Contextual vs. behavioral advertising: businesses often use both

Contextual targeting responds to the page or current search. In the outdoor retailer’s example, a guide to choosing a hiking daypack could provide the context for an advertisement for a bag. A search for hiking daypacks could also supply that context. Eurostat groups both forms under its contextual measure, so these findings should not be read as a study of display advertisements alone.^[2][3]

Behavioral targeting uses information about earlier online activity or a profile. For the outdoor retailer, the audience could be selected because its earlier browsing suggests an interest in hiking equipment, such as boots, tents or maps. Geotargeting selects by location. The survey asks about each method separately, and a business may report more than one.^[2][3]

These categories describe how an audience is selected. They do not identify a particular platform, placement, creative format or budget. A method’s name also does not establish what data a particular implementation uses or which privacy obligations apply. Those details belong in the campaign’s actual setup and review.

The national result is consistent: all 18 EU countries with the required 2024 figures have a calculated minimum above 70% among behavioral advertisers. Even the smallest of these minima remains above that threshold when we allow for rounding in the published inputs. Nine EU countries lack the method data needed for this comparison; they are shown as unavailable in the table.^[1][5]

Ireland provides the lowest minimum, about 71%, while Latvia’s is about 92%. These numbers describe the least overlap that could fit each country’s published figures. Ireland’s actual overlap could be higher than Latvia’s minimum. The table therefore answers how much co-use the figures require, rather than ranking countries by their actual co-use.

The distinction is useful when reading an advertising comparison. The methods have different definitions, but the evidence does not support treating their business users as separate populations. A company using behavioral targeting will often also have contextual advertising somewhere in its activity. That does not mean it combines both methods in the same advertisement or campaign.

Start with a simplified, fictional group of 100 online advertisers. Suppose 75 use contextual targeting, 40 use behavioral targeting, and 80 use at least one of contextual, behavioral or location-based targeting. Adding the first two groups gives 115 memberships. Those memberships must fit inside a group of only 80 businesses, so at least 35 businesses must appear in both groups.

The exact overlap is still unknown. It could be anywhere from 35 to 40 businesses in this example. Dividing the minimum 35 by the 40 behavioral advertisers gives 87.5%. This is the smallest share of behavioral advertisers that could also use contextual targeting under the example’s assumptions. The calculation uses the size of the groups, so it does not tell us which particular businesses appear in both.

The published EU aggregate gives a similar illustration. Contextual advertising is reported by 76.83% of online advertisers, behavioral advertising by 41.57%, and at least one of the three methods by 81.82%. The implied overlap is between 36.58% and 41.57% of online advertisers. Among behavioral advertisers, that puts the minimum near 88%. These are calculations from published estimates, not exact population counts.^[1][5]

For the main country finding, we use each country’s own figures. We keep the EU aggregate separate because its coverage follows the available national data, rather than a simple average of all 27 countries. Neither comparison tells us how much businesses spend on each method or which one produces their best results.

A large overlap among behavioral advertisers does not mean that nearly all contextual advertisers use behavioral targeting. The two groups are different sizes. In the published EU aggregate, contextual advertisers are the larger group. The same possible overlap represents roughly 48% to 54% of contextual advertisers, compared with a minimum near 88% of behavioral advertisers.^[1][5]

This is why the denominator belongs beside the headline. Saying that “most behavioral advertisers also use contextual” is supported by the country calculations. Reversing that sentence would ask a different question. For the EU aggregate, the possible share of contextual advertisers also using behavioral targeting falls on either side of half, so the figures do not establish a majority in that direction.

For someone evaluating a proposal, the practical response is to ask which group a cited statistic describes. A percentage of businesses that advertise online is different from a percentage of all businesses, and both differ from a percentage of a campaign’s budget. Keeping those distinctions visible prevents an adoption statistic from becoming a spending recommendation by accident.

The useful conclusion is narrower and more concrete: many businesses that report behavioral targeting also report contextual targeting. Their experience demonstrates that the categories can coexist at business level. It does not establish that a business already using contextual advertising should add behavioral targeting. That decision still needs a reason tied to its audience, content, resources and goals.

The pattern is not confined to the largest advertisers. For businesses with 10–49 people employed, the estimates require more than half of behavioral advertisers to use contextual advertising in every one of the 18 reporting countries. In 16 countries, that minimum exceeds 70%. Ireland and Cyprus are the two exceptions to the 70% threshold; both still exceed half.^[1][5]

The separate adoption rates help explain the setting. In the published EU aggregate, 39.72% of small online advertisers report behavioral targeting, compared with 54.60% of large online advertisers. Contextual adoption is 76.43% and 80.85%, respectively. Small means 10–49 people employed and large means at least 250. Each percentage uses online advertisers of the stated size as its denominator.^[1]

At country level, large advertisers report higher behavioral adoption in 17 of 18 countries. Slovakia is the exception: 37.73% of large advertisers report it, against 45.72% of small advertisers. This is a descriptive pattern, not proof that size causes a business to adopt a method. Industry mix, available staff and other explanations are not tested here.

