§

Spark Ads vs In-Feed Ads: which TikTok format wins in 2026

Spark Ads vs In-Feed Ads on attribution, creator licensing, CTR benchmarks, brand-safety review, and the decision tree for TikTok operators in 2026.

Soft sage monotone editorial cover with bold serif headline reading Spark vs In-Feed and mono eyebrow AD-STACK · FORMAT TEST.

Every TikTok media buyer eventually has the same fight with their finance team. Spark Ads cost more per impression, run inside a creator’s handle, require a licensing dance with the creator, and convert at rates that often look unbelievable on the dashboard. In-Feed Ads cost less, run inside the brand’s ad account, and look exactly like every other branded ad on the platform. Which one wins. The honest answer is “it depends on what you’re buying” — and most operators get this call wrong because they pick the format the agency partner is comfortable with rather than the format the campaign actually needs. This is the operator’s decision framework for 2026.

TL;DR

  • Spark Ads are paid promotions of organic posts owned by a creator’s handle, with the creator’s permission via an authorisation code. They convert higher and cost more.
  • In-Feed Ads are standard brand-account ads that look native to the feed but live inside the advertiser’s ad account, not a creator’s handle. They cost less and scale further.
  • Spark Ads win on cold-audience CTR, comments-based social proof, and any campaign where the creator’s authority matters to the offer.
  • In-Feed Ads win on creative-iteration speed, brand-controlled assets, and any campaign at volume where you don’t want to negotiate with creators for every variant.
  • The decision tree at the end of this piece maps your campaign type to the right format. Run it before your next test.

What Spark Ads actually are

Spark Ads, launched in 2021 and now the default format for creator-led TikTok advertising, are paid promotions of existing organic TikTok posts. The post lives in a creator’s handle (or your own brand handle), the ad runs against the creator’s followers and audience signal, and the engagement (likes, comments, shares, profile visits) accrues back to that handle.

The mechanics that aren’t obvious from a first read:

  • The creator owns the post. You’re not licensing the file; you’re licensing the right to spend money against a post they own.
  • Authorisation is via a creator-issued code. The creator generates a Spark Ads code from their TikTok account (Settings → Privacy → Ad Authorization → Manage). The code typically has a 7- to 365-day expiry depending on how the creator set it. You enter that code into TikTok Ads Manager when you build the campaign.
  • The creator can revoke at any time. Spark Ads code revocation kills your ad mid-flight. We’ve watched a $40k/day campaign go dark at 2pm because a creator unauthorised. Bake this risk into your campaign contracts.
  • Engagement stays with the creator. Likes, comments, shares, follows all credit the creator’s account. This is a feature (the creator’s authority compounds) and a bug (your brand handle doesn’t grow from your own ad spend).
  • The ad shows from the creator’s handle, not yours. The “Sponsored” label appears, but the brand attribution is the creator’s name on the post.

What In-Feed Ads actually are

In-Feed Ads — the older format — are standard ads built inside the advertiser’s TikTok Ads Manager, served into the For You feed, and attributed to the brand’s account. The advertiser uploads creative, sets targeting, sets bid, and ships.

The mechanics:

  • The brand owns the post. Every asset, every variant, every iteration lives inside the brand’s ad account.
  • No creator dependency. You don’t need a code, you don’t need a relationship, you don’t need permission. The asset can be commissioned UGC, in-house production, AI-generated, or anything else.
  • Engagement does not accrue back to a public-facing creator. The post is an ad, not a published organic post. Comments are typically disabled or moderated. Shares accrue to the ad, not to a creator’s follower count.
  • The “Sponsored” label is clearer. Users know they’re seeing a brand ad. This is sometimes a feature (no ambiguity), sometimes a bug (lower watch-through).
  • Creative iteration is fast. Swap the asset, save the variant, ship. No external party in the loop.

The seven-criteria decision matrix

Run any candidate campaign through this matrix before you pick a format. The summary takeaway: Spark Ads win on cold-audience CTR and social-proof-driven offers; In-Feed Ads win on volume, speed, and brand control.

