The challenge
A web hosting company was recording every sales lead and support query by hand, and its staff were losing hours each week to it. Each conversation meant collecting the same details: name, email, hosting needs, traffic volume and technology requirements. Staff asked the same questions again and again, and when they were busy, details got missed.
Every enquiry arrived through the same channel. A prospect asking about a dedicated server and a customer reporting a problem landed in one stream, which created confusion between sales and support. Sales opportunities were delayed, and support tickets sometimes went missing.
The back office was manual too. Customer data and tickets were synced into WHMCS by hand, which caused delays and errors.
Recommending the right plan was harder than it sounds. Requirements like storage above 50GB, traffic above 100K, or .NET versus Linux each change the answer, and staff found it difficult to match those combinations to the right product. The advice a visitor got depended on who replied.
Response time was the symptom customers saw. Replies were delayed, often by hours. The client wanted immediate support with no wait time.
What we built
Amatec built a chatbot in Zoho SalesIQ that greets every website visitor, collects structured details, recommends a hosting plan or opens a support ticket, and syncs the result to WHMCS. It was designed in Flowbuilder, the visual flow tool inside SalesIQ, as a chain of conversational cards.
The conversation runs in four stages:
- Greeting. Opening cards introduce the bot and ask what the visitor needs.
- Data capture. Buttons and predefined options collect name, email, hosting needs, tech stack and expected traffic as structured fields, so nobody retypes them later.
- Routing. A decision tree splits the conversation. Sales questions go to plan recommendations. Complaints go to ticket creation.
- Action. Each branch ends in an answer or an API call. Custom scripts call WHMCS to create a support ticket with the full details and notify the right department.
The recommendation logic took the most care. Staff had struggled to match storage, traffic and technology to a product, so we wrote conditional logic that weighs all three together and suggests the best plan on the spot, for example a dedicated server or VPS hosting. The suggestion is logged in the CRM along with the lead.
A few more pieces complete the build:
- Leads are captured automatically and synced to WHMCS through its API, so there is no manual lead entry.
- Natural language processing lets the bot interpret what visitors type as well as the buttons they click.
- Additional cards add security measures and access control to the flow.
The bot went live as a fully working feature on the client’s website, ready for immediate use.
The results
The chatbot saved the hosting company’s staff 10+ hours weekly on repetitive data entry and replaced replies that often took hours with instant ones.
- Staff saved 10+ hours weekly on repetitive data entry
- Response time went from delayed, often hours, to instant replies with guided flows
- Support tickets are created instantly in WHMCS with full details, so complaints are no longer missed or delayed
- Immediate plan recommendations reduced drop-offs, and faster lead qualification improved conversions
- The bot handles unlimited simultaneous visitors
Sales and support now have separate lanes. A prospect gets a plan suggestion while they are still on the page. A customer with a problem gets a ticket with every detail already filled in, and the right team hears about it straight away.
Plan advice is consistent. The same storage, traffic and technology inputs produce the same recommendation every time, whoever would have answered the chat.
The hours saved went to better use. The team moved its focus from admin to growth work, and the flow framework can be extended to new workflows as the client needs them.