Taboola today published The Data Advantage Report, a survey of 326 senior marketing and advertising leaders in the United States and United Kingdom. According to Taboola, 81% agreed that their organization tends to keep targeting the same audience segments as budgets grow, and 92% said they were likely to shift budget to the open web if it offered a true closed-loop targeting environment.

In Short

Taboola, an advertising technology company, today published a survey of 326 senior marketers at large companies, and 81% of them agreed that they keep aiming ads at the same groups of people even when they have more money to spend. It matters to anyone who buys or sells ads on news sites and similar pages, because the marketers said they would move an average of 28% of their budget there if they could follow one person from first click to purchase. Nothing about how ads are bought changes today, and the numbers are opinions collected by a company that sells ads on those sites, so the report shows intent rather than money actually moved.

What Taboola published

Taboola (Nasdaq: TBLA) today released the report together with a press release, and the report itself is dated October 2026. According to Taboola, a single purchase decision now runs across mail, news, search, price aggregators and, increasingly, large language models, while U.S. adults spend nearly 13 hours a day with media on the way to purchase. The press release links the 13-hour figure to an outside page without naming the study, and the same statistic appeared in Taboola's Realize ID announcement of September 22.

The report's three key findings, according to Taboola, are these. First, 81% of marketing and advertising leaders "admit they tend to keep targeting the same audience segments over time", even though budgets are growing by 9% on average. Second, 79% describe their performance marketing channels as not fully integrated, and 75% say it is not very easy to recognize the same customer across several touchpoints before conversion. Third, 92% say they would be likely to shift budget to the open web if it offered a true closed-loop setting, defined in the report as the ability to identify one consumer's journey from first touch to conversion. The report adds that 48% would reallocate more than a quarter of their budget.

Adam Singolda, Taboola's chief executive, framed the results around a gap in visibility. "Outside closed ecosystems, advertisers face a significant identity blind spot," he said. He went on: "Unlocking identity and measurement on the open web allows brands to break free from targeting ruts and connect with valuable, untapped audiences." The report sums up its own reading in one line: "Close that identity gap, and almost all advertisers say they're ready to act."

How the survey was run

The report states more about method than the press release does. According to the report, Taboola commissioned a survey of more than 300 senior marketing and advertising leaders in the US and UK, working at companies with 1,000 or more employees and monthly marketing spend of $300K or more. Respondents came from three industries - automotive, banking or financial services, and eCommerce - split roughly evenly, and held roles from senior manager to vice president. They were recruited through a global B2B research panel and invited by email, with all responses collected during August 2026. The order of answers in most non-numerical questions was randomized to prevent order bias. Every chart carries N=326.

The report names Qualtrics as the independent survey company that completed the fieldwork. The three graphics attached to the press release, however, carry the logo of Global Surveyz, which administered Taboola's March 2026 survey of 200 senior performance marketers, published on May 14. Which company handled what is not explained in either document.

Some details remain absent. The report gives no margin of error, no weighting description and no split between brand-side and agency-side respondents. The demographics pages give the country split as 52% US and 48% UK, seniority as 37% vice presidents, 33% senior managers and 30% directors, and show 58% of respondents at companies spending $1M or more a month. Because respondents were drawn from three sectors with a $300,000 monthly floor, the sample describes large advertisers in specific verticals. Nothing in either document speaks to smaller accounts or to retail, travel, media or other sectors.

How the headline figures compare with the charts

Headlines compress; charts do not. Set side by side, the press release, the report's key findings and the report's charts do not always line up.

Headline figureWording in the press releaseWhat the report's charts and text show
81%Advertisers failing to reach new audiencesAgreement with a statement that the organization tends to keep targeting the same segments over time despite budget growth: 29% strongly agree, 52% somewhat agree, 6% neutral, 12% somewhat disagree, 1% strongly disagree
75%Difficulty recognizing the same non-converting customer across touchpointsFigure 6: 25% very easy, 31% somewhat easy, 12% neither, 26% somewhat difficult, 6% very difficult. The report's own text says 56% find it at least somewhat easy, 44% do not find it easy, and 32% describe it as an active challenge
81% (second use)Open to increasing ad spend on publishers and news sites if identity challenges are solvedFigure 9: 22% very open, 59% somewhat open, 13% neither, 6% resistant. The chart title and text concern openness to shifting spend from walled gardens to the open web and carry no identity condition
92%Would reallocate spend to the open web if identity and measurement challenges were solvedFigure 11: likelihood of shifting budget given a true closed-loop targeting environment - 32% very likely, 60% somewhat likely, 7% neither, 1% somewhat unlikely
Nearly 10%Advertiser budgets rising9% average increase over 12 months; 15% rose by more than 20%, 62% by 1 to 20%, 21% stayed flat, 2% fell slightly
30%New audiences cited as a top priority for additional budgetFigure 2: 30% reaching new audiences, 24% more frequency against existing audiences, 24% testing creative on existing audiences, 22% bidding harder on the same pools

The 75% is the sharpest gap. The key findings page describes it as respondents saying it is "not very easy" to recognize the same customer, which is every answer except "very easy". Under the chart's own scale, a majority of respondents rated it at least somewhat easy, and about a third called it difficult. Both readings are arithmetically defensible, but they describe different problems, and "difficulty" fits the 32% better than the 75%.

