework consisting of 4 simple questions, we’ve transformed a passive reading of an annual report into an active, critical analysis. We have moved beyond simply accepting the headline numbers and have constructed a deeper, more honest narrative.

Instead of saying, “It was a great year with 20% growth,” our story becomes:

“The company delivered impressive top-line growth, but, we did see a higher than usual employee churn”

We’ve taken raw data and turned it into a simple, but much more human friendly narrative. We have taken the full context of all of the data, not just the headline data, and built an easy to comprehend window into the true meaning of it all.

This seems like an ideal use for Generative AI if you ask me 😉