How a professional services firm scored and routed website enquiries instead of reading every one by hand

Enquiries arrived through several website forms and a handful of other channels. All of them landed in the same place and waited for someone in the marketing team to read the message, judge how serious it was and decide who should pick it up.

That review happened when someone had time for it. A prospect ready to appoint a firm sat in the same queue as a speculative question, and both waited the same length of time. Follow-up could take hours or days.

The firm already had Salesforce and Pardot, and both were underused. Leads sat in queues with nothing attached to distinguish a serious enquiry from a passing one, so the queue was worked in the order it arrived rather than in the order that mattered.

The cost showed up at both ends. Partners were pulled into low-priority questions an associate could have answered, and enquiries worth partner attention went cold while the backlog was cleared. Nobody could see which sources or forms actually produced work.

The brief was as much about who could maintain the system as about what it did on day one.

• A scoring model only the person who built it could explain.
• Longer forms that put enquirers off in order to feed the CRM.
• Routing rules buried in code that needed a developer to change.
• Automation that sent everything to partners to stay on the safe side.
• A second lead process running alongside the CRM instead of inside it.

The website forms were rebuilt around a small number of qualification questions covering company size, sector, budget range and timeline. The questions were chosen because the answers were worth scoring, and the form stayed short enough that people finished it.

Each submission passed into Salesforce and Pardot with those answers attached. A scoring model turned them into a numerical score and a priority band as the enquiry arrived, before anyone had read it.

The band decided the route. Routine enquiries went to associates, high-priority ones went to the partner who should see them, and escalation rules picked up anything that had not been actioned so no enquiry sat unanswered by default.

A scoring model set once and left alone drifts away from reality. Scores were reviewed against what the enquiries actually became, and the criteria were adjusted when the evidence said the weighting was wrong.

New form variants were added as the firm marketed different services, each feeding the same scoring and routing model rather than starting a separate process.

Training ran alongside the build rather than as a session at the end. The marketing team learned how a score was calculated, which fields fed it and what each priority band triggered.

By the end they could change the criteria, adjust routing, add a form for a new service and explain to a partner why a particular enquiry had been scored the way it was. The system stopped depending on the person who built it.

Enquiries were scored and routed as they arrived rather than when someone reached them, and partner time went to the enquiries that warranted it. The judgement calls stayed with people, but they were made against a score rather than a full inbox.

Because every enquiry carried a score, a source and an outcome in Salesforce, the firm could see which channels and forms produced real work for the first time. The reporting improved because the routing needed the data underneath it.