CriterionSpark AdsIn-Feed Ads
CTR (typical range, 2026)1.8% – 4.5%1.0% – 2.2%
CVR (typical range, 2026)1.5% – 3.8%1.2% – 2.5%
CPM (typical range, US, 2026)$14 – $28$7 – $14
Time to first variant in flightDays (creator authorisation + post creation)Hours (upload + approve)
Creative iteration speed at scaleSlow (every variant needs creator post)Fast (variants ship in account)
Attribution clarityLower (creator handle adds noise)Higher (1:1 ad-to-account)
Brand-safety review surfaceThe creator’s full handle historyThe asset itself

The CPM gap is the load-bearing number. Spark Ads typically cost 1.5-2× the CPM of In-Feed Ads because TikTok’s algorithm prioritises content from creator handles in the feed and the auction reflects that. The CTR/CVR gap usually pays back the CPM gap — but not always, and the answer depends entirely on the offer.

Where each format wins

Spark Ads win for

Cold-audience prospecting. When the audience doesn’t know your brand, a creator’s existing authority does the heavy lifting in the first second. A creator with 200k followers and three months of organic posts has built trust your brand account hasn’t.

Social-proof-driven offers. Apps, supplements, courses, fitness products, consumer fintech — categories where “did anyone real recommend this” is the load-bearing question. Spark Ads ship the social proof inside the format.

Comments-as-creative campaigns. Some of the best-performing Spark Ads we’ve seen drive their entire performance off comments — pinned creator replies, third-party testimonial threads, debates that pull the viewer in. The In-Feed format can’t replicate this because comments aren’t part of the ad surface in the same way.

Authority-led offers in regulated categories. Health, finance, supplements — where the FTC and platform compliance require a real human voice attached to the claim, a creator’s Spark Ad reads as authentic. The brand’s In-Feed Ad reads as paid claim.

Brand-safe partnerships with on-platform talent. When the creator is part of the marketing strategy and not just a freelancer, Spark Ads are the format that compounds the relationship.

In-Feed Ads win for

Daily creative testing at scale. If you’re testing 30+ creative variants per week — which is the cadence we recommend for any always-on TikTok program — the creator dependency on Spark Ads slows you to a stop. In-Feed Ads ship in hours.

AI UGC and AI-character creative. Tools like Superscale, HeyGen, and Creatify produce AI-character UGC that performs well in In-Feed but can’t be uploaded to a real creator’s handle as a Spark Ad (it would be platform-policy non-compliant). For brands running AI-generated UGC at volume, In-Feed is the only format.

Static and slideshow creative. Spark Ads work best for video; static-heavy testing is a poor fit. In-Feed handles every format the platform supports.

Multi-market campaigns with rapid localisation. Translating creative across 8 languages for a Black Friday push is a single-sprint job in In-Feed. The same job in Spark Ads requires a creator per market.

Brand-controlled messaging. Pricing, claims, comparisons, and compliance language are easier to control on your own account.

Spend volumes above $50k/day on a single campaign. Spark Ads at that scale often hit creator-account follower-size ceilings; In-Feed scales further.

Spark Ads attribution — what actually shows up in the data

Attribution on Spark Ads is the part that confuses operators most often. The summary:

  • TikTok Ads Manager reports impressions, clicks, conversions, and pixel events the same way it does for In-Feed.
  • The creator’s TikTok analytics show the organic-equivalent metrics on the same post — views, profile visits, follows gained, comments.
  • Your attribution tool (Triple Whale, Northbeam, GA4) sees the click and the conversion the same way it sees an In-Feed click — but the source/medium tagging can vary depending on how the link parameters were set.

The mistake we watch operators make: assuming a Spark Ad’s lift is captured fully by the platform-reported number. The CTR signal is usually correct. The CVR signal is usually correct in-platform but understated cross-platform because some viewers click out of the ad to the creator’s profile, follow the creator, and come back to the brand days later via organic search or direct visit. That delayed click is rarely captured in TikTok’s attribution window.

