For more than fifteen years I ran Telecamera, an online retailer of security appliances that I started from nothing and grew to three offices in two cities, serving thousands of customers a month. The market was crowded and unforgiving. We survived it by being good at one thing in particular: taking care of the people who called us.
My sales team was never more than ten people, and they did not make cold calls. Their work was the work my customers do today: answering incoming leads, and reaching out to the clients who were going quiet. Roughly half business, half consumer, in a market where the customer could always find the same product somewhere else.
What I could measure
I ran that business on numbers. Every rep had a dashboard of their own: workload, hours, orders, marginal profit, customer satisfaction, task completion. They knew I was looking at the same figures they were, which meant I almost never had to look over anyone’s shoulder. People knew what was expected, where they were falling short, and where they were doing well. That alone took a large part of managing off my desk.
We reviewed the quality of one hundred percent of orders. It was expensive, and it was the reason we held an NPS of around 0.85 for fifteen years in a market that punished every mistake. I already believed in full coverage. I had paid for it.
What I could not
None of it told me what happened inside the conversations.
Coaching was the clearest symptom. To prepare a session I would pick call recordings by hand and write notes: open one, find nothing worth discussing, open another, find something, note it down, open a third and hear the same mistake again. Then the whole team would sit through feedback that was relevant to two of them. I also listened to calls live and gave people feedback on the spot, which worked, and did not scale, and surfaced the same errors over and over.
The mistakes repeated because I had no way of knowing which ones actually mattered across the team. Leads not qualified. Needs never established. Calls ending without an agreed next step. I must have said those three things a thousand times. Saying them again was not the answer. What I needed was to show each rep their own numbers and let them correct themselves, the way the dashboards already did for everything else.
Reactivation had the same shape. We knew a customer was slipping when their RFM score moved, and by then the revenue from their quiet months was already gone. Every customer has their own natural rhythm of ordering, so the guessing cut both ways: sometimes we chased people who were simply not due yet, sometimes we noticed far too late. The signal we needed was in the calls, weeks before it reached the transactions.
Lagging and diagnostic
The gap took me years to name properly. Every number I had was lagging. Revenue, RFM, satisfaction, conversion: each one an accurate report of something that had already finished happening. What a manager actually needs is diagnostic. Not that the quarter went badly, but which step is being skipped, which objection is not being handled, which relationship is cooling, while there is still time to do something about it.
I first understood that properly during an executive MBA, in of all places a course on numerical methods. The idea that you could predict an outcome and quantify your confidence in it reorganised how I thought about running a business. It is the thread that led me out of retail and into analytics, and eventually into building this.
What Sales Signals is
It is the tool I could not buy. It reviews every call, the way we reviewed every order, and turns what is said into the diagnostic measures I never had: which steps are being skipped, which objections recur, which clients are cooling. It gives reps their own numbers, so they can correct themselves without being told the same thing for the hundredth time. And it does the preparation that used to eat my evenings, so coaching starts from evidence instead of a stack of recordings.
One thing it deliberately does not do. I am in favour of automating almost everything: if a customer wants to order online, that is their choice and we should make it effortless. But when someone dials your number, they expect a person on the other end. Sales Signals does not answer your calls. It makes the people who do better at it.