How Kees Smit is building one connected AI stack for the entire customer journey

With 3,000 to 4,000 SKUs and a short, intense selling season, Kees Smit Tuinmeubelen gets a remarkable amount of work done in very little time every year. Thousands of products have to go live within a few weeks, customers expect detailed advice, and support questions spike the moment the season starts.

There is an AI tool for every one of those challenges. But the real gain appears when those tools are connected.

That gain shows up in two places. On the back end in time: content, imagery and service take far less work. On the front end in conversion: smart widgets help visitors at exactly the moment they make a choice.

That is why Kees Smit is building one connected AI stack together with Sero, with Sero CDP at its heart. Sero CDP brings customer and behavioral data together in individual profiles and makes that data usable across the rest of the customer journey. The other Sero products plug into it: from content production and product advice to personalization and service.

One unified customer profile as the foundation

A click on a product, an add-to-cart, a conversation with the chatbot. In a traditional tech stack, all of those signals live in separate systems and dashboards.

Sero CDP brings customer data together in a single profile. Click behavior, purchases and other customer details are connected, after which AI can surface patterns and promising segments. Those insights can then be activated directly on the website, in email, in advertising and in the Sero products themselves.

That makes Sero CDP the foundation the rest of the stack plugs into. Each application delivers value on its own and shares data with the others so they can build on each other’s signals.

A conversation with Sero Shopping Assistant, for example, can reveal new information about what a visitor is looking for. That information enriches the customer profile in the CDP and can then be used for more relevant personalization in other channels. Sero CDP spots the signal, the rest of the stack acts on it.

Thousands of product descriptions ready before the season starts

A connected stack starts with solid product data and content. Certainly for a seasonal business like Kees Smit, where so much work lands in such a short period.

New product information from manufacturers and suppliers often only arrives in February or March, while the selling season already begins at the end of March. The consequence: thousands of product descriptions have to be written in a matter of weeks. In the past, even the customer service team pitched in during the low season to absorb that peak. And whatever was not finished before the season started risked being left behind during the busiest months.

With Sero Feeds, Kees Smit automates a large part of that process. Feeds uses AI to optimize product titles and descriptions based on product data, best practices and predefined instructions.

“With our own prompts we add specific knowledge per subcategory and brand. That way we stay in control of how the content is structured and what it says.”

Jan Karel Ekkel, E-commerce & Performance Marketing Manager at Kees Smit

The time needed for product descriptions has dropped by roughly 80% as a result. According to Jan Karel, that is the equivalent of 2 to 3 FTE of work, year after year.

One photo shoot, endless product imagery

A similar challenge played out in product photography. The in-house photo studio at Kees Smit used to be set up again and again for different product combinations. With Sero Productshot, the team generates new imagery and variations from a smaller set of its own product photos. The photo studio remains the foundation, AI scales the output.

That removes a large part of the physical work around every individual product variant. At the same time, Kees Smit stays in control. Clear guidelines for proportions, product specifications and visual consistency safeguard quality. Every product image stays accurate and realistic, while the team gets a complete catalog online far faster.

Personal sales advice, now online too

Producing product information and content efficiently is one thing. Helping visitors pick the right option out of thousands of products is the next step.

Kees Smit previously used a chatbot plus a separate tool for customer advice. Sero Shopping Assistant replaces both. It answers buying questions, compares products and recommends from the company’s own catalog. Visitors describe what they are looking for in plain language. The assistant asks follow-up questions to understand what a customer really needs, then presents a fitting selection with an explanation. That brings the personal advice Kees Smit has offered in its stores since 1942 steadily closer to the digital experience.

“Sero Shopping Assistant genuinely takes on the role of an advisor. It now feels much more like talking to an employee in the store.”

Jan Karel Ekkel, E-commerce & Performance Marketing Manager at Kees Smit

Sero Shopping Assistant is fed with existing knowledge built up over many years. AI helps to add logical follow-up questions and matching answers on top of it. Existing expertise finally becomes scalable.

And with Sero CDP, that interaction gains a second purpose. The information a visitor shares during such a conversation is used to enrich the customer profile further.

An answer at the decision moment

The biggest leap came when Shopping Assistant also became available directly on the product page. At the request of Kees Smit, Sero developed a PDP widget that is now a standard part of the product. Through that widget, visitors can ask questions about the item they are looking at right then. Exactly where those questions arise, without digging through all the product information or opening a general chat.

That shift matters commercially. Doubt about dimensions, material, maintenance or delivery time almost always surfaces on the product page, right before the moment of ordering. An answer that appears there immediately does not have to send a visitor off to an FAQ, an email or a phone call. Kees Smit sees this reflected in its own reporting: visitors who shop on through the assistant convert better and generate clearly more revenue.

And because so many questions are answered right on the product page, the team has more room for customers who prefer to call. Even during peak moments, they are helped faster.

Visualizing it in your own backyard

When buying garden furniture, it helps enormously to see beforehand how a product will look in your own garden, on your balcony or on your patio. Kees Smit already offered an augmented reality solution for this. But to make it work, separate 3D models had to be created for every product and variant. A time-consuming and costly process.

Sero Try Before You Buy makes that far simpler. The visitor uses a photo of their own space and points to where the product should go. Sero then places the chosen product into that situation, straight from the product page and without a separate app.

“We see this as a valuable extension of our service. Customers get more help while they orient themselves and can judge an expensive purchase properly in their own setting.”

Jan Karel Ekkel, E-commerce & Performance Marketing Manager at Kees Smit

Together, the assistant on the product page and Try Before You Buy form the layer where orientation turns into a decision. Both remove a concrete barrier in exactly that spot: one answers the question that is left, the other shows how it works out at home.

Faster to a fitting resolution

In aftersales too, Kees Smit uses AI to handle customer questions quickly and consistently.

Working from Sero Studio, Kees Smit developed a solution around its own complaints matrix. Based on the type of complaint, the applicable rules and the customer’s situation, among other factors, a fitting resolution is proposed. The customer service agent then quickly selects the right way to handle it.

Every complaint takes less digging, and comparable situations are handled more consistently. For the customer, that means clarity sooner and a resolution held to the same quality standard.

A stack that grows with you

Kees Smit did not start with a complete AI stack all at once. Individual Sero products were added step by step around concrete questions in the organization. With Sero CDP, the foundation that connects everything is now falling into place.

The products can be used independently, but they become far more powerful as part of the same stack. Data from one application becomes input for the next. Insights can be activated immediately. And every new application builds on the same picture of the customer.

Kees Smit did not have to bring all of the required AI expertise in-house to get there. Sero guided the team through the implementation, setup and rollout of the various products, and stayed involved in developing those applications further.

That is how the stack keeps pace with the ambitions of Kees Smit. Not only by automating repetitive work, but by connecting data, insights and customer interactions ever more intelligently. The result: more room to test and improve, and a customer who benefits from it directly.

“With Sero I always have a sparring partner beside me to test, tweak and optimize. On top of that, we work faster and the quality is higher.”

Jan Karel Ekkel, E-commerce & Performance Marketing Manager at Kees Smit