For a smaller team, the evidence supports keeping both methods in the discussion when they fit the task. It does not make the large-business adoption rate a target to copy. A retailer with one useful landing page and a limited budget still needs a specific reason to run each campaign.

Return to the fictional outdoor-equipment retailer, which employs 12 people. Its team wants customers to choose a daypack that suits the walks they have planned. For this example, the team has an illustrative €1,200 advertising budget. Neither the business nor the budget comes from the survey, and no sales or performance result is assumed.

One proposal would advertise the retailer’s daypack guide alongside relevant content or searches. The guide would help a reader compare capacity, fit and the equipment they expect to carry. Another proposal would advertise to an audience whose earlier activity suggests an interest in hiking equipment. That campaign would still need a clear promise and a destination that helps the person choose a suitable bag.

The retailer can consider both proposals without deciding that they deserve equal shares of the €1,200. It should compare the actual audience settings, available content, costs and review criteria before allocating the money. It may choose one proposal, both, or neither. The survey’s adoption percentages supply context for the discussion, not a spending formula.

The same distinction matters when the team reviews results. A person may encounter several pieces of marketing before buying. Campaign reports can describe outcomes under the business’s existing attribution rules, but a higher reported return for one campaign would not by itself isolate the effect of its targeting method. Differences in the offer, audience, placement and timing can also matter.

Use these prompts for each proposed campaign. Write down the actual choices rather than labeling the whole business “contextual” or “behavioral.” Use the answers to compare proposals that serve the same customer task, or to explain why different tasks need different approaches.

Customer task: What should the person understand or do after seeing the advertisement? Complete the sentence, “This campaign helps someone ___.” For the outdoor retailer, that might be choosing a daypack for a planned walk. Keep the task specific enough to guide the content.

Audience selection: What information will make a person eligible to see this campaign? Describe the proposed page context, current search, earlier activity or location in plain language. Check the actual settings with the person running the campaign. A platform name alone does not answer this question.

Promise and destination: What does the advertisement say, and where does the person go next? Identify the page that fulfills that promise. In the retailer’s example, a guide-led advertisement should lead to the relevant guidance; a product offer needs current product and ordering details.

Budget and decision: What can this campaign spend, who owns the decision, and what would lead the team to continue, change or stop it? Use the business’s own costs and goals. Do not convert a percentage of survey respondents into a recommended percentage of the budget.

Review: Which outcomes are already measured, over what period, and under which attribution rules? Note important differences between campaigns before comparing results. Use the business’s existing measurement and attribution rules. Note where the figures come from so the team can repeat the review on a consistent basis.

The finished brief should make the next discussion simpler. The team can see why a campaign exists, how its audience is chosen, what content it needs and how the business will review it. Those decisions are more useful than choosing a method solely because another group of businesses uses it.

We analyzed Eurostat’s isoc_cismt dataset for the 2024 survey period, the latest year with populated observations for the required advertising measures in our extract. The scope covers businesses with at least 10 employees and self-employed persons in selected manufacturing, construction and other non-financial industries. It does not represent every industry or firms with fewer than 10 people. The source files were retrieved on October 17, 2026.^[1][4]

We use the directly published percentage of enterprises using internet ads, PC_ENT_ADS. Every national calculation matches country, year, industry scope and business size. The 18-country headline uses the 10-or-more size group. We do not divide EU all-enterprise method figures by overall advertising participation: the optional measures have different coverage, so that calculation would change the denominator.^[1][4]

The lower bound is contextual adoption plus behavioral adoption minus adoption of any of the three methods, or zero if that difference is negative. Dividing it by behavioral adoption gives the minimum among behavioral advertisers. We also calculate the feasible upper bound and check all possible membership combinations. The source union includes geotargeting; it does not include the separate “other methods” category.^[2][5]

Austria, Belgium, Denmark, Estonia, France, Croatia, Italy, Luxembourg and Sweden lack the required method observations in this extract. Portugal’s medium-size contextual and union estimates are also unavailable and carry a low-reliability flag. Missing estimates remain unavailable in the tables and downloads. The small-versus-large national comparisons use the 18 countries with estimates for both size groups.^[1]

We allow each input, published to two decimal places, to vary by 0.005 percentage points. The all-country minimum-above-70% result survives that check. This tests rounding, not survey sampling uncertainty. The public extract does not supply the respondent-level data or joint sampling information needed for confidence intervals on our calculated bounds. Countries are not ranked by statistical significance.

The analysis adds an overlap calculation to existing Eurostat comparisons of targeting-method adoption. It does not identify exactly which businesses use both methods or why. The finding concerns methods used by the same business, potentially in different campaigns. Choosing an effective campaign still requires its own brief, costs and results.^[3][5]