Notion：有曝光數據的 3 則 LinkedIn 廣告均為創作者贊助貼文

LinkedIn publishes impressions, dates and countries only for ads shown in the EU. In the 50 Notion ads I pulled on November 4, 2026, three have those details. All three are long first-person posts by creators, with a sponsorship line inside the text. Two of them reported 100k–150k impressions each in nine days.

The rest of the LinkedIn sample reads differently: research stats, AI workshops, customer stories. Two voices, and only the borrowed one has reach numbers.

Sample frame

Source	Snapshot date	Filters	Ads collected	Texts analysed
LinkedIn Ad Library	2026-11-04	company name "Notion", cap 50, no country or date filter	50, of which 3 with EU details	first 30 texts in library order, cut at 300 characters; 25 distinct
Meta Ad Library	2026-11-04	page facebook.com/NotionHQ, `activeStatus: active`, no country or date filter	137, all active	first 30 bodies in library order, cut at 300 characters; 3 distinct

LinkedIn hit its cap of 50, so it is a slice; Meta stopped at 137, under its cap of 150. The two sources show different campaigns: on Meta, a sign-up push with a free-trial offer; on LinkedIn, B2B content for team leaders plus the creator posts. I compare those campaigns, not the two networks.

The three posts with numbers

Start	End	Days	Impressions	Format	Who speaks
2026-10-22	2026-10-30	9	100k–150k	Single image	A solopreneur and "Notion Ambassador"
2026-10-20	2026-10-28	9	100k–150k	Single image	A creator selling a "9-5fastlane" community
2026-10-27	2026-11-04	9	30k–50k	Single image	A creator writing for startups that work with creators

None has a CTA button in the export; each links out inside the text. The same eight countries top the impression list for all three, including the United Kingdom, Spain, France and Turkey.

Each reads as the creator's own content, with the disclosure lower down. The solopreneur post ends:

"Ad: this post is sponsored by Notion."

The startup post sets up a contrast with big-company hiring, then lands on the reader:

"If you're at a startup, that team is probably you and a spreadsheet."

The third opens with salary and passive income; Notion appears only after the creator pitches their own community. Two of the three mention a free-months offer: "3 months of Notion Business free" in one, "up to 3 months free access" in the other.

What Notion's own LinkedIn copy sounds like

In the 30 texts the collector kept, the brand voice is about AI at work, addressed to managers. Simple keyword rules over those 30:

Rule	Texts (of 30)
"AI" as a word	18
contains a percentage	10
mentions "agent"	10
Australia or Singapore	8
"AI Labs" workshop	4
"Answer Bot"	4
Ramp or Rakuten customer story	3

Three of the 30 texts are the openings of the creator posts above. Two typical brand lines:

"Bring a real workflow. Leave with a working Custom Agent. AI Labs Sydney, 24 November, for engineering, product, data, design and IT leaders."

"61% of Australian leaders are confident in their organisation’s AI adoption. But do their teams feel the same way?"

Across all 50 ads the formats are 29 single image, 11 video, 6 document and 4 event; the top buttons are "Learn more" (21) and "Download" (9).

Meanwhile on Meta: three texts, one ask

The 30 Meta bodies boil down to 3 distinct texts. 22 are the same Custom Agents pitch, 6 are a "Fall promotion", 2 are a short brand line. All 30 contain "AI" and 28 mention agents. The promotion opens:

"🎯 Act Fast: Get Your Notion Business Plan FREE for 3 Months → Full AI Suite Included → No Credit Card Required"

Across all 137 ads, "Sign up" is the button on 99, "Get offer" on 21 and "Learn more" on 15.

So the free-months offer shows up in two places: the Meta promotion and two of the three creator posts. The brand-voice LinkedIn texts I kept don't mention it.

What this data can't tell you

Three ads is not a pattern. EU details exist for 3 of 50 LinkedIn ads (6%). The other 47 have no dates or impressions, so I can't say whether brand posts reached more or fewer people.
Who paid for which post isn't in my export. The search for "Notion" also returns other pages: the advertiser list shows Notion (42), Notion Korea (5), Notion Studios Inc. (2) and Notion Digital Forensics (1). I rely on the disclosure inside each creator post, not on the page field.
No spend, clicks or sign-ups. Impressions are ranges, and nothing here shows which voice brings customers. Five of the eight longest-running Meta ads also share one start date (July 31, 2026), so I don't read their order as a ranking.

3 things to test this week

Write one ad in a customer's voice. Take your best case study and rewrite it as a first-person post with the disclosure at the end, then run it next to the brand version with the same image.
Put the offer where the story ends. Both creator posts that carry the offer place it after the reader's problem. Test the same offer as the headline versus as the last paragraph.
Check your competitor's EU slice first. It is the only LinkedIn data with impressions, so look there before assuming their brand posts are what gets seen.

Repeating this for a competitor

I built these scrapers. The LinkedIn Ad Library Scraper takes a company name or page URL and returns text, format, CTA, advertiser and, for EU ads, dates, impressions and countries; the Facebook Ads Library Scraper does the same for a Facebook page. Both are no-code forms with CSV or Excel export, free to try on the Apify free plan, and a run of a few hundred ads costs well under a dollar.

Not affiliated with Notion; all data is from public ad libraries.

If you sponsored one creator post this quarter, would you let them bury your name until the last line, or ask for it in the first sentence?