The silent switch

Andrés lives in Bogotá, pays a monthly plan and works from home twice a week. Since March his calls drop when he walks into the kitchen. In May he opens his operator’s app four times in one week to check how much data he has left. In June he calls about a charge he does not recognise, waits on hold, and gets an answer he only half believes. In July he walks into a competitor’s shop and leaves with his own number on a new SIM. His operator learns he has gone from the porting request. Andrés is a composite, drawn from customers like him.

To the operator, that was a silent switch. It was not silent. Andrés said it four times over four months: in the network logs, in the app’s analytics, in the billing queue and in the contact centre. Each of those belongs to a different team, measured on its own numbers. Nobody had the four on one page, with his line at the top.

Colombia shows how much moves this way. In the first half of 2025, Colombians made 3.34 million portings, according to the CRC, the communications regulator. In the second quarter of 2026 there were 1.29 million, 21.4% fewer than a year earlier. The net results were far apart: Tigo lost 259,100 lines through porting, Claro gained 207,700, and the virtual operators together gained about 17,000.

Porting is also only the departure that arrives with a form. In the first quarter of 2026 it accounted for 17.3% of all the lines the network operators retired, the CRC reports. The rest were retired for other reasons, without a porting request. An operator that counts churn from porting requests counts fewer than one in five of the lines it loses.

The regulator has also changed the end of the story. Under CRC Resolution 8171 of 2026, since 1 September a customer who ports must wait more than 30 days before porting back to the operator they left. A counter-offer the week after the switch can no longer bring the number straight back. By the time it can, the customer has a new SIM, a new bill and a month of habit. Whatever the rule was written for, it moves the operator’s chance to the weeks before the request.

Those weeks are not empty. The signals before Andrés’s switch were ordinary ones, and his operator held every one of them. Calls that drop at the same address, day after day. Repeated checks of the data balance, a sign that the plan may no longer fit. A question about the bill, and then a second one. A complaint answered but not resolved. None of these on its own means a customer is leaving. Together, on one line, within a few weeks, they make a pattern that is hard to miss.

So why is it missed? Rarely for lack of data. Network quality is the engineers’ measure, app usage belongs to the digital team, bills to finance, calls to customer care. Each team reports its own average, and an average hides the one customer whose bad month touched all four. The data exists. The customer journey that would connect it does not.

Joining it takes less than it sounds. Start from the line, not the department: one view per customer of network events, app sessions, billing contacts and complaints over the last 90 days. Look back at the customers who ported away and find the combinations that came before their request. Then decide who acts when a line shows them, with what, and within how many days. The answer is rarely a discount. For Andrés it would have been an engineer looking at his street, a plan with more data for the same money, and a straight answer about the charge.

Colombia’s figures say customers moved fewer numbers this year than last, and that operators are not losing them at the same rate. They do not say why. Each customer who left said why, in pieces, to different parts of the company. The switch is silent only to an operator that is not listening in one place.