Facebook Reels Ad Transparency: Agency Research Checklist
Audit Meta's Why am I seeing this ad experience on Facebook Reels and separate delivery explanations, Ad Library evidence and targeting assumptions.
The Advertisng Research Team studies observable public Meta and Google ad evidence, documents limitations, and turns repeatable findings into agency research workflows.
Direct answer
Meta's updated “Why am I seeing this ad?” transparency tool became available on Facebook Reels on September 17, 2026. Agencies can use the three-dot explanation to record observable delivery factors, advertiser identity and placement context, then compare the creative in Meta's public Ad Library. The explanation does not reveal the advertiser's exact targeting, bid, budget, audience list, performance or causal reason for delivery.
What changed on Facebook Reels?
Meta updated its ads-transparency announcement on September 17, 2026 to say that the newer “Why am I seeing this ad?” experience is now live on Facebook Reels. Meta had previously extended the same update to Instagram Feed, Instagram Stories and Instagram Reels.
The experience is designed to explain, in summarized topics and examples, how activity on and off Meta technologies may inform the machine-learning models used to shape and deliver an ad. Meta also says the interface provides routes to Ads Preferences and lets a viewer hide ads from a particular advertiser.
This is a transparency explanation shown in the context of an ad impression. It is not an export of the advertiser's Ads Manager settings. Agencies should treat it as viewer-side evidence about what Meta chose to explain, not as a complete record of how an advertiser configured or optimized the campaign.
Meta's original announcement described the updated Facebook experience as rolling out globally. The September 17 update says the experience is live on Facebook Reels, but it does not publish a country-by-country Reels availability table. An Indian agency should verify the menu in the relevant account, app version and impression before promising that every user will see identical wording or controls.
Where should an agency look, and what should it preserve?
Meta's documented path begins with the three-dot menu on an ad and then “Why am I seeing this ad?” On Facebook Reels, menu placement and labels may vary by app version, account or region, so record the observed path rather than presenting one screenshot as universal.
For each material example, preserve:
- observation date, local time, country and device context;
- platform and placement, including that the ad appeared in Facebook Reels;
- advertiser or Page name visible on the ad;
- the creative format, copy, call to action and destination that were observable;
- the exact explanation topics or examples shown by Meta;
- whether “About this ad,” “AI info,” payer, beneficiary or “Paid for by” information appeared;
- the matching Meta Ad Library advertiser and Library ID when available; and
- a clear split between the platform's wording, the agency's inference and any first-party client evidence.
Do not ask a researcher to record personal interests, browsing history or other sensitive account details that are not necessary for the research question. The useful unit is the ad, the placement and Meta's displayed explanation, not a dossier on the viewer.
Original artifact: Reels ad-transparency evidence register
Use this register to keep three different evidence surfaces separate.
| Evidence layer | What to record | Safe agency use | What it cannot establish |
|---|---|---|---|
| Reels impression | Date, country, placement, advertiser, creative, CTA and destination | Preserve the exact ad context a viewer observed | Campaign settings, full audience, spend, results or why another person did not see it |
| “Why am I seeing this ad?” | Exact explanation topics, examples, visible controls and screenshot context | Document Meta's viewer-facing delivery explanation | The advertiser's exact targeting selections or a complete causal model of delivery |
| “About this ad” | AI info label, payer or beneficiary and “Paid for by” disclosure when available | Add provenance and disclosure evidence to the record | That an unlabeled asset contains no AI edits, or that a label proves who created every element |
| Meta Ad Library | Library ID, Page, creative, platforms, delivery dates and observed status | Match the public advertiser and compare visible creative variants | Viewer-level delivery logic, ordinary-ad spend, conversions, revenue or profitability |
| Agency inference | Hypothesis, supporting observations, contradictions and confidence | Form a research question for a creative or media brief | A fact until independently supported |
| Client first-party evidence | Ads Manager settings, delivery breakdowns, experiments, conversions and business outcomes | Validate claims for the agency's own client | A competitor's private settings or performance |
The register prevents a common research failure: combining one person's delivery explanation, a public-library snapshot and an agency hypothesis into a single statement about competitor targeting.
How is “Why am I seeing this ad?” different from “About this ad”?
The two labels answer different questions.
“Why am I seeing this ad?” is about the viewer-facing explanation of ad delivery. Meta says it can summarize how activity on and off its technologies may inform machine-learning models and provides access to ad controls.
“About this ad” is Meta's newer unified destination for additional transparency information. Meta says it can contain AI information for ads created or significantly edited with Meta's generative tools, detected third-party AI information, payer or beneficiary details when available, and “Paid for by” disclosures for social-issue, election and political campaigns. Meta also warns that the experience may vary by region because of legal requirements.
Neither surface replaces the Meta Ad Library. Use the viewer-side menus to record the delivered context and explanation. Use the Meta Ad Library research workflow to inspect public advertiser and creative evidence. Keep the source label attached to every observation.
How should agencies interpret machine-learning explanations?
Meta describes the displayed information as summarized topics and examples of how activity may inform the models used to shape and deliver ads. The words summarized, examples and may matter.
A safe note says: “Meta displayed these explanation topics for this ad in this viewer context on this date.” An unsafe note says: “The advertiser targeted this exact interest,” unless the agency has the advertiser's own settings and the claim is supported there.
Ad delivery can involve advertiser inputs, eligible audiences, platform predictions, auction conditions, placement availability and other system factors. A viewer-facing explanation is not necessarily an exhaustive feature list or a ranking of which factor caused the impression. Preserve the wording without converting it into a precise targeting declaration.
