Stockful AI: An Inventory Assistant That Actually Knows Your Shopify Store

Your inventory app has dashboards. It has reports. It has charts and tables and filters. And yet, the question you actually want answered is probably something like: “What should I reorder this week, and why?”

Dashboards do not answer questions. They present data and leave the interpretation to you. That is fine when you have time to dig through sell-through rates and cross-reference lead times with velocity trends. It is less fine when you have 600 SKUs across three locations and a supplier email to send before lunch.

Stockful’s AI release changes how you interact with your inventory data. Instead of navigating to the right report, setting the right filters, and interpreting the right numbers, you ask a question in plain English and get an answer grounded in your actual shop data, with the specific variants, locations, and metrics behind it.

The AI features are included on every Stockful plan, with a per-billing-cycle usage allowance that scales by tier.

What is shipping: 5 AI features

This is not a single chatbot bolted onto an existing dashboard. Stockful’s AI launch includes five distinct features that work together across the entire app.

1. AI chat assistant at the top of the app. Ask questions in natural language, get streaming answers with the data behind them.

2. One-click health check that scans five dimensions of your store and surfaces 23+ specific findings you can fix with one click.

3. Explain button on every metric in the app. Not a tooltip with a generic definition, but a plain-English explanation of what this specific number means for this specific product.

4. Anomaly detection that surfaces unexpected changes (sales spikes, velocity drops, surprise stockouts) with a one-sentence hypothesis for why it happened.

5. Smart setup wizard that analyses your catalogue and sales history during onboarding and proposes configuration tuned to your store’s actual data.

Let’s walk through each one.

The AI chat assistant

The assistant lives in the top navigation of the Stockful app. You type a question, and it responds with a streaming answer that references your real products, real stock levels, and real sales data.

Behind the scenes, the assistant has access to over 20 internal tools it can call on your behalf: looking up variants, pulling sales history, checking inventory by location, generating forecasts, fetching reorder lists, suggesting transfer quantities, and more. You do not need to know which tool to use. You just ask in English, and the assistant decides how to get the answer.

Here is what that looks like in practice:

“What should I reorder this week?”

The assistant pulls your current reorder list, cross-references lead times and velocity for each product, and returns a ranked list with recommended quantities. The output is specific enough to copy into a supplier email or use as the basis for a purchase order in Shopify.

This replaces the workflow of opening the reorder report, reviewing each product, mentally accounting for lead times, and building a PO manually.

“Why did SKU ABC-123 spike on Tuesday?”

The assistant pulls the sales history for that variant, identifies the spike, and checks for adjacent signals: was it featured in a collection? Did a related product stock out, driving substitution demand? Was there a recent tag change suggesting a promotion?

The result is a one-sentence hypothesis like “Sales for ABC-123 jumped 340% on Tuesday, likely driven by the related variant DEF-456 going out of stock on Monday.” You can verify the hypothesis yourself, but the assistant does the detective work of correlating signals across your data.

“What’s at risk of stocking out in the next 14 days?”

The assistant cross-references projected stockout dates, lead times, and safety stock levels across all variants and locations. It returns a prioritised triage list: which products will stock out soonest, how many days of supply remain, and whether current lead times leave enough runway to reorder in time.

This turns a 30-minute report-digging session into a 10-second question.

“What should I exclude from tracking?”

This is especially useful during setup. The assistant analyses your catalogue and identifies SKUs that are gift cards, samples, archived products, or have erratic velocity patterns that are unlikely to forecast well. It returns a grouped list with rationale for each exclusion, and you can apply the suggestions with one click.

The assistant handles questions across six broad categories: reorder timing, diagnosis and root-cause analysis, risk surfacing, setup and configuration help, reporting shortcuts, and cross-location reasoning. The full AI documentation covers the complete range of what you can ask.

One-click health check

The health check is a different kind of AI feature. Instead of waiting for you to ask a question, it proactively scans your store across five dimensions and tells you what needs attention.

The five dimensions are:

Data quality. Are your cost-per-item fields filled in? Are there variants with missing or zero costs that will break your margin and valuation calculations?

Catalogue quality. Do you have duplicate SKUs, archived products still marked as active, or variants that should be excluded from tracking?

Tracking configuration. Are your safety stock levels, lead times, and reorder points set appropriately for each product’s velocity? Or are you using defaults that do not match reality?

Forecasting health. Are there products where the forecast is unreliable due to insufficient sales history, extreme variability, or recent major changes in demand?

Notification health. Are your Slack alerts and report schedules configured to catch problems before they become costly?

The scan surfaces 23+ specific findings, each with a severity rating (critical, warning, or info) and a one-click fix link that takes you directly to the setting or product that needs attention.

You can run the health check at any time from /settings/health-check. We recommend running it right after installation and then monthly to catch drift. Full details on each finding type are in the health check documentation.

Explain every metric

Every metric in Stockful, whether it is days of supply, velocity, ABC class, reorder point, projected stockout date, sell-through rate, or out-of-stock rate, now has an Explain button.

This is different from a tooltip. A tooltip tells you what “days of supply” means in general. The Explain button tells you what this product’s 8.3 days of supply means specifically: that at the current sales rate of 12 units per day, your remaining 100 units will last about 8 days, and given your supplier’s 14-day lead time, you are already past your reorder window.

The explanation is generated from the product’s actual data, not a canned definition. It connects the metric to its practical implication, so you understand not just the number but what to do about it.

This is useful for merchants still learning inventory management concepts, but equally useful for experienced operators who want a quick sanity check on a number that looks off. See the Explain feature docs for more.

