Skip to main content

Use AI Inventory to plan reorders

Note

This feature is currently in closed beta. Contact your Zenoti Representative to get the feature enabled.

This article explains how to use AI Inventory to make daily reorder decisions at your center. You will learn how to activate the feature, use the dashboard, understand why a product is being recommended for reorder, create purchase orders with a suggested vendor, and manage the alerts that keep your inventory healthy.

Overview

Running an inventory-heavy business means staying on top of stock levels every day. For a medspa managing injectables, or a salon with fifty or more retail SKUs, a missed reorder can mean a canceled appointment, a lost sale, or a last-minute order at premium prices. The traditional approach — pulling several reports, opening a spreadsheet, and deciding what to order from experience — is slow, error-prone, and hard to scale across centers. AI Inventory replaces that manual review with an AI-assisted dashboard. Each night, Zenoti looks at what your center has consumed, what is on hand, what is already on order, and what is coming up on your appointment calendar. The next morning, the dashboard shows a single list of products that need attention, sorted by urgency, with a recommended quantity for each one. You review the list, adjust where needed, and create purchase orders — all from one screen. Every recommendation is explained in plain English, so you can trust the number before you send the order. AI Inventory does not place orders on your behalf. Every purchase order it generates is saved in Draft status for you to review and raise. You remain in control of every decision; Zenoti does the math.

Note

AI Inventory is currently available to selected centers. Contact your Zenoti account manager to confirm availability at your organization.

Key terms

Velocity: Your average daily consumption of a product, calculated from the last 90 days of activity. Retail velocity and consumable velocity are calculated separately because the two product types are consumed through different mechanisms.

Lead time: The number of days between placing a purchase order and receiving stock. Where possible, Zenoti calculates this from your last five purchase orders for the product at that center. Where PO history is thin, the default lead time you set during activation is used.

Safety stock: A buffer stock level, measured in days, to protect against late deliveries or demand spikes.

Coverage days: The number of days of stock you want to have on hand after reordering. A higher number means fewer, larger orders; a lower number means more frequent, smaller orders.

Dynamic Reorder Point (DRP): The stock level at which AI Inventory raises a reorder alert. Calculated as (velocity × lead time) + safety stock.

Days of Stock Cover (DSC): How many days your current stock will last at your current velocity.

Recommended Order Quantity (ROQ): The quantity that brings your stock to your coverage target, accounting for what you already have and any incoming stock. Recommendations are floored at zero (never negative) and capped at three times the coverage target as a sanity check.

Confidence tier: A badge from T1 (High) to T4 (Static) that indicates how much data the system had when calculating the recommendation. Higher tiers use more adjustments; lower tiers rely on simple averages or defaults.

Operating model: How a center records stock inflows. Zenoti classifies centers automatically as PO-driven, Hybrid, or Adjustment-driven, and offers different capabilities accordingly.

Prerequisites

AI Inventory is most accurate when the center already maintains a baseline of inventory discipline:

  • Purchase orders are raised and deliveries are marked in Zenoti. The system calculates lead times from your past purchase orders, so recent PO history directly improves accuracy. 

  • The Stock Ledger is active at your center, and vendors are linked to products in the Product Master

  • For professional (consumable) products, a Bill of Materials (BOM) is configured against the services that use those products. This lets the system use upcoming appointments as a forward-demand signal. 

  • The center calendar is up to date with holidays, closures, and promotion periods, so demand adjustments account for known variations.

Limitations

The current release of AI Inventory does not yet include the following. Continue to use existing inventory workflows for these scenarios:

  • Organization-wide roll-up across centers (AI Inventory operates per center; use the center picker to switch)

  • Automatic transfer order recommendations between centers

  • Minimum order quantity, pack size, and tiered pricing constraints

  • Statistical safety stock calculations based on service-level targets

  • Advanced forecasting (ARIMA, exponential smoothing, seasonal decomposition)

  • Batch and expiry-aware recommendations (first-expiry-first-out logic)

  • Budget-based prioritization against a monthly spend cap

  • Vendor performance scoring

  • Fully autonomous ordering (every purchase order is created in Draft status and must be raised by a user)

Centers that record most stock inflows through positive adjustments rather than purchase orders still see the dashboard and a printable restock list, but one-click PO generation and vendor selection are not available to them. A persistent banner on the dashboard reads: Enable PO workflow to unlock automatic vendor selection and one-click ordering.

Activate AI Inventory

The first time you open AI Inventory for a center, a four-step wizard walks you through the settings that the system needs to start making recommendations.

At the center level, navigate to Inventory > AI Inventory.

Step 1: Set the lead time

On the first screen, choose one of two options.

  • Select Use system-calculated lead times if your center has at least three purchase orders with marked deliveries for at least thirty percent of active products. The option shows how many delivery records it will use. As new PO history accumulates, Zenoti recalculates the lead time for each product from the last five deliveries.