How can the Meta Ad Library strengthen the observation?
The current Meta Ad Library says it can be used to search ads running across Meta technologies. For ordinary commercial ads, it is strongest as a public record of active creative, associated advertiser or Page, delivery dates and platforms. Additional spending and demographic fields are available for specified regulated categories and regions, not as a universal feature of every commercial ad.
After recording a Reels impression:
- search the advertiser in the Meta Ad Library for the observed country;
- match the Page identity, copy, media, CTA, destination and approximate delivery period;
- retain the Library ID and observation date;
- record whether the exact variant was found, a close variant was found or no match was found; and
- do not treat a missing match as proof that the impression was invalid or the ad never ran.
Meta says a new ad can take up to 24 hours after its first impression to appear in the Ad Library and that updates can also take up to 24 hours to be reflected. Timing, geography, category selection and advertiser identity can therefore affect a match.
Use the ethical Facebook competitor-research guide when turning multiple observations into a market pattern. Use the ad creative analysis rubric to describe message, offer, format and proof without copying execution.
What is a defensible agency workflow?
Use a small, repeatable protocol rather than collecting isolated screenshots.
- Define the research question before opening any platform, such as whether competitors are using creator-led Reels for a specific offer.
- Capture only naturally observed ads; do not manipulate a person's account or sensitive interests to manufacture an explanation.
- Record the Reels context and Meta's exact transparency wording.
- Check “About this ad” for additional disclosure evidence when available.
- Match the advertiser and creative in the public Ad Library.
- Label each statement as platform-displayed fact, public observation, agency inference or client-only evidence.
- Compare several advertisers or repeated observations before calling something a category pattern.
- Translate the pattern into an original hypothesis, substantiated claim and first-party test.
Advertisng's public ad-intelligence workspace can organize visible creative references and research notes beside this register. It should not be presented as access to private competitor targeting or performance data.
What can public ad data not prove?
Public Meta ad data and viewer-side transparency explanations cannot prove a competitor's exact targeting selections, exclusions, custom or lookalike audience membership, customer list, bid, budget, optimization event, attribution settings, frequency cap, auction rank, delivery model weights, complete causal reason for an impression, clicks, conversions, revenue, margin, incrementality or profitability. They also cannot prove that an absent AI label means no AI tool was used.
Follow the Advertisng research methodology: preserve the source, state what was observed, label inference and retain the private evidence required for any client-performance claim.
Frequently asked questions
Is “Why am I seeing this ad?” now available on Facebook Reels?
Meta's September 17, 2026 update says the newer experience is live on Facebook Reels. Availability and wording can still vary by account, app version or region, so verify the live menu before documenting a workflow for a client.
Does the explanation reveal the advertiser's selected interests?
Not necessarily. Meta describes summarized topics and examples of how activity may inform its delivery models. Do not convert those explanations into exact advertiser-selected targeting without the advertiser's own settings.
Is “About this ad” the same as “Why am I seeing this ad?”
No. “Why am I seeing this ad?” explains aspects of delivery and links to controls. “About this ad” is a broader transparency destination that may include AI information, payer or beneficiary details and regulated-ad disclosures.
Can the Meta Ad Library confirm why a person saw a Reels ad?
No. The Ad Library can help match the public advertiser, creative, platform and delivery period. It does not expose the complete viewer-level delivery logic for an ordinary commercial ad.
Does a missing AI label prove an ad was made without AI?
No. Meta's labeling depends on its product rules, significance thresholds, advertiser use of Meta tools and detection signals for third-party tools. Absence of a visible label is not proof that no AI-assisted edit occurred.
Should an agency collect a viewer's personal activity to explain the ad?
No. Record only the minimum context required for the research question. Avoid collecting personal interests, browsing history or other sensitive account details that do not belong in an agency research record.
A restrained next step
Choose one naturally observed Facebook Reels ad from a commercially relevant Indian category. Complete one evidence register, check the matching advertiser in the Meta Ad Library, and write three sentences: what Meta displayed, what the public creative shows, and what remains unknown. Repeat across several advertisers before turning the observation into a client hypothesis.
Primary sources
- Meta Newsroom: Increasing Our Ads Transparency, viewer-facing explanation scope, controls and the September 17, 2026 Facebook Reels availability update; reviewed September 29, 2026.
- Meta Newsroom: Expanding GenAI Transparency for Meta's Ads Products, “About this ad,” AI information, payer or beneficiary and regulated-ad disclosure scope; reviewed September 29, 2026.
- Meta: Ad Library, current search scope, availability timing, delivery information and category-specific transparency fields; reviewed September 29, 2026.
- Advertisng: Research methodology, public-ad evidence scope and limitations; reviewed September 29, 2026.
Related posts
Advertisng Public Ad Corpus: Q4 2026 Research Snapshot
A dated snapshot of 257,588 stored Meta and Google public-ad records, the stricter display-ready cohort, methodology changes and agency research limits.
Best Facebook Ad Spy Tools for Agency Research in 2026
A transparent comparison of public libraries and paid research workspaces, scored for source coverage, evidence quality, organization, collaboration, and cost clarity.
GA4 Campaign Data Import Validation: Meta Ads Agency Checklist
Validate Meta Ads cost, click and impression imports in GA4 with a repeatable agency checklist for UTMs, campaign IDs, dates, currencies, totals and attribution limits.