Purchase Thresholds
Purchase Thresholds is where you define the inventory range you want to maintain for each product at each location. A threshold has two numbers — a minimum (the level you don't want to drop below) and a target (the level you want to restore to when you reorder or rebalance). These thresholds drive two downstream features: the Min/Max Replenishment plan type in the Purchase Order Planner, which recommends reorders when stock dips below the minimum, and the Balance Stock recommendation engine, which suggests transfers from locations above their target to locations below their minimum. Thresholds are scoped per (variant, location), so the same variant can have different thresholds at different stores or warehouses.
Key Concepts
Minimum and Target
The minimum is your trigger point — the level you don't want stock to drop below. The target is where you want to be after replenishment or rebalancing. For example, setting min 10 and target 30 for a product at your DTLA store means: 'Don't let DTLA drop below 10 units; when you reorder or rebalance, top it back up to 30.' Target must be greater than minimum; Alfred won't save a threshold with target ≤ minimum.
Per-Location Scope
Thresholds are tied to a specific (variant, location) pair. The same product can have different thresholds at different locations — a high-traffic flagship store might run min 20 / target 60, while a quieter satellite store might run min 5 / target 15 for the same SKU. The Purchase Thresholds page shows one location at a time; switch locations using the dropdown at the top.
Where Thresholds Are Used
Two features depend on thresholds. (1) The Min/Max Replenishment plan type in the Purchase Order Planner: when a (variant, location) is at or below its minimum, Alfred recommends ordering enough to bring projected on-hand inventory back up to the target — accounting for in-transit POs and expected sales during lead time. (2) The Balance Stock recommendation engine: it scans thresholded locations and recommends transfers from locations above target to locations below minimum. Without thresholds, neither feature can produce recommendations for the variant.
Why Thorough Setup Matters
Balance Stock recommendations require thresholds at every stocked location for the variant. If a variant lives at three locations but only two have thresholds, Alfred can suggest moves between the two configured locations but won't pull from or push to the third — even if the third has excess stock or is about to stock out. The threshold page surfaces a 'missing thresholds' count to help you find gaps.
Auto-Populate from Velocity
Instead of setting every threshold by hand, you can ask Alfred to calculate suggested min and target values from sales velocity. You pick a lookback period (7–90 days, default 30), a min days-of-cover, and a target days-of-cover. Alfred multiplies each variant's daily velocity by the day counts and rounds up. Products with no sales in the lookback period are skipped. The calculated values appear in the table as pending edits — review them, tweak anything that looks off, and save. This is a one-time snapshot; thresholds don't update automatically as sales change.
Thresholds vs. Safety Stock
A min/max threshold is a simple reorder rule, not a probabilistic safety stock calculation. It doesn't model demand variability or service levels — it just says 'reorder when you hit X, refill to Y.' For most small-to-mid stores this is exactly what you want: predictable, easy to reason about, and easy to audit. If you need finer-grained service-level math, lean on the demand-based plan types in the Purchase Order Planner instead.
How to Use
1. Pick a Location
Open Configurations → Purchase Thresholds. Use the location dropdown at the top to pick the location you're configuring. Thresholds are per-location, so you'll repeat this step for each location.
2. Set Thresholds Per Variant
Find the variant you want to configure using search or the filter chips (vendor, product type, status, tags, collections). Enter a minimum and target in the row's input fields. The save bar appears at the top showing pending changes. Click Save to commit. Repeat for additional variants.
3. Bulk-Edit Multiple Variants
Tick the checkboxes for several variants at once. Use the bulk action bar to enter a single min and target value, then apply to all selected. This is the fastest way to set the same thresholds across a category — for example, applying min 5 / target 20 to every variant in a vendor or collection.
4. Auto-Populate from Velocity (Optional)
Click 'Auto-populate from velocity' in the page header. In the modal, pick the lookback period (how many days of sales to look at), the min days-of-cover, and the target days-of-cover. Alfred calculates suggested values for every active variant with sales in the period and loads them into the table as pending edits. Review the suggestions, override any that don't fit, and save.
5. Repeat for Each Location
Switch to the next location using the dropdown and configure thresholds there too. For Balance Stock recommendations to cover a variant fully, every location where the variant is stocked should have thresholds set.
Tips & Best Practices
- ✓Set thresholds for your top-selling products first. Velocity is concentrated — a small share of your SKUs drives most of your sales, and thresholds on those have the highest payoff.
- ✓Use auto-populate as a starting baseline, not a final answer. The calculated values are mechanical — multiply velocity by days. Review the results, especially for products with unusual demand patterns (promotions, seasonality, one-off spikes).
- ✓Keep target meaningfully larger than min. A target only one or two units above min produces noisy, twitchy recommendations. A wider band (e.g., min 10 / target 30) gives stable reorder cycles.
- ✓Configure thresholds at all locations where a variant is stocked. Partial setup leaves Balance Stock unable to use the un-configured locations as either a source or a destination.
- ✓Revisit thresholds quarterly, or whenever sales patterns shift materially. They don't auto-update — if a product's velocity doubles, the threshold you set six months ago is stale.
- ✓Use the bulk-edit and auto-populate features together: auto-populate to seed values for your active catalog, then bulk-edit to tune categories of products that share characteristics.