Facebook Marketing, X and Threads: What Role Remains in B2B
Opens the chat with a prepared prompt.
In DACH B2B in 2026, Facebook marketing is no longer a channel for organic community building, it is a paid and retargeting surface for older audiences. X records the lowest engagement of all platforms analysed in the Socialinsider benchmarks and now carries only real-time and press communication, while Threads is a test with almost no additional effort because of the Instagram link, but without reliable DACH reach.
Key Takeaways
- Facebook is losing users in Germany and, with 31% weekly reach according to the ARD/ZDF-Medienstudie 2025, is still the second most used platform: the base is shrinking, but it remains large.
- Organic Facebook reach no longer carries a B2B programme, Socialinsider puts engagement rates in the tenths of a percent, and the value of the channel sits almost entirely in the paid layer.
- X reports 19.0 million ad reach for Germany, up 20.7% year on year, but the figure comes from X's own advertising data and works only as an upper bound.
- Threads has grown to 500 million monthly active users globally and sits at 5% weekly usage in Germany, eighth place, and Meta publishes no official absolute figure for Germany.
- Ad reach, weekly usage from surveys and official MAU are three different types of figure and must never appear in the same ranking.
- Meta's pay-or-consent model is subject to a DMA proceeding by the European Commission, and the commitment from December 2025 is likely to change the data basis for personalised audiences from 2026.
- The most honest budget decision for all three channels: cut the time spent on them back to monitoring and response, and keep media budget only where a measurable retargeting or reach function exists.
Facebook, X and Threads usually end up under the same agenda item in DACH companies: channels somebody set up years ago and nobody has seriously reviewed since. They belong in one decision, because the same question is open for all three, namely what exactly is still being paid for here and how much staff time that is worth.
The short answer, as of August 2026: Facebook stays as a paid and retargeting surface, X shrinks to real-time and press communication, Threads is a test with effort close to zero. On none of the three channels does organic community building justify a dedicated editorial role.
Before you compare figures: keep three types of number apart
The most expensive mistake in these discussions happens before the first argument. Presentations put ad reach values, survey data on weekly usage and official MAU figures from quarterly reports side by side as if they were comparable. Ad reach is an estimate of addressable accounts taken from the platform's advertising tools. Weekly usage is representative self-reporting. MAU is a platform-internal definition that the company sets itself. Anyone who builds a ranking out of that measures nothing. The clean basis for this is in the overview of social media usage in Germany and DACH.
X adds a second problem. The German value comes from X's own advertising data and has moved by more than 20% year on year. For planning purposes it is an upper bound at best.
Facebook marketing: shrinking base, working paid layer
For Germany, DataReportal reports 22.8 million reachable Facebook accounts at the end of 2025, down 11.5% year on year. Messenger is losing faster, down 18.2% to 9.9 million. At the same time, in the ARD/ZDF-Medienstudie 2025 Facebook remains the second most used platform in Germany with 31% weekly reach, behind Instagram. Both are true at once: the base is shrinking, and it is still the second largest in the country.
Among younger people the direction is unambiguous. In Austria, according to the Jugend-Internet-Monitor 2025 from Saferinternet.at, only 20% of 11 to 17 year olds still use Facebook, eight percentage points fewer than the year before. In the DACH region Facebook has become a channel for older audiences, and that is the only audience definition you can still plan with here.
Organically the channel carries nothing any more. Socialinsider measures an average engagement rate of 0.15% for Facebook, which fell to 0.13% by the fourth quarter of 2025; only plain status posts reach 0.20%. For Q1 2024 to Q4 2025 Emplifi reports a median of 1.4% to 2.5%, because a different denominator applies there. The difference is not a contradiction in substance. It shows that engagement benchmarks are only comparable within one source. Both data sets are global, not DACH-specific.
What keeps Facebook in the setup is the paid layer. Meta's Advantage+ suite automates audiences and placements to the point where creative quality becomes the actual control variable. Drawing on US client data, Tinuiti reports that Advantage+ Shopping accounted for 33% of total Meta retail and e-commerce spend between Thanksgiving and Cyber Monday 2025, up from 26% the year before. That is a retail signal from the US and not a B2B figure, but it shows where budgets are moving in the Meta world. How to fence in this automation sensibly instead of handing it the entire budget is covered in the pillar on paid social and performance marketing.
X: growing ad reach on paper, lowest engagement in practice
For Germany, X reports 19.0 million ad reach, up 20.7%. It is also the value to trust least, because it comes from X's own advertising tools. In the ARD/ZDF survey X sits at 8% weekly usage, in sixth place, behind Snapchat and Pinterest.
The engagement side is clearer than the reach side. For X, Socialinsider measures a fall from 0.13% in early 2025 to 0.10%, the lowest rate of all platforms analysed, with 0.12% as the annual average. For 2025 Rival IQ arrives at minus 48% year on year, the sharpest drop of any platform in its data set. Both surveys are global.
