They Had Data. They Didn't Have Answers.
The retailer was not short of data. Customer activity, product interactions, carts, orders, campaigns, and commercial performance were already being measured. Dashboards could show revenue, conversion, traffic, and average order value. Analysts could go deeper with filters and SQL.
But the question that mattered was rarely the first one a dashboard answered.
Moving from the first statement to the second meant opening several views, changing periods and segments, exporting data, asking an analyst, and reconciling different answers. Each answer tended to produce another question: Was it traffic or conversion? Which categories were responsible? Did shoppers lose interest, or did the product experience stop converting them?
The problem was not access to data. The answers were buried inside it, and finding them depended on someone knowing where to look.
An AI Analyst That Starts Working Before the Team Does
The e-commerce manager started the day with a Stormly briefing already in the inbox. It did not ask the team to inspect a page of KPIs or decide which dashboard deserved attention. It prioritized the developments that appeared meaningful and explained why.
One finding stood out: Beauty revenue had fallen even though traffic remained stable. The decline was concentrated in three brands, and the behavioral change appeared between the product detail page and add to cart rather than during checkout.
That was immediately more useful than a red revenue number. The briefing connected the commercial signal to the products and customer behavior behind it. It also gave the team a sharper place to begin: investigate the affected brands, their product pages, availability, pricing, and visitor mix.
Stormly had already completed the first analytical pass before anyone asked a question. The email did not finish the analysis. It made the next question obvious.
Beauty revenue fell while traffic remained stable
The change is concentrated in three brands and is primarily associated with weaker product-detail-to-cart conversion, not checkout abandonment.
The Email Raised Another Question. They Asked Claude.
The manager did not have to log into Stormly, construct a new report, or submit a request to the data team. The team was already working in Claude. Stormly made the retailer's analytics available there.
That is the practical value of Stormly's MCP connector: Claude becomes the conversational interface, while Stormly supplies the analytical intelligence and access to the retailer's actual project data. Each answer can lead directly to the next question.
Stormly tested traffic, product mix, conversion behavior, and the purchase journey, then returned the supported explanation: stable traffic, a brand-level concentration, and weaker movement from product detail to cart.
The analysis narrowed the issue to the brands that mattered instead of treating Beauty as one undifferentiated category.
Stormly routed the request to the appropriate e-commerce analysis, applied the period and audience comparison, and showed where the journeys diverged.
The investigation could now compare the retailer's product and behavioral evidence with relevant market signals before the team chose a response.
Stormly automatically chooses between a matching e-commerce report and custom SQL, depending on the question, runs the analysis against the selected project, and returns an interpreted answer to Claude. This is not an LLM guessing from a dashboard screenshot. It is a conversational route into the same analytical engine the team uses for rigorous reporting.
Customer Behavior Meets Market Context
Internal analytics can show that searches, product views, or conversion for a brand changed. The more valuable question is whether customers are losing interest in that brand generally, or whether they remain interested but are failing to buy it from this retailer.
What did customers actually do?
Stormly examines product views, searches, carts, orders, pricing, product mix, funnels, campaigns, and customer segments in the retailer's connected data.
Is demand changing more broadly?
Relevant search and shopping trends provide an external reference point, helping separate an on-site problem from a brand or category shift.
Reconsider inventory exposure, promotional priority, and the balance between brands in the category.
Investigate pricing, availability, product-page quality, sizing, proposition, UX, and competitive pressure.
Look at the product proposition and the information customers encounter before committing.
Focus on checkout friction, delivery expectations, payment behavior, and campaign-specific journeys.
The same revenue change can require four very different responses. Connecting commercial performance, customer behavior, and market context helps the team test the explanation before acting.
E-Commerce-Specific Analytics
The conversation is useful because the analytics beneath it understands commerce. Stormly works with SKUs and variants, products, brands, categories, carts, orders, revenue, campaigns, returns, funnels, repeat buyers, retention, and customer journeys.
The retailer can move from a question in Claude to a purpose-built analysis without translating the business problem into a generic chart specification. More than 50 ready e-commerce reports provide the analytical foundation, and custom reports are available when the business model calls for something specific.
Explore Stormly's ready reports to see the wider analytical library.
New Arrivals Performance
Which newly launched products are gaining traction in their first hours, days, or week, and which need merchandising support?
Cart Abandonment
Which SKUs, brands, or categories create strong cart intent but fail to convert, and where is revenue potentially recoverable?
Revenue Contribution
Which products, categories, and brands are genuinely driving revenue instead of merely appearing in successful orders?
Cross-Selling Analysis
Which products are bought together often enough to support a meaningful bundle, recommendation, or merchandising opportunity?
Root Cause & Conversion Journey
Why did revenue, orders, or conversion change, and where did behavior shift between product view, cart, checkout, and purchase?
Aha-Moment & Repeat Buyers
Which behaviors distinguish buyers from non-buyers, and what do returning customers do or purchase differently?
What Actually Changed for the Team
Analytics stopped being a destination the team had to remember to visit. It became an intelligence service that watched the business, brought forward the important changes, and stayed available for the next question.
- Something changes
- Someone notices
- Dashboards and filters are checked
- An analyst receives another request
- More questions create another report
- A meeting turns the analysis into a decision
- Stormly detects the change
- The AI briefing arrives
- A manager asks why in Claude
- Stormly analyzes actual behavior
- A follow-up question tests the explanation
- The team acts, with a specialist report when needed
From Looking for Insights to Acting on Them
The most important outcome was not another isolated conversion claim. It was a different operating rhythm for analytics.
Meaningful changes could be detected while they still mattered. Managers received an explanation rather than a KPI alert. Routine follow-up questions no longer had to begin in an analyst queue, and business users could explore richer questions without first deciding which report or SQL query to request.
- From reactive to proactive: Stormly surfaces changes before someone has to discover them manually.
- From dashboards to answers: conclusions arrive first, with the evidence available when the team needs to inspect it.
- From analyst queue to immediate exploration: Claude and MCP give managers a natural path into the retailer's actual data.
- From isolated KPIs to connected reasoning: products, customer behavior, funnels, commercial performance, and relevant market context can be considered together.
- From generic BI to e-commerce intelligence: the underlying analysis already understands the concepts the team uses to run the store.
The team spent less time assembling the analysis and more time deciding what to change. Stormly did not replace the retailer's data stack or the specialists responsible for it. It became the intelligence layer that made their work available sooner, more proactively, and through the tools the wider team already used.