Anomaly detection with hypotheses

Stockful’s anomaly detection runs continuously and surfaces unexpected changes on a dedicated /anomalies page: a sudden sales spike for a product that normally moves slowly, a velocity drop that could indicate a listing issue, an unexpected stockout that does not match the forecast.

What makes this more than a simple alert is the hypothesis. Each anomaly includes a one-sentence “likely cause” based on the surrounding data. For example: “Velocity for Winter Jacket (Black, M) dropped 65% over the past 7 days, likely due to seasonal demand decline (same pattern observed last year).”

The hypothesis is not a guarantee. It is a starting point for investigation, grounded in the data the assistant can see. Sometimes the real cause is something external that does not show up in your Shopify data. But having a data-backed hypothesis saves you from starting every investigation from scratch.

Anomaly alerts can optionally be sent to Slack on your schedule. You configure the channel and frequency in /settings/ai. Full details in the anomaly detection docs.

Smart setup wizard

Onboarding has always been the awkward part of inventory apps. You install the app, it syncs your products, and then it asks you to configure a list of settings you do not have answers for yet: What should your safety stock be? What lead times should you set? Which products should you exclude?

The smart setup wizard analyses your catalogue and sales history during onboarding and proposes answers for you. It examines product velocity to suggest appropriate low-stock thresholds, identifies SKUs that should be excluded from tracking, and sets initial configuration tuned to your store rather than generic defaults.

You review and approve everything before it is applied. The wizard proposes; you decide.

Pricing: included on every plan

Every AI feature described above is included on every Stockful plan: the $19.99/month Starter plan, the $39.99/month Growth plan, and the $99.99/month Pro plan.

Each plan includes a per-billing-cycle allowance of merchant-initiated AI interactions — chat turns and Explain clicks — sized to the plan. Starter includes 500 per cycle. Growth includes 1,000. Pro includes 2,500. The allowance resets when your Shopify billing period rolls over.

Background AI work — anomaly hypotheses, the weekly summary digest, smart setup analysis — does not count against your allowance. Only things you actively trigger do. If you reach 80% of your allowance, Stockful surfaces an upgrade nudge. If you hit 100%, AI chat pauses until the next billing cycle while everything else in Stockful keeps working.

The allowance scales with the plan because heavy AI usage has real underlying cost. Most stores will be well under the cap — but the structure means a one-person shop on Starter and a multi-location operator on Pro both get AI included, sized to how much they are likely to use it.

This is worth emphasising because it is unusual. Most apps that add AI features either charge extra for them or gate them behind higher tiers. Stockful’s position is that AI should make the existing product better for everyone, not create a new revenue line.

Trust and data

AI features that touch your business data need clear boundaries. Here is how Stockful handles it.

Opt-in by default. The weekly AI summary digest and anomaly Slack alerts are off by default. You choose whether to enable them, which Slack channel to use, and what day and time they fire, in your local timezone.

Master toggle. A single switch in /settings/ai disables every AI surface in the app: the chat assistant, the Explain buttons, the health check, and anomaly detection. If you prefer the traditional dashboard-only experience, one click turns AI off entirely.

Inference only. Your shop data is sent to frontier AI models for inference only. It is not used to train models. No data leaves your shop until you actively invoke an AI feature (ask a question, run a health check, click Explain).

Full details on data handling, retention, and the technical architecture are in the privacy and data documentation.

How it compares

The AI landscape in Shopify inventory apps is still early. Here is how the current options stack up:

Shopify Sidekick is Shopify’s built-in AI assistant. It can answer general questions about your store, but its inventory capabilities are shallow. Community feedback is mixed: merchants report that Sidekick sometimes claims it can run inventory reports but produces empty results. It is a general store management assistant, not an inventory specialist.

Prediko’s Pia is the closest competitor to Stockful’s AI assistant. Pia is a genuine conversational agent that can surface stockout risks, recommend reorders, and let you dig into data points through chat. It is well-integrated with Prediko’s forecasting and purchase order system. If your primary need is AI-assisted purchasing and supply planning, Pia is strong. Prediko starts at $119/month.

Stockful’s AI covers a broader surface area than chat alone: the health check, Explain buttons, anomaly detection with hypotheses, and the smart setup wizard are all unique to Stockful. The focus is on understanding and diagnosing your inventory, not just planning purchases. And every AI feature is included on every plan, with a per-cycle usage allowance that scales by tier, starting at $19.99/month with a 14-day free trial.

Inventory Planner by Sage and Stocky (sunsetting August 2026) do not offer AI chat, health checks, or anomaly detection.

If you need AI-powered purchasing with PO automation, Prediko is worth evaluating. If you need AI-powered inventory intelligence and diagnostics at a lower price point, Stockful is built for that.

Getting started

Three steps to your first AI-powered inventory answer:

Step 1: Install Stockful (or open the app if you already have it). The AI chat panel is in the top navigation bar.

Step 2: Run the health check. Go to /settings/health-check and click “Run scan.” Review the findings and fix anything marked critical. This ensures the AI has clean data to work with.

Step 3: Ask the assistant something. Try “What should I reorder this week?” and confirm the answer references your real variants, with actual quantities and lead times.

You will know within 60 seconds whether the AI is working with your data or generating generic advice. It should be your data.

Your inventory data already has the answers

Your shop already contains enough data to answer most of the inventory questions you have. The problem has never been a lack of data. It has been a lack of time and tooling to extract the answers.

The AI assistant, health check, Explain buttons, anomaly detection, and setup wizard are five ways of closing that gap. Each one takes data you already have and turns it into something you can act on.

Every feature is live now. Every feature is included on every plan.

Start your free 14-day trial at stockful.app. Explore the full AI feature set at stockful.app/ai.