In online learning, 95 people out of a 100 quit. A network of health workers from malaria-endemic countries spent one summer proving that this is a choice, not a law – and they did it with WhatsApp, phone calls, and one stolen handset.

The usual explanation for online learning’s completion rate is the learner: too busy, too poor a connection, not motivated enough. (There is much more to this story.)

This is an account of twelve consecutive Wednesdays in the summer of 2026 when a group of nurses, district physicians, nutritionists and community facilitators – scattered across 16 countries, mostly strangers to each other – tested a different explanation. Their theory was that people do not drop out because they lack motivation. They drop out because nobody is standing next to them. What follows is what they did about it, what it cost them, where they failed, and one line written by a man in eastern Cameroon that tells you whether any of it was worth doing.

Learn more: Honor Roll: 189 TGLF Ambassadors are inventing new ways to turn the tide against malaria

Seventy people were on the other side of that phone

The course opened on a Wednesday.

By that evening, Kevin Kalonda’s phone was gone.

He is a clinician in the Democratic Republic of the Congo, an immunisation specialist, and over the preceding weeks he had talked seventy colleagues into signing up for a free online course on malaria.

Seventy people, most of whom he knew only as names in a WhatsApp group.

That group was on the phone.

He bought another handset the next day.

Then he sat down with a problem he described without drama when he next spoke to his colleagues: “With the WhatsApp group, so far, I count 70.”

70 people who had just started something difficult, and one man trying to see all of them at once.

“I said to myself, right, the work has suddenly become difficult.”

What he did was build a poll.

Four options, and he read them out later like a man describing a piece of equipment he had made himself.

Where are you with the course?

One: I haven’t started.

Two: I’ve started but haven’t finished the opening section.

Three: I’ve finished it.

Four: I have a problem and I need help.

Option four is the entire story.

A nurse in Bukavu learns from a doctor in Bouaké

Kalonda is an Ambassador of the Geneva Learning Foundation, a Swiss non-profit that runs free certificate courses for health workers in countries where training budgets are thin and travel is expensive.

The courses work by inversion: instead of an expert lecturing a room, participants write up what they have actually done in their own districts, read each other’s accounts, and comment on them.

A nurse in Bukavu learns from what a district doctor tried in Bouaké.

The Foundation calls the people who complete these courses Scholars, and the network now runs to tens of thousands of them.

Ambassadors are Scholars who choose to do more.

The role goes out by invitation only, to alumni who have shown up consistently and then completed a further four-week programme in leadership and community facilitation.

Their job is to carry learning opportunities into their own professional networks, and – the part that turns out to matter – to stay beside the colleagues they recruit until those colleagues finish.

Between the beginning of June and the middle of August 2026, the francophone Ambassadors met every single Wednesday.

Twelve Wednesdays, no gaps, through the middle of a northern summer that means nothing to a health worker in Kinshasa or Yaoundé.

Around 300 hours of accumulated talk.

16 countries dialling in.

Almost none of them were in the same room, and most of them had never met.

They arrive twenty minutes early, just to see whether the line will hold

The meeting opens before it opens.

For the first 20 minutes there is no agenda – civil society leader and TGLF Ambassador Patrice Bamulenga, who facilitates, calls it the partie social, and its real function is to let people find out whether their connection will survive the next hour.

People arrive early to test the line.

Some type a greeting and vanish.

Some greet the room and then explain why they cannot speak.

“Hello Charlotte and everyone, I’m in a regional meeting on maternal death surveillance, I won’t be able to speak, but I can listen and follow with my headphones.”

That is Olivier Djah, a district medical officer in Côte d’Ivoire, attending two meetings at once.

He had enrolled 113 people in the malaria course.

He had built the list himself, contacts and email addresses, and he was working through it.

Two people never missed a single Wednesday: Nicolas Kashama, a nutritionist in the Miti-Murhesa health zone near Bukavu, and Noelly Watusadisi, a district physician.

Twelve out of twelve, on top of full working weeks, in a part of the world where the internet is a negotiation rather than a utility.

Kalonda gave the role its name at one of these meetings, and the phrase stuck to the whole cohort.

To support people properly, he said, you have to understand the tool they are using and you have to be near them. Les ambassadeurs de proximité. Ambassadors of closeness.

“I prefer that description,” he said.

“It’s better that we be ambassadors of proximity.”

He had an operational version too, which is less romantic and more useful.

Don’t wait until Friday to ask whether Wednesday’s task got done.

Monday, Tuesday, you start messaging them.

