Perplexity runs three ad units in 2026: sponsored follow-up questions in the related-questions carousel, sponsored citation sources in the sidebar, and native answer integrations for a narrow set of brand-safe verticals. The auction prices on a blend of relevance and bid, creative is question-shaped not banner-shaped, and measurement centers on cited-in-answer rates rather than impressions. The surface is small versus ChatGPT but has the clearest buyer workflow of any answer engine — which is why most 2026 generative-surface test budgets start here.
Start monetizing your AI app in under an hour
With Thrad, publishers go from first API call to live ads in less than 60 minutes. With fewer than 10 lines of code required, Thrad makes it easy to unlock revenue from your conversational traffic the same day.
Perplexity ads, explained — Thrad 2026
Perplexity's 2026 ad product is the most public experiment in answer-engine monetization. Sponsored follow-up questions sit inside the related-questions carousel, sponsored sources earn disclosed placement in the citation sidebar, and a handful of brand-safe verticals get native answer integrations. This piece explains every unit, how pricing works, and what the buyer workflow looks like.
Perplexity is the answer engine that made ads real. While ChatGPT was still debating whether to introduce sponsored results and Gemini was running internal experiments, Perplexity shipped three disclosed ad units — sponsored follow-up questions, sponsored citation sources, and vertical native integrations — and published pricing and policy documents explaining how they work. In 2026, those units are the clearest picture any marketer has of what answer-engine advertising actually looks like from the buy side.
This explainer walks through each unit, the auction and pricing model, the measurement stack, and the buyer workflow a 2026 brand follows to run their first Perplexity campaign. It's written for marketers and ad-ops leads who have heard the phrase "LLM ads" for two years and want a concrete answer about what they can actually buy today.
What is a Perplexity ad?
A Perplexity ad is a disclosed, question-shaped placement that sits adjacent to or inside an AI-generated answer. The three units in 2026 are the sponsored follow-up question in the related-questions carousel, the sponsored citation source in the sidebar, and the vertical-specific native answer integration for a short list of approved categories. All three are explicitly labeled "Sponsored," all three go through human creative review, and all three price on a blend of bid and relevance rather than bid alone.
The important thing to internalize is that a Perplexity ad is not a banner, a text ad, or a sponsored-search result in the Google sense. It's a placement that has to work as an answer — a sponsored follow-up question has to read like a question a real user would ask, a sponsored citation has to be a source a real answer could plausibly cite. Creative that doesn't clear that bar gets throttled in the auction regardless of bid.
Unit | Where it renders | Creative format | Typical CPM |
|---|---|---|---|
Sponsored follow-up question | Related-questions carousel | Natural-language question + destination | $15–$35 |
Sponsored citation source | Answer sidebar | Source URL + title + snippet | $20–$45 |
Native answer integration | Inline in generated answer | Vertical-specific (product card, booking, etc.) | Negotiated |
How does the Perplexity ad auction work?
Perplexity runs a second-price auction that takes semantic relevance as a first-class bid modifier rather than a downstream filter. Each impression opportunity — a user issues a query, the system identifies the related-questions carousel slot or the citation sidebar slot as monetizable — is scored on two dimensions simultaneously: how much the advertiser is willing to pay, and how closely their creative matches the user's intent. An ad with a lower bid but a tighter semantic match can outrank an ad with a higher bid and a looser match.
This is a meaningful break from classical search auctions, where quality score is a threshold gate rather than a continuous bid modifier. In Perplexity's model there is no fixed threshold; the relevance score continuously bends the effective CPM up or down. Advertisers that write question-shaped creative which mirrors real user phrasing pay lower effective CPMs for higher-intent traffic. Advertisers that write generic promotional copy pay a premium and still lose impressions to better-matched competitors.
Perplexity's auction treats "does this creative belong in an
answer to this query?" as a real-valued bid input, not a pass/fail
gate. That single design choice is what makes question-shaped
creative economically cheaper than banner-shaped creative on the
same surface.
Why does Perplexity matter in 2026?
Perplexity matters because it's the only answer engine with a
mature, self-serve ad product and public measurement primitives in
2026. ChatGPT is larger by an order of magnitude on DAU, but its
ad product is still rolling out unevenly across sponsored search
and shopping surfaces and is not fully self-serve for most buyers.
Gemini has been running internal experiments since 2025 but has not
shipped a public auction. Copilot's ad product is coupled to the
Microsoft Advertising stack but does not yet expose generative
surfaces as a distinct inventory line.
In that landscape, Perplexity is where a 2026 marketer goes to actually run an answer-engine campaign end-to-end — set up an account, upload creative, configure bids, see impressions served, measure what happened. The surface is small, but the workflow is real. For most brands, running one quarter on Perplexity is a cheaper and faster way to learn answer-engine mechanics than waiting for ChatGPT's self-serve product to stabilize.
The second-order reason Perplexity matters is that its public documentation has become a de-facto reference for the category. Its disclosure rules, its creative policies, and its measurement definitions are being referenced by IAB Tech Lab and by every brand-safety vendor building for answer engines. Even marketers who never spend a dollar on Perplexity end up reading Perplexity's policy docs to understand how the category is converging.
What does the buyer workflow look like?