  • Select Set a default lead time to enter an organization-wide default. The field is pre-filled with five days. If certain categories typically take longer to arrive, override the default by category — for example, Injectables at seven days while keeping Shampoo at five.

Note

Lead time determines how far in advance you are alerted. If your vendor typically takes five days to deliver, AI Inventory alerts you when you have about five days of stock left — not after you have already run out.

Step 2: Set the safety stock

Enter how many days of safety stock you want to carry. Use the quick-set buttons for 0, 2, 3, 5, or Custom.

Zenoti recommends two to three days for most salons and medspas. Safety stock is your insurance policy — it protects against a delivery running a day late or demand running a little hotter than usual. Setting zero days means no buffer. For injectables, wax, color, and other products that directly affect service delivery, two to three days is strongly recommended.

Step 3: Set the coverage days

Select how many days of stock you want on hand after reordering. Options are 7, 14, 30, 60, 90, and Custom. Thirty days is the default.

This setting controls order size. A small coverage target means frequent small orders; a large coverage target means occasional large orders. For most salons and medspas, thirty days strikes a reasonable balance between carrying cost and ordering frequency.

Step 4: Confirm and activate

Review your selections on the summary screen and click Activate AI Inventory. Zenoti calculates the reorder point for every active product at the center, and the dashboard is populated. You can change any of these settings later from the inventory configuration page.

Use the dashboard

At the center level, navigate to Inventory > AI Inventory.

The top of the dashboard shows a row of tiles summarizing the state of your inventory. Below Reorder Point is the count of products that have dropped below their dynamic reorder point but still have stock on hand. Stockout / Negative is the count of products at or below zero. Avg Days of Cover is the mean days of stock cover across your active products. Pending Orders is the count of open purchase orders that have not been fully received. Click any tile to filter the list below.

Below the tiles is the Products Needing Attention list. This is the heart of the dashboard and the list you will work through each day. It shows the products that need action, sorted with the most urgent first. Data integrity issues (anything with on-hand at or below zero) are pinned to the top, followed by critical-risk items (where stock will run out before a delivery can arrive), followed by warning items (where the buffer is getting thin).

Each row shows the product name, whether it is retail or consumable, the category, current stock, the dynamic reorder point, days of stock cover, a risk badge, the recommended order quantity, and a confidence tier. You can sort by any column and filter by category, product type, or confidence.

Note

If no products are below the reorder point, the dashboard shows an "All clear" message. Your inventory is in good shape; check back after the next calculation.

Understand why AI Inventory recommends an order

Every recommendation is explainable. Click Why this? on any row to open a panel that shows the reasoning.

The panel starts with a one-sentence summary in plain English. For a retail product, something like "Order 132 units to cover 30 days of expected demand based on your current sales rate of 5 units/day." For a consumable, "Order 45 vials to cover 30 days of usage. You have 28 Botox appointments booked this week, which is higher than your usual rate."

Below the summary is a breakdown of every input the calculation used: velocity, lead time, safety stock, coverage days, current stock, and incoming stock. For each input, the panel shows whether the value was derived from your data, taken from a configured default, or set as a manual override. For consumables with a BOM linked to upcoming appointments, the panel also shows the forward-demand rate and which rate — historical or forward — the system used.

Finally, the panel shows the confidence tier for the recommendation and explains what it means. A T1 recommendation might read "High confidence — based on 142 days of consumption history and system-calculated lead time from 7 deliveries." A T3 might read "Low confidence — this product has only 22 days of history. The recommendation uses a simple average without trend adjustments. Review this quantity manually before ordering."

Use the panel to decide whether to accept the recommendation, adjust the quantity, or investigate further. If a value looks wrong — for example, a lead time that is shorter than your vendor actually takes — set a manual override at the product level and AI Inventory will use your value on the next calculation.

Create purchase orders from the dashboard

Once you have reviewed the recommendations, you can create purchase orders in a few clicks.

Select the products you want to order using the checkboxes. To select every item below the reorder point in one step, use Select All below DRP. Then click Prepare Order. A side panel opens with a vendor allocation grid, where each product is pre-assigned to a vendor using the Most Frequently Ordered strategy.

At the top of the panel, switch the vendor strategy to match how you want the order allocated:

  • Most Frequently Ordered picks the vendor with the most past POs for each product and is the default.

  • Lowest Cost picks the vendor with the lowest last purchase price.

  • Fastest Delivery picks the vendor with the shortest average lead time.

  • Custom lets you choose the vendor manually for each row.

Regardless of which strategy you pick, the panel always shows vendor frequency, last purchase price, and lead time for every row, so you can compare as you work.

Review each row. If a row needs a different vendor, select one from the dropdown. If a recommended quantity looks off, edit it inline — the line total updates automatically. If your edit differs from the recommended quantity, AI Inventory displays a brief acknowledgment showing the percentage change and the resulting days of coverage. The footer of the panel shows the total items, the number of vendors, and the estimated cost.

