Skip to main content

Tightly's Recommendation Engine

Cagla Sener avatar
Written by Cagla Sener
Updated over a week ago

Tightly’s Replenishment Recommendation Engine helps you restock at the right time, with the right quantity, and from the right supplier. It combines your past sales, current stock levels, and supplier data to make sure you never run out of your best-performing products.

What drives recommendations?

Tightly looks at three main areas:

  1. Profitability

    • Revenue performance

    • Profit margins (cost vs. sell price)

    • Sales consistency

  2. Demand Forecasting

    • Historical sales velocity

    • User overrides via Demand Planning

    • Events like promotions or seasonal changes

  3. Operational Parameters

    • Supplier lead time

    • Replenishment frequency (daily/weekly/monthly)

    • Safety stock buffer


Priority rules

When deciding what to recommend first, Tightly follows this hierarchy:

  1. Replenishment Sets → your manual rules always win.

  2. Composite Score → higher-scoring products (best performers) get priority.

  3. Lead Time → faster suppliers are preferred.

  4. Default Supplier → if you’ve set a default supplier, we stick to it.


🚫 When recommendations are NOT shown

Tightly skips recommendations in these cases:

  • Not enough sales history → at least 7 days of sales within the most recent 31 days (when stock was available).

  • Zero demand → products with no sales velocity.

  • Missing key data → no supplier assigned or corrupted/missing stock levels.

  • Budget exceeded → recommendations that push you over your monthly/category budget won’t appear.


What data matters most?

  • Must-have data:

    • Active product catalog with SKUs

    • Real-time inventory levels

    • Valid sales history (7+ days within recent 31 non-stockout days)

  • Accuracy boosters:

    • Correct lead times (variant > supplier > org default)

    • Current unit costs (for budget and profit checks)

    • At least 60+ days of sales data for seasonal trends

  • Optional but powerful:

    • Budget configuration by category and overall

    • Default suppliers set for variants


How budgets are applied

  • Tightly checks monthly category and overall budgets.

  • Recommendations are processed in order of composite score.

  • Each recommendation’s cost = recommended quantity × unit cost.

  • If a product doesn’t fit the budget, it won’t be shown.

  • Exceptions: missing costs (skipped from budget checks) or replenishment sets (always prioritized).


How it works (behind the scenes)

  1. Sales velocity → calculates demand from your past 365 days (or uses your overrides).

  2. Day-by-day simulation → projects inventory 90 days forward, factoring in sales, stock, and deliveries.

  3. Safety stock → dynamic buffer to protect against demand spikes and supplier delays.

  4. Composite score → ranks products by revenue, volume, and profitability.

  5. Final priority check → applies supplier preferences, budgets, and replenishment set rules.


Credits and recommendations

  • Calculation phase (free): Tightly calculates recommendations automatically in the background.

  • Generation phase (credits): When you generate recommendations, credits are consumed (1 credit per unit recommended).


Inventory health statuses

Tightly also classifies your stock health:

  • Critical → Stockout or falling below safety stock

  • Caution → Overstock (too much stock)

  • Healthy → Inventory levels are balanced


Key takeaways

  • Manual rules (Replenishment Sets) always override smart suggestions.

  • At least 7 days of recent sales are required for a product to qualify.

  • Recommendations won’t exceed your budgets.

  • Credits are only used when you choose to generate recommendations.

  • Tightly automatically manages bundles and supplier preferences.

Did this answer your question?