AI configurator for a technical category: the Haier AC case study
How we built the configurator + savings calculator on haier-ac.ro: 350+ users in month 2, 3-5 minute sessions, less pressure on the call center. What worked and what's replicable.
350+
configurator users in month 2
3-5 min
average length of a configuration session
↓ call center
less pressure: customers figure it out themselves
The challenge: a technical category for an uncertain customer
Air conditioning is a purchase most people make rarely and without knowing the criteria. How many BTU for a 20 m² room? Inverter or not? Which energy class is worth it? For a typical customer, those questions are a wall. And a site that shows them 40 models and a set of technical filters doesn't solve the problem — it dumps it on their shoulders.
The result, without a guidance system: either the customer leaves “to think about it,” or calls the call center with basic questions, loading the team. Both are costs. Our goal for haier-ac.ro was to turn this confusion into an assisted decision, right on the site.
The solution: configurator + savings calculator
We built an AI configurator that starts from the customer's need, not from specifications. Simple questions — what room size, what kind of usage, what conditions — translated internally into technical criteria and matched with the right units. On top of the configurator, a savings calculator that shows what the customer gains over the long run, shifting the conversation from price to value.
The two layers answer the two questions that stall the decision: “what should I choose?” (the configurator) and “why is it worth it?” (the calculator). Together, they do what a good salesperson would — available around the clock, with no pressure.
Results and what they mean
In month 2, over 350 users ran the configurator, with 3-5 minute sessions. That range is the real signal: people don't spend 3-5 minutes on a filter — they abandon it in 20 seconds. They spend it on a conversation that helps them. The configurator produced real engagement, not traffic passing through.
The side effect mattered just as much: less pressure on the call center. The basic questions — the ones the team used to field over the phone — were absorbed by the configurator, and the customers who did reach sales were already clear on what they wanted. The configurator didn't replace the team; it freed them up, leaving people for what matters: negotiation, special cases, closing.
The pattern applies to any technical category
- 01
Need questions, not spec questions
You start from what the customer understands, not from technical jargon. The pattern is identical whatever the product.
- 02
Recommendation on product data
The matching logic depends on clean data. What differs between categories is the data, not the mechanism.
- 03
Value calculator
Savings, ROI, total cost — every technical category has a value to demonstrate, not just a price.
Frequently asked questions
What does the Haier AC configurator do?
It guides the customer toward the right air conditioning unit, starting from simple questions about their need (room size, usage, conditions) rather than from technical specs. Then a savings calculator shows what they gain over the long run. Instead of leaving the customer to wrestle with BTU and energy classes, the configurator takes on the burden of the decision and turns it into a guided conversation.
What results did it produce?
In month 2, over 350 users ran the configurator, with 3-5 minute sessions each — the length of a real advisory conversation, not a few-second filtering pass. The important side effect: less pressure on the call center, because people figure things out on the site and reach the sales team better prepared.
Why is air conditioning a good category for a configurator?
Because it's a technical category where the typical customer doesn't know the criteria. Nobody wants to become a BTU expert just to cool their bedroom. There are many interdependent variables (room size, insulation, sun exposure, usage), and the wrong choice costs — either too little capacity or too much. Exactly the context where guidance beats filter-based selection.
Which part mattered most — the configurator or the calculator?
They worked together. The configurator answers “what should I choose?”, and the savings calculator answers “why is it worth it?”. The calculator was central to keeping customers engaged for 3-5 minutes, because it shifts the conversation from price to value — the customer sees what they save, not just what they pay. Without the calculator, the configurator would have solved only half the decision.
How much of this model is replicable for another category?
Almost all of it. The pattern — simple questions instead of specs, recommendation logic built on product data, a value calculator, a bridge to action — applies to any technical category with a complex decision: equipment, B2B solutions, products with many variables. What differs is the specific questions and the product data. The framework stays the same.
Conclusions
Haier AC shows what it means to treat a technical category with respect for the customer: you don't ask them to become an expert, you guide them. The configurator plus the savings calculator turned a stalling decision into a 3-5 minute conversation that ends with an informed choice — and freed up the call center at the same time.
And the pattern is replicable: any category where the customer doesn't know the criteria and the decision has stakes can benefit from the same model. The question for your business: do your customers in the technical category leave uncertain, or are they guided toward a decision?
Dan Cristian Alexandrescu is the founder of Websem, an agency that builds AI platforms and systems for serious business. The Websem team delivered the configurator + savings calculator for Haier AC Romania and other AI tools for categories with a technical decision.
Want a configurator like Haier's for your category?
30 minutes to see whether the pattern applies to your products — plus 3 concrete actions. No strings attached.