In every case, the obvious problem was not the real one.

The job

Sales associates at our partner stores needed to sign up for a global tech brand's sales training program and then pass its product quiz. When I started, about a quarter of the people in scope had signed up.

What was found

The obvious fix was more reminders. Going through the sign up path one step at a time showed where people actually got stuck, such as the steps that asked for a personal email and a separate program sign up. Later, some people who had passed the quiz were missing from the completion report.

What was done

I rebuilt the sign up guide around those steps and sent reminders by stage, so each person got the one step they were missing. For the quiz, I ran a tracker for each person, collected screen recordings as proof of passing, and followed up with trainers and district sales managers. I also showed the missing passes to the team that runs the report, and the report was corrected.

What happened

Sign ups went from 26% to 51% in one week. Quiz completion went from 19% to 94% in 10 days.

The honest note

Part of the rise to 94% came from counting people who had already passed, not only from new passes. That correction is also why the final number can be trusted.

The job

Sales staff at our partner stores promote our financing offers on their own social media. Each retail partner had its own poster template, so only staff from 13 partners had one to post.

What was found

The obvious fix was to make a template for each new partner. That work grows with every partner and every monthly offer, so most partners would always be waiting for theirs.

What was done

I replaced the 13 templates with one. Each sales associate adds their own store logo and details, and a short demo video shows how in about a minute.

What happened

Sales staff from all 52 active partners could post, not only 13. Use of the toolkit rose for the first time since February.

The honest note

Being able to post is not the same as posting. The next number to watch is how many stores post each month.

The job

Run Home Credit's first presence at a motorcycle expo, a three day trade show in Cebu, from a booth that could display motorcycles but not sell them on the spot.

What was found

I picked the project up midway with no handover. Before the event, I found that the existing enquiry form sent leads to a system that does not process motorcycle loans, so none of the booth leads would have reached the team that handles them.

What was done

I rebuilt the plan that four agencies bid against, set the booth and print specifications, and got a dedicated lead capture form live before day one.

What happened

The booth captured 456 financing leads from 11,510 visitors.

The honest note

These are leads captured at the booth, not approved loans, and I report them that way.

The job

Launch Home Credit's first social commerce collaboration ad with a retail partner, a format the company had not run before.

What was found

The ad itself was not the slow part. The new format had never gone through operations clearance, so the people approving it needed to see exactly how it works.

What was done

I wrote the operations explainer for the new format, then turned the launch into an end to end playbook with 10 stages and a 16 point launch checklist.

What happened

The first ad went from key visual to live in four days. The same setup was then repeated with two more partners, and the second repeat took two days.

The job

A retail partner kept having products rejected from our online marketplace, 42 of them.

What was found

The obvious fix was to resubmit. Checking the listings against the manufacturers' sites and the partner's own website showed that the partner's own site contradicted what it was sending us.

What was done

I traced the problem back to the source data instead of the submission process, and had it fixed there.

What happened

All 42 products were relisted and went live.

The job

The school needed enrollment inquiries for the school year 2026 to 2027 intake. I had three weeks of production, from the first post to the final report.

What was found

Parents were not dropping off from a lack of interest. Reading the full inbox showed conversations going quiet at one point: right after tuition and fees were shared, especially when a family asked about scholarships or financial aid and there were none to point to. That is an affordability barrier, not an awareness problem, and it changed what the school needed to fix next.

What was done

I made 23 pieces in three weeks: 16 posts, 3 videos and the enrollment posters, plus reply templates for the inbox. Targeting followed the school's actual bus service area instead of a radius around the map pin, and stayed narrow even when Meta pushed to widen it. The boosts ran through Ads Manager instead of the Boost button, which kept the campaign objectives and the organic engagement on the posts. The kickoff primer, content calendar, approval deck, mid campaign report and final report all landed on schedule.

What happened

The campaign reached 105,841 unique people from about 186,000 impressions. 127 parent inquiries arrived in Messenger, and 77 inquiry threads were compiled and handed over for follow up. Of the six boosted pieces, the enrollment poster brought in the most conversations at the lowest cost.

The honest note

The inquiry count could have been reported as 106% of the goal. Inquiries are not enrollments, so the report was reissued as a reach and inquiry report with the drop off point marked. The headline number is smaller, and the school board can trust it.