Three funnels, three constraints: worked examples
The diagnosis is easier to trust once you have watched it run on real-shaped numbers. Here are three operators, each convinced their problem lived somewhere it did not, and what the constraint read actually told them.
The founder who kept buying traffic
A solo B2B SaaS founder runs paid search and gets 4,000 visitors a month. Visitor-to-lead is 1.6% (above the 1.4% benchmark, so the front of the funnel is healthy), giving 64 leads. MQL-to-SQL sits at 11%, well under the 15% median, producing 7 opportunities. Win-rate is a respectable 28%, so 2 customers a month. His instinct was to double ad spend, which would have cost real money to lift the one stage already beating its benchmark. Run the constraint read and the answer is obvious: the 11% mid-funnel is the binding stage. He instruments his follow-up and finds the median response time to an inbound lead is four hours. He wires an agent to reply in under a minute and tightens his scoring to the two attributes that actually predict a close. MQL-to-SQL moves to 17%, opportunities jump from 7 to 11, customers from 2 to 3, a 50% lift in new customers for the price of a process change and zero extra traffic.
The agency that was winning and shrinking
A productised-service operator closes well: 22 new clients a quarter, win-rate above benchmark, pipeline full. Yet revenue is flat. The front-of-funnel dashboards all look green, so the constraint is invisible from where she is standing. Extend the view past the sale and it surfaces immediately: quarterly logo churn is running at 14%, so she is losing roughly 12 clients for every 22 she wins. She has been paying acquisition prices to replace revenue she already owned. The exploit move is not more leads; it is a structured onboarding moment in week one and an at-risk flag on any account that goes quiet for ten days. Churn falls to 8%, net adds nearly double, and not one euro of new acquisition spend was needed. The constraint was never demand. It was the leak.
The operator with no demand at all
A third founder has a beautiful product, fast follow-up, and a 35% win-rate on the handful of deals that reach him. The trouble is the handful: 300 visitors a month, 4 leads, 1 customer. Every mid- and back-of-funnel stage is above benchmark, which is precisely the trap, because it tempts him to keep polishing a sales motion that is already excellent. The constraint is raw demand at the very front. Here the honest move is to stop optimising conversion entirely this cycle and pour every agent at demand and lead generation, because no amount of mid-funnel craft matters when only four leads arrive. Same framework, opposite conclusion: the constraint read sent each of these three founders to a different stage, and in every case it sent them away from the work their instinct wanted to do.