The 81% on openness to open web spend also changes character. In the press release it is conditional on identity challenges being solved. In the report it is a stand-alone attitude measured before the hypothetical is introduced, which makes it closer to the earlier 81% from Taboola's May survey, where respondents agreed they would raise open web investment if it offered agentic automation comparable to search and social. That earlier figure split 49% strongly and 32% somewhat. The same percentage, then, has now appeared against three different framings in under five months. Whether that reflects a stable appetite or a stable question format cannot be settled from the published material.

The headline 81% and 92% figures are both top-two-box sums: 29 plus 52, and 32 plus 60. Fewer than one respondent in three chose the strongest option on either the targeting or the likelihood scale. Both describe what respondents say about themselves; neither measures reach, frequency or audience overlap.

What the full report adds

Budgets

According to the report, 77% of respondents saw budgets increase over the past 12 months. The report also contains a caveat that the press release does not. "As we did not ask respondents where this additional budget is being directed," it states, "it could be that some are 'spending more to stay in place'". Rising costs, in other words, may account for part of the growth. The press release phrase "budgets are rising nearly 10%" rounds the 9% average up and carries no baseline.

Among those with extra money, 30% named reaching new audiences as their primary desired outcome. Yet 70% named something else, with frequency, creative testing and heavier bidding on existing pools clustered at 22% to 24% each. The report itself reads that spread as advertisers "spreading budget across several approaches at once", and as a sign that extra budget is partly used "to keep pace". Pressure to show finance leaders that spend reaches new audiences is split: 46% feel very or somewhat pressured, 32% feel unpressured and 21% are neutral.

Integration and identity

Only 21% describe their performance channels as fully integrated, with 52% saying somewhat integrated, 10% neither, 12% somewhat fragmented and 5% completely fragmented. The report cautions that the 52% "likely reflects a lower bar than it sounds", since connecting search and social is easier than unifying identity across display, open web and email.

On the obstacles, 35% named limited standardization of audience and identity data across the open web as the top barrier to understanding consumer intent outside walled gardens. Loss of third-party cookies drew 21%, the absence of a privacy-safe identity solution 17%, fragmented publisher data 16% and lower reach 11%. Asked how difficult understanding intent is without an identity solution, 49% said it was not easy (17% very easy, 34% somewhat easy, 14% neither, 29% somewhat difficult, 6% very difficult).

Why spend stays put, and how much might move

The top reason spend stays in closed platforms, at 18%, is lower confidence in targeting accuracy. Another 17% pointed to difficulty proving incremental ROI, 14% to satisfaction with other platforms, 13% each to tracking, brand safety and internal processes, and 12% to a lack of expertise.

The final chart is the only one that attaches a quantity to the intent. If the open web offered true closed-loop targeting, advertisers said they would reallocate 28% of current ad budget on average. By band, none of the respondents (0%) chose zero, 9% chose 1 to 10%, 43% chose 11 to 25%, 38% chose 26 to 50% and 10% chose more than 50%. Treating the top band as 60% and using band midpoints gives roughly 29%, close to the stated average. The May survey had put the average expected open web allocation at 24%, among 200 performance marketers rather than 326 senior leaders. The new figure is a stated share of budget under a stated hypothetical. The report reads it as a "substantial, immediate reallocation", yet the chart carries no timeframe and no definition of the "current ad budget" being reallocated.

Identity as the stated bottleneck

Recognizing one non-converting visitor across touchpoints is a job for identity resolution, the process of deciding which cookies, devices, emails and addresses belong to one person or household. The links are then kept in an identity graph. Closed platforms have an easier time because users log in; on the open web, the same person may read a news article on one device, search on another and click an ad on a third, with nothing tying the events together.

How well that linking works is contested. A Truthset study for CIMM and Go Addressable, cited by FreeWheel on February 28, 2026, put IP-to-postal-address links at 13% accuracy. Simulations by LiveRamp and the MMA, published on July 20, 2026, modelled a campaign with a true return of $1.50 per dollar reading as $0.43 when half of matched users were linked to the wrong person. The 50% precision case was labelled an illustrative lower bound, and LiveRamp sells identity tools. The direction of the finding, though, is that identity errors distort the measurement on which spending decisions rest.