On top of that comes a cost question that many teams only notice when they set up tooling. The X API bills pay per use from USD 0.001 per owned read and USD 0.20 per post containing a URL, while creating a post without a link costs USD 0.015; prices can change. For a monitoring setup that is manageable, for automated publishing across several accounts it adds up. The comparison with the other platforms is in the article on social media APIs, access, limits and costs.
X still works for real-time topics, press and media contacts, technical communities and status updates during outages. For predictable reach, content distribution and lead generation the channel drops out, because engagement in the tenths of a percent carries no distribution and the German reach figure is too uncertain to plan with.
Threads: large globally, eighth place in Germany
Threads is the only one of the three channels for which an independent third-party measurement exists alongside the platform's own figures. In June 2026 Meta reported 500 million monthly active users, and the quarterly report for the third quarter of 2025 gave 150 million daily active users. Both are Meta's own figures. Independently of those, for January 2026 Similarweb measures around 141.5 million daily mobile users against around 125 million for X, which means Threads has overtaken X on mobile.
In Germany the picture is different. The ARD/ZDF-Medienstudie 2025 measures 5% weekly usage for Threads, and 11% among 14 to 29 year olds, which is eighth place. There is no official absolute figure for Germany, Meta publishes no country data for Threads. Anyone who gives you one has estimated it.
Since January 2026 Meta has been rolling out Threads Ads worldwide with no named market exceptions; DACH is therefore covered in all likelihood. The pragmatic approach: set up the account through the existing Instagram profile, mirror existing text content, watch for three months. If nothing happens, it costs you nothing beyond the decision. For niches beyond the Meta world it is worth looking at the alternative social media platforms in the DACH region.
The three channels compared directly
Platform | German scale (end of 2025) | Type of figure | Weekly usage DE 2025 | Engagement (global) | Role in B2B |
|---|---|---|---|---|---|
22.8 million, minus 11.5% | Ad reach (DataReportal) | 31%, rank 2 | 0.13% to 0.15% (Socialinsider) | Paid reach and retargeting, older audiences | |
X | 19.0 million, plus 20.7%, upper bound | Ad reach from X's own advertising data | 8%, rank 6 | 0.10% to 0.12% (Socialinsider) | Real time, press, crisis communication |
Threads | no official German figure | global 500 million MAU (Meta) | 5%, rank 8 | no reliable benchmark | Low-cost test via Instagram |
The column with the type of figure is the most important one. It stops anyone from deriving a statement about relevance in Austria or Switzerland out of 22.8 million and 500 million.
What this means for your budget
Channel | Keep it if | Freeze it if | Realistic effort |
|---|---|---|---|
older audience, an e-commerce share or existing retargeting pools in Meta | there is no media budget and the page is run purely organically | paid setup plus responding to messages, no editorial calendar | |
X | press, public sector, a developer audience or outage communication in the business model | lead generation is the only goal of the account | monitoring plus posting when there is an occasion |
Threads | Instagram is being used anyway and content can be carried over | nobody runs Instagram | minutes per week, no separate plan |
As a rule of thumb from practice: a channel without media budget and without a defined trigger needs no editorial calendar, it needs one person who checks the inbox once a week. That is an honest cost allocation. Half a day of editorial work per week for 0.13% engagement is the most expensive item in many social budgets, because it never shows up as a cost anywhere. As soon as several channels draw media budget in parallel, a rule-based split beats annual planning in a spreadsheet; how to semi-automate that is covered under budget allocation agents for paid ads.
Regulation: what is added at Meta and X in 2026
Meta's pay-or-consent model is subject to a European Commission proceeding under the Digital Markets Act. In December 2025 Meta committed to offering EU users an option with less personal data and limited personalised advertising from January 2026. For advertisers that means, in our assessment, smaller and blurrier custom audiences, and the scale of it cannot be planned for at present. Anyone writing retargeting pools into annual planning as a fixed quantity is better off remeasuring them monthly from the first quarter of 2026, instead of setting them up once and only reporting on them afterwards.
Alongside that, the usual obligations apply to each of the three profiles: the imprint requirement for company profiles under the DDG and the Medienstaatsvertrag in Germany, the ECG and the Mediengesetz in Austria, advertising disclosure under the UWG, DSA obligations on the platform side and the transparency requirements of the EU AI Act for synthetic content from 2 August 2026. Two of these are regularly underestimated. The disclosure requirement under section 5a UWG already applies where there is consideration or an excess of promotional intent, and under section 8 paragraph 2 UWG liability also falls on the commissioning company, not only on the creator or the agency. For Swiss audiences the revised Datenschutzgesetz comes on top, which sets its own requirements for data processing and is not automatically dealt with by GDPR-compliant consent. None of this replaces a legal review, in particular not for the specifics in Austria and Switzerland.
Common mistakes
Deleting the channel outright: A deleted Facebook profile gives away the brand name and the review history. A frozen profile with a current imprint and a pointer to the active channels costs nothing and prevents mix-ups.
Reading reach figures as proof of audience: 22.8 million reachable Facebook accounts in Germany are not 22.8 million decision makers. The question is always how many people from your target segment are in there, and only the campaign manager answers that, not the market study.