The buyer workflow has four stages: account approval, creative build, campaign launch, and measurement review. Approval takes 3–10 business days for a first-time advertiser and involves a policy review of the brand and the vertical. Regulated categories (healthcare, financial services, political) have an extra review step and a narrower set of eligible ad units. Once approved, the advertiser accesses a self-serve console where they can build sponsored follow-up questions, upload citation sources, and configure audience and vertical targeting.
Creative build is where most first-time advertisers spend too much time. A sponsored follow-up question is typically 6–14 words, phrased as a real user would phrase it, and points to a destination URL. Writing one well requires reading the actual related-questions feed for your keyword cluster, noticing the phrasing patterns, and matching them. A question that reads as copywriting ("Looking for the best CRM?") will underperform a question that reads as native curiosity ("Which CRMs integrate with HubSpot without Zapier?").
Campaign launch is standard: budget, schedule, bid ceiling, and the set of queries or topics the campaign should be eligible on. Perplexity allows both broad topical targeting and narrower keyword-cluster targeting. Most 2026 buyers run a broad topical match for discovery during week one, then tighten to the best- performing clusters by week two.
Measurement review centers on three numbers: impressions served, click-through rate on the sponsored question or source, and the cited-in-answer rate for the advertiser's domain. The third number is the one that's new and specific to answer engines — it's the share of relevant answers in which the advertiser's source was pulled into the generated response, whether or not a click occurred. Cited-in-answer rate is often a more meaningful indicator of influence than click-through, because answer engines resolve many queries without the user clicking at all.
Common misconceptions about Perplexity ads
"It's just sponsored search." No — the auction weights semantic relevance continuously, creative is question-shaped, and measurement includes cited-in-answer rates that have no analogue in classical search.
"The surface is too small to matter." True on raw DAU, but the buyer base is disproportionately high-intent — researchers, analysts, technical buyers — and the measurement stack is mature. For B2B and considered-purchase verticals, the surface punches above its DAU weight.
"You need a massive budget to test." False — most 2026 pilots start at $5k–$15k for the first month, which is enough to collect directional data on creative, bid, and vertical fit.
"Creative is impossible to write." It's different, not impossible. The discipline is writing in the user's phrasing, not the brand's phrasing. Every major agency that has shipped on Perplexity has a short internal guide on this.
"Citations can be gamed." Perplexity's citation ranking weights authority signals, originality, and direct-match quality. A thin page built for sponsored placement will get served, but it will not be cited organically in answers — and the cited-in-answer rate is what buyers actually pay for.
What comes next for Perplexity and answer-engine ads
The most likely 2027 trajectories for Perplexity's ad product are (1) native answer integrations expanding beyond the initial set of verticals into healthcare, B2B SaaS, and local services, (2) a programmatic API that lets DSPs plug into Perplexity inventory through the IAB Tech Lab AI Ad Network reference architecture, and (3) tighter measurement primitives around answer influence — including post-answer attribution windows and a standardized cited-in-answer measurement specification.
Beyond Perplexity itself, the category will see convergence: ChatGPT's sponsored-question product will start to look more like Perplexity's, Gemini's eventual public auction will borrow Perplexity's auction design, and the IAB category-level standardization will make cross-platform buying feasible. The platforms that moved first on disclosed ads and public documentation will shape how the category works for everyone.
How to get started on Perplexity
A lean 2026 starter campaign looks like this. Set aside a pilot budget of $10k over four weeks. Pick one product or service line with clear, non-regulated intent. Submit for advertiser approval early — the 3–10 day window is the critical-path dependency for a week-one launch. While approval is pending, spend a day reading the actual related-questions feed on Perplexity for your core keyword cluster: capture 30–50 real question phrasings so your creative has a native voice to imitate.
Build 8–12 sponsored follow-up questions that map to the phrasing patterns you captured, plus 4–8 sponsored citation candidates that point to your best existing content assets (long-form explainers perform best; thin product pages underperform). Launch with a broad topical match, a mid-market bid ceiling, and conservative daily pacing. By end of week two you should have enough data to cut the bottom third of creative, double-down on the top third, and narrow targeting to the best-performing clusters. By end of week four you should have a baseline for CPM, CTR, and cited-in- answer rate that you can use to size a permanent budget line or plan a broader test across ChatGPT and Gemini.
If the cited-in-answer rate is what surprises you most, you've absorbed the core lesson of the surface: answer engines reward content worth citing, not content optimized for clicks. That changes both what you advertise and how you write.
perplexity advertising, perplexity sponsored questions, answer engine ads, llm ads perplexity
Citations:
Perplexity, "Introducing sponsored questions," 2024. https://www.perplexity.ai
Perplexity, "Publisher revenue program," 2024. https://www.perplexity.ai
IAB Tech Lab, "AI Ad Network Reference Architecture," 2026. https://iabtechlab.com
eMarketer, "Answer engine ad spend forecast, 2026," 2026. https://emarketer.com
Digiday, "Inside Perplexity's ad product," 2025. https://digiday.com
Adweek, "How brands are buying Perplexity," 2026. https://adweek.com
WARC, "Measurement for answer engines," 2026. https://warc.com
Be present when decisions are made
Traditional media captures attention.
Conversational media captures intent.
With Thrad, your brand reaches users in their deepest moments of research, evaluation, and purchase consideration — when influence matters most.
Date Published
Date Modified
Category
Advertising AI
Keyword
perplexity ads