“We accompany them. We do not do the work for them.”

Then they argued about it, which is the best evidence that any of this was real.

Rameaux Nkollo Mbachan, a community facilitator in Cameroon, had been watching his fellow Ambassadors solve problems on behalf of the people they were meant to be coaching, and he told the room to stop.

“We accompany them, we don’t give them the direct solution. “We accompany them. We do not do the work for them.”

A Scholar who is handed every answer never learns to navigate the platform, and never learns anything else either.

Louise Nkusu Lusangu came at it from the opposite side.

A Congolese nurse living in South Africa, she had listened to Kalonda’s four-day follow-up rhythm and did not like the pressure in it.

You cannot push people, she said.

Go gently.

Kalonda answered her in the chat, in writing, and conceded the point without abandoning his method:

“We follow them without subjecting them to any pressure whatsoever. The Ambassador must always make sure of the quality of the information their Scholars produce.”

Nobody was in charge of that exchange.

Nobody had to have it.

They woke her at two in the morning, and she got up

What accompaniment costs is easier to see one person at a time.

Kashama had a Scholar who could not submit his peer review.

He tried to walk him through it over WhatsApp, and WhatsApp would not carry it.

So he went.

“I was obliged to follow him, given that he wasn’t far from where I live. I accompanied him and I think we submitted his observations together.”

He also tells the Scholars he supports to work at night, when the connection steadies.

Lusangu’s colleagues in eastern DRC study in the small hours, and when they do, they wake her.

“They got up at two in the morning to wake me, me here in South Africa. Because I love my career, I got up at two in the morning, we worked together until four.” One morning she joined the Wednesday meeting the day after xenophobic violence had shut her area down – a death, the shops closed, everyone indoors.

She reported that the atmosphere had settled, and then she said this: “We are here. We are here. We are not going to move.”

Mireille Mbiya Kabundji works in Tanzania, where the colleagues she recruits need Swahili, the course is in French, and she thinks in both plus English.

Two of her recruits could not get themselves registered.

“I fought,” she said.

“I fought, I showed them how to move forward.” She does all of it from a handset – she apologises for it at the start of every meeting, “as usual, I won’t be long, because I always follow you on the phone” – and then she offered to record tutorials for everyone else in the same position.

“There are a lot of people who don’t have a tablet.”

Marlène Kapinga Mulumba, who runs continuing education for a health service in DRC, had two recruits who told her they had no time.

She gave them a number instead of an argument: fifteen to twenty minutes a day.

“And it works.”

And Krou Mathieu Nguessan, in Côte d’Ivoire, recruited ten people, watched four of them actually start, and could not do the course himself because he had moved from a district post into teaching.

He kept following the four anyway.

“I went hand to hand,” he said.

“I even went to their workplace.”

The five weeks when the record shows nothing happening

Then, in July, the record went quiet.

The Ambassadors log their outreach in a register – one line per act of dissemination, what was shared, through which channel, to roughly how many people.

In June the register filled up: dozens of entries, peaking in the third week.

From the first week of July, across five weeks, it received exactly one.

Read the register alone and you would conclude that the summer had defeated them.

The attendance tells the opposite story.

Those same Wednesdays in July drew 41, 29, and 47 people – three of the four best-attended meetings of the entire twelve weeks.

Nobody had left.

The work had changed.

The course was now running, which meant the job was no longer recruiting strangers but keeping the enrolled ones moving: answering questions, fixing registrations, translating technical terms for a nurse whose French was thinner than her clinical judgement, phoning people who had gone silent, sitting with someone whose progress bar had frozen at 82 per cent because nobody had told them to mark the activity as complete.

The register has no field for any of that.

It counts recruitment.

It cannot see accompaniment – which is very possibly the part that turns a sign-up into a finished course.

In August, once the pushing resumed, the entries came back: more than a hundred in four weeks.

“There are not only eight of us”

The Ambassadors were harder on themselves than any outsider would dare to be.

Kapinga Mulumba said the thing everyone had been avoiding.

She had noticed how few of her peers were turning up.

“We need to mobilise ourselves, because there are many of us Ambassadors. The Foundation has appointed more than two hundred Ambassadors. There are not only eight of us. Now is the time for us to activate.”

Charlotte Mbuh, who facilitates the programme from Cameroon, read the funnel out loud most weeks without softening it – how many had enrolled, how many had actually started, how many had answered the week’s questions, how many had begun reviewing their peers.

Each number smaller than the last.

She also told the room, plainly, when they had let her down: she had posted something on the community channel and nobody responded.