For brands serious about Spark Ads, run a geo holdout incrementality test early. Most we’ve seen show Spark Ads under-credit themselves by 15-30% on lift versus the in-platform number.

Creator licensing — the part that breaks campaigns

Spark Ads run on a creator-issued authorisation code. The code mechanics:

  1. Creator generates code in TikTok app under Settings → Privacy → Ad Authorization. The default code lasts 60 days; creators can extend to 365.
  2. Advertiser enters code in TikTok Ads Manager when setting up the Spark Ad campaign.
  3. Code unlocks the specific post for paid promotion under the advertiser’s ad account.
  4. Code expires at the set date. The ad stops serving. There is no warning.
  5. Code can be revoked by the creator at any time, immediately stopping the ad.

The contractual pieces nobody handles well in their first Spark Ads campaign:

  • Length of authorisation. Match the code to the planned campaign length plus a buffer. If you’re running a 30-day campaign, get a 90-day code. The code is free to extend; the cost of mid-campaign expiry is everything you spent.
  • Revocation penalty. Build a clause into the creator contract that imposes a penalty for unauthorised revocation. Otherwise the creator has a free option to torpedo a campaign.
  • Exclusivity windows. Some creators run Spark Ads codes with multiple competing brands simultaneously. Read the contract; pay for category exclusivity if it matters.
  • Right to edit. The post can’t be edited after a Spark Ads code is issued without invalidating the code. If you suspect a typo or claim issue, you must catch it before the code is generated.

CTR and CVR benchmarks (US, 2026)

Pulled from our internal benchmark set across DTC, mobile apps, and SaaS clients. These are typical mid-funnel performance ranges — top-decile creators and top-decile creative can run 2-3× higher.

CategorySpark Ads CTRIn-Feed CTRSpark Ads CVRIn-Feed CVR
DTC supplements2.4% – 4.2%1.4% – 2.1%2.1% – 3.5%1.5% – 2.4%
Mobile apps (UA)1.8% – 3.6%1.0% – 1.8%1.5% – 2.8%1.2% – 2.0%
Consumer fintech2.0% – 3.8%1.2% – 1.9%1.4% – 2.5%1.1% – 1.8%
SaaS lead gen1.5% – 3.0%0.9% – 1.5%1.2% – 2.0%0.8% – 1.4%
Fashion / apparel2.2% – 4.5%1.4% – 2.5%1.8% – 3.0%1.6% – 2.6%

The longer benchmark set, including CPM and CPA bands, sits in our TikTok ads specs and benchmarks 2026 piece. The CTR uplift on Spark Ads is consistent across categories. The CVR uplift narrows in categories where the offer is the load-bearing decision (SaaS lead gen, low-AOV apparel), and widens in categories where authority matters (supplements, fintech, fitness).

Brand-safety review — different surface, different risk

Brand-safety review on Spark Ads has to look at the creator’s full handle, not just the single post you’re promoting. A creator who shared a controversial opinion last month doesn’t disappear from search results just because you’re promoting their food review this month.

Our review checklist before greenlighting any Spark Ads creator partnership:

  • Full handle history: last 50 posts, last 6 months. Search for category-relevant controversial topics.
  • Comment moderation pattern: how the creator handles negative or abusive comments on their content.
  • Other brand partnerships: who else is in their Spark Ads pipeline this quarter.
  • Geographic coverage: do they post in markets where your brand has compliance concerns.
  • Platform policy compliance: has the handle had any community guideline strikes.

For In-Feed Ads, the brand-safety review is the asset itself. Faster, cheaper, simpler — and the only surface that exists.

The decision tree

A short flow to run before your next test:

  1. Is the campaign always-on or a one-off? Always-on with 20+ variants/week → In-Feed. One-off launch with a creator partner → Spark.
  2. Does the offer benefit from creator authority? Supplements, fitness, fintech, regulated → Spark. Standard DTC, SaaS, ecom → In-Feed.
  3. Is the creative AI-generated UGC? Yes → In-Feed only. No → either.
  4. Are you testing 30+ variants/week? Yes → In-Feed. No → either.
  5. Is your CPM budget room for 1.5-2× current cost? Yes → Spark for the CTR lift. No → In-Feed.
  6. Will you run incrementality testing within 90 days? Yes → Spark gets a fair shot. No → In-Feed (Spark looks worse than it is without lift testing).

