The situation
The business is a mid-market B2B SaaS company of around forty people, selling a workflow product into operations teams at larger organisations. Inbound was healthy on paper. Between the website, a content programme, two paid channels and a steady trickle from partner referrals, around six hundred new leads landed in the CRM every month. On the surface that looked like a pipeline problem solved.
The reality inside the SDR team was different. Four SDRs split the inbound queue between them, and each one was spending roughly three hours a day reading through form submissions, LinkedIn enrichments, website behaviour and free-trial signups, trying to decide which contacts were worth a call and which were students, job hunters, competitors or one-person shops who had ticked the wrong box. Twelve hours a week, per rep, gone before a single dial. The head of sales had run the maths and it came to nearly two full working days a week of qualification work for every person who was supposed to be selling.
The worst part was not the lost time. It was what the lost time was costing them. Because the team was working through the queue in roughly the order it arrived, the genuinely high-intent leads, the operations directors at eight-hundred-person organisations who had read three pricing pages and downloaded the integration guide, were sitting unread for two or three days behind a wall of low-fit signups. By the time an SDR got to them, half had already booked a call with a competitor or gone cold. The team could feel it happening and could not get out from under the volume to fix it.
What we did
The starting point was their own data. The business had four years of closed-won and closed-lost history sitting in the CRM, and that was enough to build a qualification model that recognised the shape of a good fit for their specific product. Company size, industry, the role of the person filling in the form, the pages they had visited, the signals from the enrichment provider, the wording of any free-text fields. None of these mattered much on their own. Together they predicted intent with enough accuracy to be trusted with the triage job the SDRs had been doing by hand.
The model runs against every new lead the moment it hits the CRM. It scores, tags and routes in seconds. High-fit leads land directly on an SDR's calendar with a short brief explaining why they scored the way they did and what to focus on in the first call. Medium-fit leads go into a nurture sequence that warms them up over a few weeks. Poor-fit leads are politely declined or sent self-serve resources, with no human time spent. The SDRs see only the conversations worth having.
We were deliberate about which calls stayed human. Anything involving a named enterprise account, an existing customer expansion, a partner-sourced lead or a request for a custom demo bypasses the model entirely and goes straight to a person. The model is there to absorb volume, not to replace judgement on the deals that matter most. The head of sales reviews the model's decisions weekly and feeds corrections back in, so it keeps getting sharper as the product and the market move.
The result
Three months in, the numbers were straightforward. Manual triage time dropped from twelve hours a week per rep to under thirty minutes. Conversion from qualified lead to first meeting went up by three and a half times, mostly because the right leads were now being reached within hours instead of days. Pipeline value attributed to inbound roughly doubled over the quarter, and the cost of the project paid itself back three and a half times over inside the ninety-day window.
The honest part is what happened to the time the SDRs got back. It did not all turn into extra dials. Some of it did, and the team is having more first conversations than ever, but a real chunk of the recovered hours has gone into things they never had room for before. Following up properly with prospects who said no six months ago. Writing better outbound sequences for the named accounts they actually want to win. Sitting in on demos with the AEs to learn the product more deeply. The head of sales describes it as the team finally doing the job they were hired to do, rather than working as a human filter for the inbox.