When you are ready, click Generate Purchase Orders. Zenoti creates one Draft purchase order per vendor, tagged with AI Inventory as the source. Open each Draft PO, verify the details, and raise it.

Once a PO has been raised, the corresponding alerts move to In-Transit state and are suppressed from the dashboard until the expected delivery date plus two days. If the stock is not received within that window, the alerts reopen.

For centers that do not raise purchase orders

If your center does not raise POs in Zenoti — or if your organization is hybrid and some products are tracked only through adjustments — the flow is slightly different. When you click Prepare Order, instead of the vendor allocation grid, a restock summary opens. You can print the summary, copy it to the clipboard, or export it as CSV, and use your usual ordering process outside Zenoti. When the stock arrives, record it with a positive stock adjustment; the alert resolves automatically once on-hand rises above the reorder point.

Manage alerts

An alert passes through several states during its life. When a product first drops below the reorder point, the alert is NEW. Once you have viewed the dashboard, it moves to ACTIVE. If you are not ready to act on an alert, click the snooze icon on the row and pick a duration; the alert moves to SNOOZED and does not reappear for up to fourteen days.

When you raise a PO for the product, the alert moves to IN-TRANSIT and stays there until the expected delivery date plus two days. If the stock arrives and on-hand rises above the reorder point, the alert RESOLVES automatically. If the expected delivery window passes without the stock arriving, the alert REOPENS so you can follow up with the vendor.

Note

If a critical alert is already in-transit for a product, AI Inventory does not send duplicate alerts for that product until the state changes. This keeps you from being pinged about the same issue every day while waiting for a delivery.

Override AI Inventory values

AI Inventory applies lead time, safety stock, and coverage days in this order of priority, with the most specific setting winning: SKU > Category > Center > Organization. The values you set during activation become the organization defaults. You can override them later at any level. Lead time can be overridden at the organization, category, center, or SKU level. Safety stock and coverage days can be overridden at the organization, center, or SKU level. Every override is logged with who made it, when, the old value, and the new value, so you can audit inventory decisions over time.

Note

If a recommendation looks consistently off for a particular product — for example, the lead time used does not match what your vendor actually takes — set a manual override at the SKU level rather than adjusting the recommendation each time. The override takes effect on the next calculation.

Understand confidence tiers

Every recommendation carries a confidence badge based on how much data the system had to work with.

  • T1 (High) applies when there are ninety or more days of consumption history and a system-calculated lead time from at least five POs. This tier uses the full model — day-of-week pattern, growth trend, center calendar, and, for consumables, forward demand from booked appointments.

  • T2 (Medium) applies when there are forty-six to ninety days of history. The system still adjusts for growth trend and day-of-week, but the shorter window means less stability.

  • T3 (Low) applies when there are only fourteen to forty-five days of history. At this tier, the system uses a simple moving average without trend adjustments and relies on your configured default lead time.

  • T4 (Static) applies when the product has no consumption history. For a brand-new product, AI Inventory uses the average velocity of other products in the same category at your center as a proxy. If no comparable products exist, it falls back to your configured defaults.

Note

Treat T3 and T4 recommendations as starting points. Open the Why this? panel to see which inputs are derived and which are defaults, and adjust quantities based on your judgment. As the product builds up consumption history, the confidence tier improves automatically.

Understand when a recommendation is not shown

There are a few situations in which AI Inventory deliberately suppresses a recommendation rather than show one based on unreliable data. In each case, the dashboard explains what needs to change.

  • If a product has fewer than fourteen days of consumption history and no comparable product in the same category to use as a proxy, the row shows a Needs more data label instead of a recommendation.

  • If the current stock is at or below zero, the system shows a data-integrity alert — "Fix stock data before reordering" — and no reorder recommendation. A negative on-hand usually means a checkout or consumption was recorded without a corresponding receipt. Fixing the underlying data gives you accurate recommendations going forward.

  • If velocity is zero but stock is still on hand, the product is flagged as Slow Moving and excluded from reorder alerts.

  • If the last stock audit is more than 180 days old, a warning badge notes that stock quantity may not reflect physical inventory. Consider conducting a stock audit.

  • If a consumable product has no BOM linked to any service, the forward-demand feature cannot apply, and the Why this? panel notes that the recommendation is based on historical consumption alone.

Email alerts

AI Inventory sends two kinds of email to help you stay on top of stock without needing to open the dashboard every day.

The weekly digest summarizes the state of stock across the center (configurable to daily). It names the center, lists the count of products below the reorder point, shows the top five most urgent items with one-line explanations, and gives an estimated total restock value. Clicking any item deep-links to the dashboard with that product's Why this? panel expanded.

The critical alert is sent if any product is out of stock with non-zero velocity in the last thirty days. The email subject makes the situation clear — "URGENT: 3 products are out of stock at Downtown Center" — and the body lists the affected items. If a PO is already in-transit for a product, the critical alert for that product is suppressed until the PO is either received or reopens.