“Even the Ambassadors – even you, dear Ambassadors, you abandoned me. To this day, not one like, not one comment.”

The sharpest exchange was about honesty.

Some of the reach figures in the register were not credible.

Lusangu raised it as a family matter – dirty linen gets washed at home, she said – and then said it anyway: between us, let us be serious about the numbers we put in that register.

If you write fifty and you have not followed up fifty, the Foundation will publish fifty.

“Let’s give the number we are actually following up.”

Bamulenga backed her: the Foundation does not need anyone inflating their figures.

And Watusadisi turned the diagnosis on herself.

Colleagues hesitate, she said, because we are not transparent with them.

She had been transparent with one, answered all his questions honestly, and watched him decide it was not for him.

She kept the practice.

Ninety-five people in a hundred quit. Here is what moves that number.

Somewhere in this there is a fact about online learning that reframes everything above it.

Reda Sadki, who founded the Foundation, put it to the Ambassadors in July.

“In online learning, 5% completion is the norm. Anything above that should be attributed to the scaffolding, the effort that Ambassadors and everyone else involved makes to support learners.”

Five per cent.

That is the industry’s ordinary outcome – 95 people in a 100 sign up for something and never finish it.

Measured against that, the diminishing numbers Mbuh read out every Wednesday are not a disappointment.

They are the scaffolding showing up in the data.

There is a cleaner test available, and it is the reason this particular summer is worth writing about.

The malaria course ran in two languages at once.

Same Foundation, same content, same launch.

The French version had a self-governing Ambassador assembly meeting every Wednesday for twelve weeks.

The English version had no equivalent.

By the end of June, the French stream had enrolled 1,466 people.

The English stream had 496.

Be careful with that.

The two language communities may not be the same size to begin with, the Foundation’s own mailing lists feed both, and English-speaking Ambassadors may simply have worked without filling in a register – absence of record is not absence of effort.

The comparison is an association, not a proof.

What can be said is narrower and still remarkable: the stream with a room enrolled roughly three times as many learners as the stream without one, and organised peer-to-peer persuasion is the most plausible reason.

He thought the silence meant he had been rejected

Letters to commend the francophone TGLF Ambassadors went out on 25 August.

They are formal, in the way that matters: Foundation letterhead, a reference number, a statement of what the Ambassador did between June and August and that they did it on top of a full-time professional post.

The letter recommends them to their colleagues, their management and their partners.

It puts their name on a public roll of honour where anyone – a supervisor, a ministry, a future employer – can verify it.

To understand why a letter is not a small thing, go back to the first Wednesday of the series, in June, when someone who had not been in contact with the Foundation for a while typed a single line into the chat.

Marcellin Makaya:

“For me, it’s important to have these meetings. That period of silence felt so long that I thought I had been rejected.”

More than two hundred Ambassadors have been appointed.

Nine of them carried most of this summer.

It is tempting to read the gap as indifference.

Makaya suggests something else entirely – that silence gets interpreted, and that people who have gone quiet are often waiting to find out whether anyone noticed they were gone.

The room meets on Wednesdays.

The names go on the wall.

A town that was fishing with its mosquito nets

Which brings this back to what any of it is for, in a place none of the Ambassadors mentioned above has ever been.

In the malaria course, participants answer six long questions about their own practice.

One of them asks about a new tool you put in place and how it was received.

A Cameroonian participant, Arthur Fidelis Metsampito Bamlatol, answered it by writing about the distribution of insecticide-treated bed nets at Yokadouma, a town in the east of the country where he had been working.

On paper, a distribution campaign.

In practice, an argument, because in that locality the nets were not being used as nets.

Households took them and used them “every time, for other purposes, and fishing in particular.”

He is honest about how that went.

“It was not easy to convince them.”

What moved it was not a better message but a different channel and a different language: sustained broadcasting on local radio, in the local dialect.

Nothing switched.

It shifted.

“The situation gradually improved.”

And the lesson he draws reads like a method rather than a sentiment.

Change is slow, patience counts, awareness-raising has to be continuous.

And do not do the persuading yourself: get the community leaders who have already adopted the practice to talk to the ones still refusing it.

Which is, almost word for word, what a Congolese clinician with a stolen phone had been arguing about with a Cameroonian facilitator all summer, in a meeting Bamlatol never attended, about a doctrine he had arrived at on his own.

The last line of his answer is the only outcome in this story that was ever the point:

“Today, several families sleep under a mosquito net in that locality.”