Most operators end up with a portfolio: 60-70% In-Feed for daily creative testing and always-on volume, 20-30% Spark Ads for creator-led prospecting, 10% experimental formats (TopView, Branded Effects, etc.). The ratio shifts by category — fitness and supplements run heavier on Spark; SaaS and B2B run almost entirely In-Feed.

Tooling note

If you’re shipping creative for either format at volume, the tooling matters. For AI-character UGC that feeds In-Feed, Superscale, HeyGen, and Creatify are the production-grade options — covered in detail in our best AI UGC tools ranking. For Spark Ads, your tooling is the creator relationship: the contract, the code management, and the post-approval workflow.

Frequently asked questions

What’s the main difference between Spark Ads and In-Feed Ads on TikTok?

Spark Ads are paid promotions of organic posts that live in a creator’s TikTok handle (with that creator’s authorisation), while In-Feed Ads are standard brand-account ads that live inside the advertiser’s ad account. Spark Ads inherit creator authority and engagement; In-Feed Ads give the brand full control over the asset.

Do Spark Ads cost more than In-Feed Ads?

Yes, typically 1.5-2× the CPM. Spark Ads also usually deliver higher CTR (1.8-4.5% vs 1.0-2.2%) and higher CVR, which can offset the CPM gap depending on the category and offer. For cold-audience prospecting and authority-led offers, the CPM premium is usually worth it.

How does Spark Ads creator authorisation work?

The creator generates an authorisation code from their TikTok account (Settings → Privacy → Ad Authorization → Manage), with a 7- to 365-day expiry. The advertiser enters the code in TikTok Ads Manager when building the campaign. The code can be revoked by the creator at any time, which immediately stops the ad.

Can I use AI-generated UGC in Spark Ads?

No, not directly. Spark Ads require an organic post on a creator’s real handle, and AI-character UGC uploaded to a creator’s account would be non-compliant with TikTok’s authenticity policies. AI UGC creative runs as In-Feed Ads exclusively. If you want AI creative under a creator’s voice, you have to commission the creator to record original content.

Which format works better for mobile app user acquisition?

In-Feed Ads at volume — the daily creative testing cadence required for mobile UA is incompatible with creator-dependent Spark Ads. Spark Ads can play a role at the top of the funnel for cold-audience prospecting, but the bulk of mobile UA spend usually lives in In-Feed.

How do Spark Ads affect attribution?

Platform-reported attribution treats Spark Ads the same as In-Feed Ads — clicks and conversions flow through the same pixel and the same conversion API. The under-counting risk is on delayed conversions: viewers who follow the creator instead of clicking through immediately, then convert days later via direct or organic search. Run incrementality testing to capture the true lift.

Letters from readers

  1. Q·01 How is ad-stack funded?

    We pay for every tool seat ourselves at the public plan tier, and the journal is reader-supported via the newsletter. No vendor pays for placement, and no review is sponsored.

  2. Q·02 Why benchmark on the same brief instead of letting each tool play to its strengths?

    Because the only fair variable in a head-to-head test is the tool. Letting each vendor pick their best demo brief is how the AI ad category got into its current marketing-led mess — every tool wins on its own showcase. Same brief means you can actually compare cost-to-published across the field.

  3. Q·03 How often do you re-test tools that have shipped major updates?

    Every quarter. Reviews carry a 'last tested' date in the byline. If a tool ships a meaningful capability change between quarterly cycles, we publish a field note rather than waiting — but the score on the main review only moves at the next full re-test.

  4. Q·04 Can I send in a tool to be reviewed?

    Yes — send a note via the contact link in the footer. We can't promise coverage of every submission, and being suggested has no bearing on the eventual verdict. Vendors who pay for seats themselves rather than offering us free credits are evaluated identically.