In the past few months of talking with retailers across Australia and New Zealand, one thing keeps coming up - too much stock, in the wrong places, stuck, for too long.
Every CFO we’ve met expresses the cause differently, but the theme is always the same.
“We’re not short on data —we’re short on foresight. And one of the things lack of foresight leads to is extra stock sitting somewhere, piles of cash that should be utilised elsewhere.”
Most retailers aren’t suffering from a lack of systems. They’re surrounded by them. One company we spoke with had six separate forecasting systems running at once. ERPs and demand-planning tools - built to manage structured data like sales, SKUs, suppliers, and lead times - are often so rigid that teams end up exporting everything into spreadsheets or Power BI just to make sense of it.
The result? Two parallel worlds.
The structured world of models, reports, and approvals—slow, and not as accurate as they profess. And the unstructured world—what buyers and planners see, hear, and feel on the ground: a retailer’s IT outage, a buyer pre-ordering before a holiday, a shipment delay, a sudden weather shift, or roadworks that kill foot traffic.
The problem is that these two worlds rarely meet. Businesses end up over-buying because of what they think they know, and remain under-prepared for what they should know.
Capital gets trapped in aging inventory. Buyers doubt themselves and their systems. Suppliers become unwilling bankers. And the cycle repeats.
Which is why we built RabbitHawk to work the way the world works. A system that unites the structured and the unstructured into a single adaptive intelligence.
RabbitHawk learns from every constraint and every outcome (as granular as SKU by location in store x time) blending statistical precision with real-world context. It accounts for lead times, MOQs, promotions, local quirks, even external signals like weather or social trends.
This isn’t a LLM next-word probability model. Instead it’s the conversion of human experience and off-system knowledge into forecastable data.
This is no small shift.
Gartner, IBM, IDC, Deloitte, MIT and others estimated that 80-90% of enterprise data sits outside of structured data models. Text, emails, images, audio, video, social media posts, and web logs. And in this unstructured data we see lead times shift. Supplier performances vary. Weather changes. Promotions misfire. Roadworks, holidays, sporting events, politics. Specifically with inventory management, the judgment of experienced buyers—the kind that keeps businesses alive —rarely makes it into a database. And so we return to teams avoiding ‘smart’ systems, downloading data and guessing inside spreadsheets.
Yet decisions still have to be made, every day: allocating stock, placing POs, setting promotions. Waiting for perfect information means waiting too long.
No one can wait for certainty.
But now retailers don’t have to wait for foresight.
