Notes from the inside of retail decisions.
Why we built Solya, how we think about decisions, agents and data, and what we’re seeing on the ground with retail teams.
How to RFP a retail decision platform
Most retail decision-platform RFPs pick the wrong vendor because they test the wrong thing. Here is a reusable structure that selects on adoption and execution.
Dynamic pricing in retail: rules dressed up, or actual decisions?
Most platforms sold as dynamic pricing are rule engines with a slicker UI. Here's a buyer's test to tell rule-based pricing apart from decisions.
Decision platform vs forecasting platform vs BI
Retail buyers conflate three distinct categories: BI describes, forecasting predicts, a decision platform decides. Here is what each is actually built for.
Sport retail runs two inventory logics at once
The core wants continuous replenishment; the seasonal drop wants sell-through and markdown. Run both on one logic and you lose on both walls.
Decision intelligence vs business intelligence
Decision intelligence and business intelligence are different paradigms, not tool tiers. Here is the structural line between them.
Data sovereignty and GDPR for a retail decision AI
A retail decision layer mostly runs on aggregates, not customer identities — which makes sovereignty and GDPR an architectural choice, not a tax.
A CFO's guide to retail AI ROI: which P&L line, and when
Most retail AI ROI decks don't survive a finance review. Here's how to tell cost-saving AI from margin-generating AI — and where it actually moves the P&L.
Continuous replenishment vs. the weekly meeting
The weekly replenishment cadence is an artifact of a pre-data era. Every week your team meets to decide is a week the network drifts further from optimal.
Build vs buy a retail decision platform: an honest framework
You have a warehouse and a data-science team, so building the decision layer in-house looks obvious. Here's the framework that tells you when it isn't.
Big-ticket retail: slow turn inverts the playbook
In furniture and big-ticket home retail, slow turn and heavy logistics amplify the cost of every wrong allocation and late markdown.
AI grocery decisions: why perishables break the order model
In grocery, the question was never how much to order. It is how much to order given spoilage, shelf-life, and the markdown clock running all day.
Managing the DIY long tail: slow movers and project baskets
In DIY retail, cutting a slow-moving SKU can lose the whole project basket. Velocity per SKU is the wrong lens for a long-tail catalogue.
AI assortment planning: the highest-leverage decision in retail
Assortment is set once a season and lives 6–9 months. No in-season tuning recovers a bad pre-season call — which is exactly where most assortment AI fails.
AI assortment and allocation in beauty retail
Beauty's SKU explosion makes assortment and store-level allocation the make-or-break calls. Here's why spreadsheets break and what a decision layer changes.
Retail returns are a feedback signal, not a logistics problem
Most retailers handle returns as a reverse-logistics flow. Treated as a signal feeding allocation and buying, the same returns recover 200–400 bps of margin.
Demand sensing in retail: forecasting for decision-readiness
Forecast accuracy is the wrong KPI. A 92%-accurate forecast that lands after the replenishment cut-off is worth less than a 78% forecast that lands in time.
Why retailers lose money between stores without knowing it
Most retail leadership teams discuss network performance store by store. But the biggest leaks aren't inside the stores — they're between them.
Why your BI tools don't make decisions (and never will)
Retail dashboards have never been clearer, and yet the same problems keep eating margin. Here's why BI tools can't fix what they were never built for.