Starting Threads with its own editorial calendar: That turns a zero-cost test into a project with a reporting obligation. If Threads needs its own plan in order to work, it does not work for you.
Comparing engagement rates across sources: Socialinsider, Emplifi and Rival IQ calculate with different denominators. Set yourself a target within one source or against your own previous period, never against a benchmark from another study.
What you decide on again in six months
Three signals are enough for the reassessment. First: does Facebook paid still produce retargeting volume, or have the audience sizes fallen below the threshold at which campaigns deliver after the DMA changeover? Second: does X still generate reactions from the press or a specialist audience, or is it only bots and competitors posting there? Third: do your own Instagram insights show any measurable referrals from Threads at all, or does it stay at zero?
For each of the three questions, define in advance which value counts as a yes: a minimum size for the retargeting audience, a number of press or specialist reactions per quarter, one measurable referral from Threads. Without that threshold the reassessment becomes a matter of taste, and the channel stays in the plan out of habit. With X the cost side belongs in the same decision: if monitoring or publishing run through the API, the invoice for the last six months goes on the table next to the reaction figures.
Anyone who answers these three questions twice a year needs no debate of principle about the value of Facebook marketing. The figures decide faster than any opinion in the room.
Data & Statistics
Facebook Deutschland: 22,8 Mio. Nutzer Ende 2025, minus 11,5 % gegenüber dem Vorjahr
DataReportal Digital 2026: Germany (2025)Facebook Messenger Deutschland: 9,9 Mio. Nutzer, minus 18,2 % YoY
DataReportal Digital 2026: Germany (2025)Facebook Deutschland: 31 % wöchentliche Nutzung ab 14 Jahren, Rang 2 hinter Instagram
ARD/ZDF-Medienstudie 2025 (Media Perspektiven 31/2025) (2025)Facebook-Engagement global: 0,15 % Durchschnitt, Rückgang auf 0,13 % im vierten Quartal 2025, Status-Posts 0,20 %
Socialinsider Social Media Benchmarks 2026 (2026)Facebook-Engagement global: Median 1,4 bis 2,5 % zwischen Q1 2024 und Q4 2025 (abweichender Nenner)
Emplifi 2026 (2026)Advantage+ Shopping: 33 % des Meta-Retail- und E-Commerce-Spends zwischen Thanksgiving und Cyber Monday 2025, nach 26 % im Vorjahr (US-Daten)
Tinuiti Black Friday Cyber Monday Recap (2025)X Deutschland: 19,0 Mio. Ad Reach Ende 2025, plus 20,7 % YoY (X-eigene Werbedaten, als Obergrenze zu lesen)
DataReportal Digital 2026: Germany (2025)X Deutschland: 8 % wöchentliche Nutzung ab 14 Jahren, Rang 6
ARD/ZDF-Medienstudie 2025 (Media Perspektiven 31/2025) (2025)X-Engagement global: Rückgang von 0,13 % Anfang 2025 auf 0,10 %, Jahresschnitt 0,12 %, niedrigster Wert aller analysierten Plattformen
Socialinsider Social Media Benchmarks 2026 (2026)X-Engagement: minus 48 % im Jahresvergleich, stärkster Rückgang aller Plattformen im Datensatz
Rival IQ Social Media Industry Benchmark Report 2025 (2025)X-API-Preise: 0,001 USD pro eigenem Read, 0,015 USD pro erstelltem Post, 0,20 USD pro Post mit URL
X Developer Platform (API Pricing) (2026)Threads: 500 Mio. monatlich aktive Nutzer (Juni 2026)
Meta Newsroom (2026)Threads: 150 Mio. täglich aktive Nutzer (Quartalsbericht Q3 2025)
Meta Q3-2025-Earnings (via Social Media Today) (2025)Threads: rund 141,5 Mio. mobile Tagesnutzer gegenüber rund 125 Mio. bei X (Januar 2026)
Similarweb (via TechCrunch) (2026)Threads Deutschland: 5 % wöchentliche Nutzung (14 bis 29 Jahre: 11 %), Rang 8
ARD/ZDF-Medienstudie 2025 (Media Perspektiven 31/2025) (2025)Österreich: 20 % der 11- bis 17-Jährigen nutzen Facebook, minus 8 Prozentpunkte gegenüber 2024
Saferinternet.at Jugend-Internet-Monitor 2025 (2025)Meta sagt im Dezember 2025 zu, EU-Nutzern ab Januar 2026 eine Option mit weniger personenbezogenen Daten und eingeschränkter personalisierter Werbung anzubieten (DMA-Verfahren)
Europäische Kommission, Digital Markets Act (2025)FAQ
Is Facebook marketing still worth it in B2B in 2026?
How many users does Facebook have in Germany?
How many users does X have in Germany?
Is Twitter, or X, still relevant for companies?
Is Threads worth it for companies in the DACH region?
How much organic reach does Facebook still have?
What does Meta's pay-or-consent model mean for advertisers?
Want to go deeper?
Get new analyses straight to your inbox, or see how we put this knowledge to work for companies.