TikTok Search

TikTok’s search function has grown into a genuine discovery channel. In Adobe Express research fielded in early 2026, 49% of US consumers said they had used TikTok as a search engine, up from 41% two years earlier.¹ That figure measures whether someone has ever done it rather than how often, which is the distinction most commentary on TikTok search loses. For younger users it is a real starting point for discovery, though not a replacement for Google.

How does TikTok search work?

TikTok’s search bar returns a mix of video content, creator profiles, and trending sounds ranked by a combination of relevance and engagement signals. The ranking factors are not publicly documented in detail, but the observable signals include: keyword matches in captions, spoken words in the video, on-screen text, hashtags, engagement rate (likes, comments, shares, saves), and recency.

Unlike Google, TikTok’s search results are heavily influenced by personalisation. Two users searching the same query may see significantly different results based on their watch history and engagement patterns. This makes TikTok search less predictable than Google search from a content strategy standpoint.

TikTok also surfaces a “Search” tab and in-video search prompts that suggest related queries after a user watches a video, which extends discovery beyond the search bar itself.

TikTok’s own figures are more useful than most of the consumer surveys quoted on this topic, because the platform is measuring its own product rather than asking people to recall behaviour. It reports that “57% of users utilize TikTok’s search functionality and 23% search for something within 30 seconds of opening the app”.² The clearest evidence that the company treats this as real search intent is commercial: its Search Ads Campaign product sells “sophisticated keyword-based ads which specifically target TikTok’s search results page”. A platform does not build keyword-targeted advertising against a surface nobody uses deliberately.

One structural change is worth tracking on a platform whose ranking behaviour was already undocumented. TikTok’s US operations were separated into the TikTok USDS Joint Venture, which closed on 22 January 2026, with Oracle, Silver Lake and MGX taking managing stakes and ByteDance retaining 19.9%. Under the arrangement Oracle would “retrain, test, and update the content recommendation algorithm on U.S. user data”.³ Neither TikTok nor the joint venture has published what that means for search ranking specifically, and no before-and-after evaluation of ranking behaviour exists publicly. The defensible reading is narrow: US users are served by a system trained on a different data set from the global app, so what you observe in one market should not be assumed to hold in the other.

What do users search for on TikTok?

The content categories that perform best in TikTok search reflect the platform’s strengths: authenticity, visual demonstration, and personality.⁴

• Video tutorials: the most common search use case on TikTok, covering everything from makeup to software to cooking
• Product reviews: short, opinionated reviews from real users are trusted more than polished brand content
• Personal stories: first-person accounts of experiences, decisions, and outcomes
• Influencer recommendations: what people with established credibility in a niche recommend

Fashion, beauty, food, finance, and software are among the categories where TikTok search is most active. Categories that require depth or precision (legal, medical, complex technical) are less suited to short-form video and see lower search engagement.

TikTok versus Google: what the data says

The narrative that “TikTok is replacing Google” is not supported by current data. The share of Gen Z respondents saying they are more likely to rely on TikTok than Google fell from 8% in 2024 to 4% in 2026, and when the same survey asked which platforms people find most helpful for searching, Google led at 85%, followed by Reddit at 29%, ChatGPT at 26%, YouTube at 24% and TikTok last at 16%.¹ Adobe also found more consumers preferring ChatGPT to Google (14%) than TikTok, so the substantive challenge to Google among younger users is coming from AI assistants rather than from social video.

A widely repeated trio of figures is often used to argue the opposite: 67% of Gen Z using Instagram to search, 62% TikTok and 61% Google. Those come from a SOCi survey of 1,002 US 18 to 24 year-olds published in March 2024, they are scoped to local results specifically, and they are multi-select responses that sum to 190%.⁵ They describe which platforms young people have used for local discovery, not how search demand divides between them, and they are now over two years old. They do not support a claim that the three platforms hold comparable shares of search.

Different platforms serve different query types. TikTok wins for discovery-led, lifestyle, and recommendation queries. Google wins for transactional, factual, and high-stakes queries. A user researching a product may find it on TikTok and then search Google for price comparison and reviews before purchasing. These platforms are sequential, not mutually exclusive.

How do you optimise content for TikTok search?

Start from TikTok’s own search data. Creator Search Insights, launched in March 2024, “gives creators insight into topics people are searching for on TikTok” and is reached by typing that phrase into the app’s search bar.⁶ Its most useful view is content gaps, which TikTok defines as topics “searched for often, but aren’t featured in a large number of videos on TikTok”. It is the closest thing the platform offers to a keyword tool, and it reports the platform’s own query behaviour rather than a third-party estimate, which is the same reason TikTok’s usage figures are more reliable than the consumer surveys quoted around this topic.

Captions and spoken words matter. TikTok indexes the text of captions and the words spoken in the video. A clear, spoken statement of the topic in the first few seconds, matched by an accurate caption, improves discoverability. This is the equivalent of a title tag and opening paragraph for TikTok.

On-screen text. Text overlays are indexed and contribute to keyword relevance. Putting the core topic or question as on-screen text in the first few frames serves both search and the growing share of users who watch without audio.

Direct answers early. TikTok search users are impatient. A video that answers the core question within the first 5 to 10 seconds holds viewers long enough to generate engagement signals. Videos that spend time on introductions before getting to the answer see higher drop-off rates.

Hashtags. TikTok has not published documentation setting out how hashtags feed its search index, so treat the widespread claim that they carry a defined indexing role as practitioner inference rather than platform guidance. What is reasonable on the evidence: relevant, specific hashtags help TikTok categorise content, and a few precise ones serve better than a cloud of generic ones. Unlike Instagram, TikTok has not capped how many you may use.

Authenticity over production. High production values do not correlate with TikTok search performance. Content that feels authentic and first-hand performs better than polished brand video. This is partly cultural and partly algorithmic: genuine engagement (comments, saves) correlates with authentic content more than with professional production.

TikTok search and AI citations

TikTok is cited in AI answers, but at a fraction of the rate of Reddit or YouTube, and the gap varies sharply by query language. Profound’s March 2026 analysis of 3.25 billion citations found TikTok made up 3.1% of the social citations in Google AI Overviews for English prompts, rising to 9.0% for Arabic and 15.9% for Spanish, roughly a fivefold increase over the English baseline.⁷ Those are shares of social citations specifically, not of all sources, and social citations were 15.3% of AI Overviews citations overall.

Two things follow. TikTok is not absent from AI citation pipelines, so the older claim that it simply does not appear is wrong. But for an English-language audience it remains a minor citation source next to text-led platforms, and the usual explanation is that video without an accessible transcript is harder to retrieve at passage level than text is. If your market is Spanish or Arabic-speaking, the weighting changes materially.

TikTok’s larger contribution remains indirect: it drives brand awareness and search behaviour that produces queries on platforms where citations are easier to earn and measure.

Footnotes

1. Gen Z Preference For TikTok Over Google Drops 50%, Data Shows — Search Engine Journal ↩ ↩²

2. Introducing Search Ads Campaign — TikTok For Business ↩

3. After a shaky start, TikTok’s U.S. joint venture lands on its feet — CNBC ↩

4. Is TikTok the New Search Engine? How to Optimise for Discovery in 2026 — SEO Sherpa ↩

5. Gen Z Dumping Google For TikTok, Instagram As Social Search Wins — Forbes ↩

6. Get inspired with Creator Search Insights — TikTok Newsroom ↩

7. How query language reshapes AI citations — Profound ↩