Bots and AI traffic: how to protect ad budgets in 2026

The context: why non-human traffic has become a budget issue

For years, bot traffic was treated as a minor nuisance. By 2026, the situation has structurally changed. AI agents—from shopping assistant crawlers to answer engine spiders—are generating traffic volumes comparable to human traffic in many e-commerce categories. According to a analysis published by Digiday , digital retail brands are facing a real head-scratcher: non-human traffic messes with retargeting strategies, but not all of this traffic is totally worthless.

The problem is therefore twofold. On one hand, advertising platforms charge for impressions and clicks on bot-generated sessions, eroding the budget without producing conversions. On the other hand, indiscriminately blocking AI agents means excluding potential touchpoints with technology-assisted buyers. Therefore, the answer can be neither 'ignore' nor 'block everything'.

For Italian SMEs, the impact is proportionally heavier. Paid media budgets are often limited. Therefore, every single percentage point of unqualified traffic entering the retargeting funnel turns into a direct, measurable cost.

The numbers reshaping marketers' priorities

The phenomenon isn't uniform across sectors. The most affected categories are e-commerce, travel, and finance — these are the verticals where AI shopping agents are most active. The problem is set to grow, not shrink.

Additionally, the cost per acquisition (CPA) for retargeting campaigns becomes distorted when the audience pool includes bot sessions. Advertising platforms — Google Ads, Meta, programmatic — do not automatically filter all non-human traffic uniformly. Consequently, marketers who do not proactively intervene pay an invisible premium on every conversion.

A second relevant data point concerns the quality of first-party signals. If the tracking pixel records bot sessions, lookalike audience models and smart bidding strategies are fed with polluted data. Therefore, the effect is not limited to the immediate cost: it extends to the quality of future campaigns. This is why we at SHM Studio we consider first-party data cleansing an operational priority, not an optional activity.

Not all bots are created equal: an operational taxonomy

Before defining a strategy, it is necessary to classify non-human traffic. There are at least three distinct categories, with very different implications for marketers.

• Malicious bots : scrapers, click fraud, automated competitor intelligence. They should be blocked without exception. They consume budget and pollute data without any commercial value.
• Neutral crawlers : search engine bots, web archives, monitoring tools. They don't run up direct ad bills, but they can totally skew organic traffic stats if you don't sort them out properly.
• Commercial AI agents : shopping assistant crawlers like those integrated into Amazon Alexa Shopping AI or in other recommendation systems. These agents gather product info for real buyers. Blocking them could hurt your catalog's visibility on AI-assisted channels.

Therefore, the strategic decision is not binary. It requires granular segmentation based on user agents, behavioral patterns, and access frequency. In particular, commercial AI agents deserve different treatment compared to malicious bots.

Strategic reading: the dilemma dividing marketers

The debate among paid media professionals reflects a real tension. On one hand, those who argue for the need to aggressively block all non-human traffic to protect data purity. On the other hand, those who argue that AI agents represent an emerging channel of commercial distribution, and that excluding them is equivalent to giving up future visibility.

Similarly to what happens with custom purchase notifications — as analyzed in our in-depth article on Alexa Update Me When and shopping notifications — AI agent behavior is turning into way more advanced forms of shopping middlemen. Ignoring this traffic now could mean missing the boat later.

However, the pragmatic stance is that the two needs aren't mutually exclusive. You can block bad bot traffic, segment neutral traffic, and open up dedicated channels for commercial AI agents. But this setup takes some tech investment and structured data governance. It's definitely not something you can just throw together in a week.

A useful parallel comes from the world of content. The issue of visibility in AI systems — addressed, for example, in the analysis on ChatGPT Ads and the implications for Italian SMEs — shows how companies that ask themselves the right questions early on gain a structural advantage. The same principle applies to managing non-human traffic.

The cost of indecision: what happens if you don't take action

Not taking an explicit stance on bot traffic has concrete and measurable consequences. First, the CPA of retargeting campaigns progressively increases as the audience pool becomes polluted. Second, smart bidding platforms optimize towards distorted conversion signals, generating a compounding negative effect over time.

Furthermore, the quality of attribution reports deteriorates. Bot sessions that complete micro-conversions — product page views, add-to-carts — are recorded as real events. As a result, the attribution model overestimates the effectiveness of certain channels and underestimates that of others. The resulting budget allocation decisions are systematically distorted.

Operational implications: three priorities for Italian SMEs

Turning this analysis into concrete actions takes a structured approach. Here at SHM Studio we have set three operational priorities for companies looking to tackle this issue systematically.

First priority: current traffic audit. Before any intervention, it's necessary to quantify the percentage of non-human traffic on the company's digital properties. Tools like Google Analytics 4—with advanced filters—and specialized solutions like DoubleVerify or Integral Ad Science allow for granular segmentation. This analysis must be conducted separately for each channel: organic, paid, direct, referral.

Second priority: configuration of exclusions in paid platforms. Google Ads and Meta offer brand safety and invalid traffic exclusion options that many SMEs don't fully utilize. Specifically, retargeting campaigns should explicitly exclude audience segments with anomalous behavioral patterns. This intervention, integrated with a strategy of google ads campaigns well-structured, it can significantly reduce CPA within a few weeks.

Third priority: define an explicit policy for commercial AI agents. This is the newest and least managed part. Companies need to decide which AI agents to allow, with what frequency, and on which sections of the site. This decision has both technical implications — robots.txt configuration and rate limits — and commercial ones, related to visibility in AI-assisted channels. The topic intertwines with emerging engagement strategies, such as those analyzed in the context of TikTok and new engagement formats for brands : content distribution changes, and access policies must adapt accordingly.

Tools and resources for a data-driven strategy

The bot detection tool market is mature, but choosing the right tool depends on the context. For SMEs with limited budgets, combining Google Analytics 4 with advanced filters and the native invalid traffic exclusion features in paid platforms is an accessible starting point.

For companies with higher traffic volumes or high-margin e-commerce, dedicated solutions such as DoubleVerify they offer greater granularity and more detailed reports. These platforms classify traffic in real time and can integrate directly with programmatic DSPs, cutting spending on unqualified impressions before they are bought.

Finally, for those managing LinkedIn campaigns — where bot traffic is structurally different from the open web — the considerations change. The LinkedIn campaigns operate in a more controlled environment, but not immune to the problem. Here too, audience segmentation and monitoring engagement patterns remain essential practices.

To explore the entire landscape of paid media and advertising strategies, our dedicated hub on advertising, paid media and video collects the latest analysis on these topics.

Towards digital traffic governance: the framework that is still missing

The current debate among marketers reflects the absence of a shared standard. There is not yet a universally accepted taxonomy for classifying commercial AI agents, nor a communication protocol between sites and agents that allows for transparent access management. This gap is structural and will not be filled in the short term.

Consequently, companies waiting for an industry standard before acting are falling behind competitively. Conversely, those who build internal traffic governance today—even if imperfect—acquire data, skills, and processes that will become a differentiating asset when the standard arrives.

We at SHM Studio recommend treating this issue as a structural component of the strategy of Digital marketing , not as a tech issue to just hand off to the IT crew. Figuring out which agents to let in, which ones to block, and how to tweak your paid campaigns accordingly are total business decisions that hit your ROI straight on. So, marketing management needs to drive them, backed up by the right technical know-how.