In the past few months of talking with retailers, manufacturers, and analysts across Australia, New Zealand, and the U.S., one pattern keeps surfacing—companies are swimming in data but starving for clarity.
Every executive phrases it differently, but the sentiment is near universal:
“We have more data than we know what to do with—but not the kind we can actually use to make decisions.”
And they’re right.
Gartner, IBM, IDC, and MIT have all estimated that between 80 and 90 percent of enterprise data sits outside structured systems - beyond the reach of the ERPs, CRMs, and BI dashboards we depend on. That’s the hidden 85%.
It’s all the emails, text notes, supplier messages, PDFs, spreadsheets, product images, store videos, IoT readings, and post-meeting comments that never make it into a database. The stuff that actually explains why things happened - but which most systems can’t even read.
The Two Worlds Problem
Structured data is the world of rows, columns, and schemas.
Sales by SKU. Margin by week. Supplier lead time in days. Clean, logical, and precise - but slow to update and painfully incomplete.
Unstructured data is the world of everything else. It’s what happens when a buyer sends a message saying, “Factory delay, shipments pushed back two weeks.” It’s the retailer noticing foot traffic collapse because of roadworks. It’s the sudden demand surge because an influencer wore the brand on TikTok.
These aren’t random anecdotes - they’re cause and effect. But because they don’t fit the model, they get lost outside it.
So businesses end up planning on the structured world - and reacting to the unstructured one.
IDC says 85% of enterprise data is unstructured, and yet only about 15–20% of it ever gets analysed. In retail, that’s the difference between seeing that a SKU is slow-moving and knowing why. In healthcare, it’s the insight buried in a doctor’s handwritten note. In finance, it’s a risk warning sitting inside an email thread no one ever mined.
The world’s most valuable signals sit in places our systems don’t look.
Why Forecasting Breaks Down
Traditional forecasting tools are designed for the structured world.
They assume clean data, stable lead times, and predictable behaviour.
But the real world doesn’t work that way. Lead times shift. Suppliers miss deadlines. Promotions pull demand forward. Weather changes demand curves. And none of this lives inside an ERP field.
That’s why, when we sit with planning teams, the same story repeats: “The system gives us one number. But it doesn’t feel right.”
And so they export the data to Excel, cross-check it against what they know, and spend days re-building what the system should have known in the first place.
That three-week forecasting cycle you see in mid-market retail? It’s not a process problem. It’s a data comprehension problem.
Where RabbitHawk Fits In
RabbitHawk was built for that messy middle—where structured data ends and the real world begins.
Our forecasting and optimisation engine connects directly into systems like NetSuite, but then goes further - ingesting unstructured signals alongside structured ones.
It learns from every constraint and every outcome—SKU × Store × Week × Context—blending statistical precision with lived experience.
It treats unstructured data (emails, comments, notes, trends) as part of the forecast itself, not as noise to be cleaned out.
In other words, it unifies the model and the market.
It doesn’t just predict what will sell—it understands why, and how those drivers might shift next week.
It’s what happens when AI stops being theoretical and starts working the way retail really does.
The Broader Truth
Across every sector we’ve looked at—retail, finance, healthcare, manufacturing - the same imbalance holds.
Around 80–90% of operational data lives outside structured systems, yet almost all of our analytics budgets chase the 10–20% we already understand.
Healthcare’s “dark data” sits in clinician notes and imaging files.
Finance’s sits in contracts, transcripts, and communication trails.
Retail’s sits in everything from markdown memos to supplier updates.
And that 85% is where foresight hides.
From Data to Foresight
Perfect data doesn’t exist.
Lead times change, files go missing, stores miscount stock.
But decisions still have to be made - every day.
Allocating stock.
Setting promotions.
Authorising purchase orders.
The best companies don’t chase certainty; they build systems that thrive in uncertainty.
RabbitHawk helps teams make faster, better decisions when the data is incomplete, the timeline is short, and the stakes are high - by bringing the 85% back into play.
Because in the end, business performance isn’t limited by how much data you have.
It’s limited by how much of it you can actually use.
And that’s where foresight—real, usable foresight—begins.
