Four documented employee advocacy programs, and the deal each one offers

Four employee advocacy programs are documented well enough to check, two of them German and two American. Taking part is voluntary in all four. What separates them is what the company puts in and what it asks back. OTTO offers a seminar place and asks for nothing. Telekom funds up to a fifth of the working week. Dell requires its own training before anyone posts about the company. Starbucks pays selected baristas and sends them briefs.

That range is the useful part, and it is what the listicles flatten. Every one of these companies left the decision to post with the individual. What each of them was willing to spend, and what it asked for in return, is the thing you actually have to settle before you copy anything.

Disclosure before anything else: we build Authira, employee advocacy software for teams of three to thirty people. All four companies below are far larger than that, none of them uses it, and this page recommends nothing. They are here because they published something checkable, which most examples on this topic did not.

The four programs that are actually documented

A note on what is missing before we start. The first page of results for this question is almost entirely vendor blog posts, and reading them back to back shows most of them citing each other rather than a source. Every case below rests either on the company's own page or on dated trade press with a named author. Where a figure is not published, this page says so rather than borrowing one.

Deutsche Telekom: around 300 volunteers, some on company time

Telekom's ambassador initiative is the best-documented program in the set, because the company publishes the details itself. Its page, dated 25 October 2024, describes an initiative founded in 2015 “as a grassroots movement” that now has around 300 active ambassadors from over 15 countries.

Two details are worth more than the headcount. The participants “organise themselves independently of hierarchies and departments”, so this is not a marketing team directing people. And volunteering “takes centre stage, with many of them getting involved alongside their main job”, while “some areas even allow up to 20 percent of working time to be used for ambassador activities”.

That last clause is the expensive part, and it is the one most summaries leave out. Telekom pays nothing extra for the posts. It pays in working hours, which for a fifth of a week is a substantial standing commitment.

One caution if you go looking for confirmation. Several German-language pages report around 150 ambassadors instead. The 300 figure is the one on Telekom's own site, so that is what this page uses.

OTTO: 100 sign-ups for a seminar nobody had to attend

The German retailer OTTO built its Jobbotschafter program from roughly 2017. The clearest account of it is a PR-Journal interview dated 20 January 2020 with Eugenia Mönning, then press spokesperson for HR. It reports around 200 participants at the time, and more usefully it describes how the thing started.

When the program was announced internally, 100 employees signed up for the voluntary seminar within a short period. Nobody was assigned. The principle Mönning gives for it is the most quotable line in any of these four programs: “Niemand muss. Jeder kann. Damit muss auch das Unternehmen lernen umzugehen.”

Nobody has to, everybody can, and the company has to learn to live with that. The third sentence is the one that costs something. A genuinely voluntary program means accepting that some of the people you most want to see posting will decline, and that you have no lever to change it.

Starbucks: the one that pays

Starbucks launched Green Apron Creators in 2024, and it is the clearest example of a company treating this as paid creative work rather than unpaid advocacy. In June 2026 it became the first brand to pilot a custom Creator Network inside TikTok's Content Suite, an arrangement that lets Starbucks send briefs to selected participants and compensate them through advertising revenue sharing.

Two trade publications reported it independently within a day of each other, both quoting Erin Silvoy, Senior Vice President of Global Marketing (Marketing Dive, 23 June 2026 and Tubefilter, 24 June 2026). Starbucks' own newsroom post on the program refuses automated retrieval, so both reports here are secondary, and they agree on every fact used above.

What neither publishes is how many baristas take part. If you see a participant number for Green Apron Creators, it did not come from Starbucks or from either of these reports. This is also the case that sits furthest from what most people mean by the category, because paid contributors under brief are closer to a creator program than to employees posting on their own account.

Dell: the policy came first

Dell appears in almost every list on this topic, usually with a decade-old training statistic attached. The document you can actually read is its Global Social Media Policy, effective November 2023, and it is worth reading because it is not what the lists imply.

The policy sets out five principles. One of them is “Protect Information”. It then states the training requirement plainly. Employees are “required to take this and other SMACU courses” if social media is part of their role, and equally if they discuss the company's products or its competitors anywhere on it. Breaches carry disciplinary consequences up to termination.

That is a compliance instrument with a training program inside it, and for a company of Dell's size it is a reasonable design. It is worth separating from the Telekom and OTTO model, though, because the two answer opposite questions. Dell's policy governs people who were going to post anyway. Telekom and OTTO are trying to persuade people to start.

What each company puts in, and what it asks back

Put the four side by side and one thing they share stands out first. Every one of them left the decision to post with the individual. Nobody in these programs is required to publish anything, including at Dell, where what is mandatory is the training rather than the posting.

So the interesting column is not whether people had a choice. It is what each company was willing to spend to make that choice attractive, and what it asked for in return. Ordered from the cheapest commitment to the heaviest:

The last column is where the category quietly stops being one thing. OTTO and Telekom are running the same practice at different budgets, and the gap between a seminar and a fifth of the working week is money rather than intent. Dell adds governance on top without changing who decides. Starbucks is doing something else: once a company briefs the content and pays for it, the post is advertising with a colleague's face on it, and the audience reads it that way.

None of these is the correct answer, and the useful thing about seeing them in a row is that it forces the question early, while it is still cheap to change your mind. What goes wrong is rarely picking the wrong one. It is never picking, then asking people to volunteer and quietly counting who did.

What none of these examples tells you

All four companies employ tens of thousands of people. Telekom has an internal community large enough to organise itself, OTTO ran seminars, Dell built a course catalogue and Starbucks negotiated a bespoke arrangement with TikTok. If you have twelve people, none of that is available to you, and copying the visible parts of a large program is how a small one gets abandoned.

There is a more useful absence here too. Looking specifically for a documented program at a company under roughly 1.000 people returns vendor case studies describing their own customers, with no independent record and usually no date. This page publishes no participation benchmark for the same reason. The numbers in circulation come from surveys that do not state their sample or their fieldwork window, and a figure like that is worse than no figure at all, because it gets planned against.

So the transferable part is not the machinery. It is the arrangement, the written guidance and the decision to ask rather than assign. Those cost nothing and they are what the four programs above actually have in common underneath the differences.

The smallest version that transfers

If you want to start from these examples rather than admire them, four things carry over to a team of any size.

Settle what you are putting in, and write it down. Decide what the company spends on this, whether that is time, training or money, and what you are asking for in return. Say it in a sentence before you approach anyone. Most programs skip this and end up implying one thing while measuring another.

Ask individuals, do not announce a program. Both German examples began with people opting in to something specific. A launch email to everyone produces a spike and then the collapse that kills most programs around week three.

Give people the guidance before they need it. Dell's policy is overbuilt for a small team, but the instinct behind it is right: people post more freely when the boundaries are written down. A short social media policy does that job at any size.

Expect the blockers to be personal, not structural. The reasons an individual stops posting are mostly about confidence and time, and no policy or platform resolves them. If you are the person who has to sell this internally, the launch side of it has its own guide.

One terminology note, since two of the four examples are German. Their sources call this a corporate influencer program, which is the label that established itself in German-speaking markets for the same practice. The definition and the two operating models are set out separately, along with where the two terms actually differ.

Employee advocacy examples: FAQ