The business behind the survey

Taboola's case rests on the open web, the inventory outside search and social. Realize reaches over 600M daily active users, according to Taboola, across publishers such as NBC News and Yahoo and OEMs such as Samsung and Xiaomi. The product line has widened through 2026: Realize+, an agentic system, was added on April 23, and on August 19 Realize took over programmatic display on NBCNews.com and TODAY.com.

The press release does not mention Realize ID. The report does, in its closing "About Realize" section. According to the report, Realize ID is Realize's identity and signal infrastructure, which clusters cookie-based IDs, mobile advertising IDs, publisher IDs, hashed emails and RampID into one persistent profile and processes hundreds of millions of such clusters daily. The report says campaigns built on its directly observed signals have produced CPAs up to 40% lower and conversion rates up to 144% higher, and it ends with a prompt to start a Realize campaign. No baseline, sample or period accompanies either figure. The same section says that because Taboola's code sits on both the publisher page and the advertiser's site, signals write back to the same ID, "producing a closed loop" - the same term the survey used for its 92% question. The Realize ID announcement of September 22, 2026 cited an uplift of up to 2.4x, also without a disclosed methodology. It entered a market that already had at least seven identifiers in circulation, and Taboola was named on April 27, 2026 as one of four launch partners testing PayPal Ads ID.

The financial backdrop is mixed. On August 5, 2026, Taboola reported second-quarter revenue of $476.8 million, $15.2 million below the floor of its guidance, after a Google spam policy removed a product it had expected to add more than $20 million of ex-TAC gross profit in the second half. The same results counted 2,081 scaled advertisers with average revenue per advertiser of about $197,000, effectively flat on a year earlier. The format is under wider pressure; Google told a court that open-web display fell from over 40% of AdWords display impressions in January 2019 to 11% in January 2025.

Why this matters for the marketing community

Surveys of stated intent are cheap to run and slow to confirm. The Data Advantage Report shows that large advertisers in three sectors describe a latent appetite for the open web, hedged by identity and measurement. It does not show spend moving. Taboola's own August network figures point the other way for now: spend per advertiser is not expanding. The distance between "likely" and money actually moved is the number nobody has published.

There is also the question of who owns the answer. A company diagnosing an identity problem while selling an identity product is not unusual in this industry, and it does not invalidate the figures. The report is more forthcoming than the press release on sample, timing and question design, and candid about two limits: budget destination was not asked, and "somewhat integrated" is a low bar. Yet it closes with a product call to action, and the 75% and the conditional 81% are worded more strongly in the release than the charts support. PPC Land's reading of Realize ID noted that the product is tied to Realize and to the inventory and conversion data Taboola controls, a shape closer to identity inside a walled garden than to the shared identifiers the open web has tried to build. If identity is the gating factor for open-web budgets, where that identity lives becomes a commercial question as much as a technical one.

For publishers, the stakes are plain. News and similar inventory stands to gain if even part of the stated intent converts; a 28% average would be large if it were real. For measurement vendors, the survey adds another data point to a debate in which accuracy figures, from Truthset and LiveRamp alike, keep arriving from parties with products to sell. The percentages are tidy; the conditions attached to them are not!

Will the 28% survive a version of the question that sets a budget, a timeframe and a named alternative? Will respondents who said "likely" in August 2026 still say so when a closed-loop environment is actually on offer? Neither is answered in the published material.

Timeline

Summary

Who: Taboola (Nasdaq: TBLA), led by chief executive Adam Singolda, which commissioned a survey of 326 senior marketing and advertising leaders in the US (52%) and UK (48%) at companies with 1,000 or more employees and monthly marketing spend of $300K or more, in automotive, banking or financial services, and eCommerce. The report names Qualtrics as the survey company; the press release graphics carry the logo of Global Surveyz.

What: The Data Advantage Report, in which, according to Taboola, 81% agree their organization tends to keep targeting the same audience segments despite budget growth, 79% say performance channels are not fully integrated, 81% are very or somewhat open to shifting spend to the open web, and 92% say they would be likely to shift budget given a true closed-loop environment, with an average of 28% of budget reallocated. The charts show agreement and likelihood scales, not measured behavior, and the press release's 75% figure corresponds to 32% describing identification as difficult in the report's own chart.

When: Published today, October 5, 2026. Responses were collected during August 2026.

Where: The open web, meaning publishers, news sites and device makers inside Taboola's network of more than 600M daily active users, as reached through respondents in the United States and United Kingdom.

Why: Taboola argues that fragmented identity and measurement keep large advertisers in closed platforms. The report arrives 13 days after Realize ID and supports the company's case for open web budgets, with the caveat that the company publishing the diagnosis also sells related products, and that the report itself says budget destination was not asked.