I have opened a market from a single desk, inherited a region that had lost its way, and stepped into a fast-scaling one on an interim basis. Three companies, three very different products, three different motions. The lesson underneath all of them was the same, and it is the one most founders learn a year too late: revenue follows evidence of belief, not effort.
Here is what each one taught me.
Protegrity: regulated buyers move on proof, not pitch
In 2018 I opened Protegrity's first Asia Pacific office, in Singapore, as General Manager for the region. The product was enterprise data-security: tokenisation and encryption for the world's largest banks, insurers and healthcare providers, sold into compliance-bound environments. You do not talk a risk officer at a bank into that. Nothing moves on narrative. It moves when a peer institution has already done it, when the security team has watched the architecture survive their own scrutiny, when the regulatory box is demonstrably ticked (as reported at the time by SecurityBrief Asia). Building the region from one person meant I could not afford to chase everyone. The only accounts worth my week were the ones where belief could spread: a reference customer whose adoption would give the next three permission to move.
In any considered purchase, your pipeline is not a function of how many people you contacted. It is a function of where proof is compounding. Concentrate there.
DataRobot: a turnaround is a demand problem wearing an activity costume
Later, as Regional Vice President for Asia at DataRobot, I inherited a region that needed to be reset. The instinct in a struggling territory is always to do more: more outbound, more meetings, more pipeline reviews. But activity poured into the wrong motion does not fix the motion. It drains the tank faster and hides the leak behind a bigger number. The recovery did not come from volume. It came from being honest about which segments were truly converting belief into commitment, and moving the team's scarce time toward them.
What told us the truth was engagement, not assertion. DataRobot was an evaluation-led enterprise platform, so the buyer left a behavioural trail: whether the trial kept being used after the first week, whether new colleagues from the account showed up unprompted, whether the team ran their own data through it rather than the sample, how fast they came back with the next question. Those generalised engagement signals predicted outcomes far better than anything a champion said on a call, or anything sitting hopeful in the forecast. So we closed the loop: read what buyers did, feed that evidence back into where the team spent its next hour, watch what happened, and adjust again. That feedback loop is how you get the time back. Every hour reclaimed from a deal the behaviour had already written off went to one where belief was truly building. You stop selling to what people say, and to what you hope, and start responding to what they do.
"[Out of all the VPs,] you were most on top of all the forecast updates."
When growth stalls, resist the urge to add activity. Diagnose the decision, not the diary. Read what buyers do, feed it back into where your hours go, and most "we need more leads" problems turn out to be "we are working the wrong ones" problems.
Grafana: I modelled my own team, and capacity was the hidden leak
On an interim basis I led APAC go-to-market for Grafana, a leading, late-stage observability company. Grafana grows bottoms-up: engineers adopt the open-source software long before anyone speaks to sales, so the demand signal already exists, written in usage. The temptation is to read that as "so just go and close it." I wanted to know where my team's time was really going, so I modelled our sales capacity against our stage-by-stage conversion. The result was uncomfortable: roughly 74% of the team's total selling capacity was being spent on activities with about a 50/50 chance of ever producing revenue. Not because anyone was lazy, but because we let too many opportunities through the first gate, and every one of them drew down the same finite pool of hours.
The fix was not more activity. It was qualifying harder at the very first stage. John McMahon, the architect of the MEDDIC methodology, puts it bluntly: it is almost impossible to hit revenue numbers repeatedly without a voracious qualification process. When I re-ran the model with tighter qualification at stage one, the same team, working the same hours, would have produced well over half as much revenue again. That uplift was not new headcount or new leads. It was capacity being handed back, reclaimed from deals that were never going to close.
A second finding was hiding in the timing data. The opportunities that were going to advance mostly did so quickly: around four in five deals that reached the next stage got there within roughly twenty days of entering the one before. The ones that sat for months rarely converted. The behaviour told the truth long before the forecast did.
Your constraint is never leads. It is where your finite hours are spent. Qualify ruthlessly at the top, read the behaviour rather than the hope, and protect capacity for the few opportunities where belief is truly moving.
The throughline
Three products, three motions, one pattern. In every case the deals that closed were the ones where belief was already spreading inside the buyer's world, and my job was to find that belief, feed it proof, and get out of its way. This is not a soft observation. It is how contagion works: adoption travels through networks of trust, not through the volume of messages sent, something Christakis and Fowler have spent two decades demonstrating in stakes far higher than software. And it is why forecasts built on what sellers report are so unreliable. We see what is in front of us and mistake it for the whole picture; Kahneman named the trap, "what you see is all there is." A pipeline is a hall of mirrors made of exactly that.
Why this matters before you hire
Here is the part founders miss. In the early days, you are the best demand sensor your company will ever have. You feel belief spreading, or you feel it stall, in your gut, on the call, in the thread that suddenly goes quiet. That instinct is real and it is valuable. The mistake is not trusting it. The mistake is leaving it as the only place that sense lives. When the founder is the sole instrument, the company can only detect as much demand as one person has hours to feel. That is the ceiling. Founder-led sales is the strongest early engine there is, right up until it becomes the only one.
What we do at Celerio is take that instinct, the one you have already proven works, and build it into a system that reads the same signals at a scale you cannot personally reach: what buyers do, weighed as evidence, rolled into a forecast you can take to a board. You keep the judgement. The production stops depending on your hours.
Proof beats narrative. It did at Protegrity, at DataRobot, and at Grafana. It will at yours.