OpenAdLibraryOpenAdLibrary
Affiliate & Media Buying

How to Advertise on Taboola: Setup and Measurement Checks

Prepare a Taboola campaign, verify tracking without ad spend, check creative requirements and choose a budget tied to your own business assumptions.

How to Advertise on Taboola: Setup and Measurement Checks

To advertise on Taboola, set up a Realize advertiser account, confirm that your offer and target market are eligible, configure measurement, and prepare a campaign with a controlled budget and truthful creative. Review the account's current requirements before enabling delivery. This guide separates preparation from the decision to spend.

1. Confirm the offer, account and buying constraints#

Record the legal advertiser, target countries, landing domain, billing currency and person responsible for the campaign. Check the account's funding, billing and verification requirements directly. A deposit, a daily spending cap and a recommended test budget are different things; this guide does not assign a universal minimum to all accounts.

Check both the advertised category and the destination against the current advertising policies. Rules can depend on the product and geography. A competitor's visible ad is not approval for your offer.

Decide what result you need to measure before building the campaign. A purchase, qualified lead and page view answer different business questions. Include refunds, rejected leads and fulfilment costs when deciding what a conversion is worth.

2. Configure and verify measurement before spending#

Realize's tracking setup documentation points to the Tracking area and Account Tracking Setup. Available installation methods include tag-manager, manual and supported commerce integrations. Select the method appropriate to the actual site; do not install duplicate copies through several methods.

Configure the relevant event and its value/currency where applicable. Use the site's consent controls and the applicable platform instructions. Server-to-server reporting is useful for supported measurement paths, but it is not permission to bypass consent or to send events you cannot substantiate.

Use a non-spending verification sequence:

  • Keep advertising delivery disabled. Open your own test destination directly, not through a paid placement.
  • Use an approved test account or test transaction. Avoid creating a real charge or a real customer lead solely for this check.
  • Run the platform's Tracking Test Tool and inspect whether the expected event is received. Confirm the account/pixel, event name, value, currency and absence of duplicate events.
  • Check the consent-denied path separately. Record what is sent, what is suppressed and what the chosen integration documents.
  • Record the test time and the displayed result. Event receipt demonstrates instrumentation; it does not demonstrate that paid attribution or campaign optimization works.

The platform says tracking status can take time to update. A browser request or an “active” indicator alone is not a complete purchase-event test. Follow the Tracking Test Tool instructions and investigate discrepancies before enabling spend.

3. Choose a strategy for the objective#

The current bid-strategy guide lists Fixed Bid, Enhanced CPC and Maximize Conversions, with an optional target CPA for the latter. It recommends Maximize Conversions; it does not establish this article's former mandatory sequence of fixed bidding, 15–30 conversions, then automation.

Choose from the strategies actually available for the campaign objective and account. Confirm what the entered bid or target controls. A target CPA is an optimization target, not a guaranteed price per acquired customer. Verify conversion instrumentation before asking an algorithm to optimize toward that event.

4. Set a budget you can evaluate#

Start with an explicit maximum acceptable loss and a business hypothesis. For illustration only, a $40 contribution per valid sale and a 1% click-to-sale rate give a break-even CPC of $0.40. At a 0.5% rate, that falls to $0.20. These are arithmetic scenarios, not expected Taboola results.

Choose a spending cap and review point consistent with that uncertainty. A budget ten times a target CPA does not guarantee ten conversions or completion of a learning phase. Do not treat fixed 3–5-day waits or a universal 100-click threshold as proof that a result is reliable. Allow for your conversion lag and predefine the evidence needed to continue, pause or revise the experiment.

The Taboola cost worksheet explains this calculation in more detail.

5. Prepare creative for the actual placement#

The thumbnail policy states a 400 × 350 minimum for horizontal thumbnails and a 5 MB limit. Its current 2026 creative guide recommends 1200 × 674 and headlines of 35–45 characters, with 60 as the maximum. Recommendations and acceptance minimums serve different purposes; inspect the selected format's requirements in the uploader.

1000 × 600 is 5:3; 1200 × 600 is 2:1. Neither is 16:9. Keep the subject within safe crops and use the previews rather than assuming a single source image will fit every placement. Other formats, including vertical or motion creative, have different specifications.

Use accurate branding and a headline that matches the destination. An image's visual interest does not substantiate a health, tax, insurance or financial claim.

Historical tax-deadline creative

June 2026 historical example from this guide. The tax-forgiveness deadline is an unverified advertiser claim. It illustrates why urgency must be checked, not copied.

Historical medication-and-memory creative

Historical health creative. A medical-sounding authority and list format do not establish the underlying claim. Do not reuse its premise as a supported recommendation.

Historical insurance creative

Historical example of audience-specific wording. Check actual eligibility and disclosures for the policy being promoted.

These examples are not measured winners. Their capture history cannot reveal CTR, conversions or profitability.

6. Review the campaign before enabling delivery#

Confirm the destination, objective, conversion event, currency, spending cap, schedule, target regions, device eligibility, exclusions and creative previews. Record the settings you approved. Split campaigns when you need separate budgets or an interpretable experiment; a mandatory desktop/mobile split is not a universal rule.

Review campaign and item statuses individually. Realize describes an aim of one business day for content review, with longer review possible. Approval, funding, eligibility and enabled delivery are separate conditions.

After launch, evaluate first-party cost and outcomes with conversion lag accounted for. Compare CPA, qualified outcomes and contribution, not just cheap clicks. Public archive research can help generate creative hypotheses; it cannot replace this measurement.

Frequently asked questions

How can I test Taboola tracking without spending on ads?
Keep delivery disabled, open your own test destination directly and use the documented Tracking Test Tool with an approved test event. Verify the account, event, value, currency and duplicate behavior. Event receipt alone does not prove paid attribution.
What image dimensions should I use for Taboola?
The cited thumbnail policy specifies a 400 by 350 minimum for horizontal thumbnails and a 5 MB limit; the 2026 creative guide recommends 1200 by 674. Confirm the selected format in the uploader and inspect cropping. Neither 1000 by 600 nor 1200 by 600 is 16:9.
Must a new campaign use manual bidding before automation?
The current documentation does not establish that as a universal sequence. Choose from the strategies available for the objective and account, and verify the conversion event before optimizing for it.
The OpenAdLibrary Team
Written byThe OpenAdLibrary Team
Ad intelligence & native advertising research

We build OpenAdLibrary, the open ad-transparency platform. Every day our systems capture live native ads across Taboola, Outbrain, MGID, Revcontent, Teads, Yahoo and MSN, identify the real advertiser behind each one, and follow the click to its landing page. These guides distill what we see in that data so you can research the market faster.