01 · LBV · Telemarketing Coordinator · 2005 to 2012
Each agent's language
Before CRM, there was a room with more than forty agents and a headset on every head. I had come from sales and started coordinating people whose accents, histories and relationships with work were very different from mine. The dashboard showed talk time and number of acceptances. It explained little. I started paying attention to the words each person chose, the tone, the gestures they repeated while listening to the donor. For one agent, guidance became an open notebook holding their attention. For another, a recurring gesture became a listening anchor. With ongoing coaching and SPIN Selling, the operation gained about 25% in new donors and turnover fell by about 40%.
Leading starts with learning each person's language.
02 · Monchu · Communications & Branding Manager · 2016 to 2019
The party that learned to recognize who came back
A party that started with 300 to 500 people per edition reached an average of 2,000 to 3,000 in about two years. Part of that path ran through a spreadsheet. I built a score with frequency, average ticket and recency, and each tier got its own conversation: a combo for first timers who came with friends, early tickets for those who had disappeared, an annual membership plan for those who showed up every time. Each event in the portfolio called for a different mix. Sometimes one strategy, sometimes three. Research followed the same rhythm. On the first visits, I asked whether people had enjoyed it and would come back. After the fourth or fifth, I asked what drew them in and what bothered them. Those answers changed where restrooms and food stands went, whenever the space allowed. The audience started to feel noticed.
People who always come back deserve a deeper question.
03 · Monchu · Communications & Branding Manager · 2016 to 2019
The schedule the audience chose
Some parties took a long time to fill up and emptied early, leaving the bar working for an empty dance floor. The problem was not even in my area. I cross-referenced ticket validation times with the manual occupancy and exit log and saw how that audience arrived, stayed and left. The standard schedule for that music genre did not fit them. We adjusted the event to the real rhythm of their stay. Bar sales rose between 15% and 20%, and the new schedule came back in feedback as more comfortable.
The party's clock belongs to the audience.
04 · LBV · Strategic Partnerships Analyst · 2021 to 2022
Borrowed reach, owned audience
Back at LBV, I built a corporate partnership funnel from scratch, from prospecting to contract, and B2B sponsorships, in cash and in kind, funded the period's promotional projects. With influencers and celebrities, I designed co-branded campaigns that brought 15% more top-of-funnel acquisition, with no paid media, and turned borrowed reach into a segmentable audience. For those already inside, exclusive benefits funded by partners, with per-partner KPIs guiding each renewal.
A good partnership also needs a cadence.
05 · Fênix Educação · CRM & Lifecycle Marketing Specialist · 2022 to 2025
Same course, different moments
At Fênix, a request to improve the newsletter led to a broader diagnosis: a disorganized database and gaps in tracking and segmentation. I prioritized cleanup and led a four-email thematic flow across the entire base. Lead responses guided continued contact; 19% of the base was selected for CRM and nurturing. IT handled the platform API in parallel. With UX and the page team, I also proposed moving the pre-enrollment CTA earlier. The change was released, and we tracked growth in clicks, pre-enrollments and enrollments. I do not remember the percentages, and other changes happened during the same period.
The journey responds to the person's state.
06 · Fênix Educação · CRM & Lifecycle Marketing Specialist · 2022 to 2025
The test that changed my method
With a messy base in my hands and a strong urge to fix everything, I changed several things at once. Performance dropped and no one could say why. My coordinator asked for the list of what I had done. That made it clear: I had touched too many variables without isolating any. Since then I run CRM like a product backlog. Each test has one variable, one question and one reading criterion before moving on.
A test that does not explain the result is just another change.
