The True Cost of Invalid Traffic: Are Humans Seeing Your Ads?

November 14th, 2023

Invalid traffic (IVT) impacts all online advertisers. No matter how much you spend on paid media, it’s likely that a proportion of your ad clicks are generated by invalid traffic.

Invalid traffic wastes budget on visitors who will never convert to customers. It also generates spam leads, contaminating your CRM and sending your sales team on a wild goose chase. Plus, IVT skews your analytics, restricting your ability to make data-driven decisions about budget allocation. And the explosion of advertising automation is making the problem worse.

According to our 2024 Wasted Ad Spend Report, 8.5% of all online traffic from May 2022 to May 2023 was invalid, equating to one in every 11.7 clicks:

But not all IVT has the same impact on advertisers. So how can you measure the cost of invalid traffic on your paid marketing campaigns?

In this article, you’ll learn the true cost of invalid traffic: how it impacts your performance marketing success, and the knock-on effect on lead generation and revenue.

What Is Invalid Traffic?

Invalid traffic refers to any website visits that don’t come from a real person with genuine interest in your product or service. Invalid traffic can include:

  • Bots (good or bad)
  • Fake users
  • Accidental clicks from real users
  • Malicious clicks from competitors
  • Any click that has no chance of converting to a genuine lead or customer

Advertisers are particularly affected by invalid traffic because you’re paying to get your product seen by genuine prospects. Every invalid click costs you money that could have been spent converting a real lead.

Not all invalid traffic is bot-driven, and not all bot-driven traffic is malicious. Badly-placed ads can easily draw accidental clicks from real users, while some bad actors deliberately seek out your ads and click on them to drain your ad budget and promote their own product.

IVT is usually split into two categories: general invalid traffic (GIVT) and sophisticated invalid traffic (SIVT). Let’s explore the difference and why it’s important to understand this.

Get a Free Invalid Traffic Audit

Ads are for humans, not bots. Get a demo and save up to 25% of your advertising budget by automatically eliminating fake ad engagements across all paid channels.

Understanding General Invalid Traffic (GIVT)

General invalid traffic includes the more “acceptable” types of IVT that serve a non-malicious purpose for online users. This includes:

  • Googlebot and other search engine web crawlers
  • SEO spiders
  • Analytics crawlers
  • Automated bot traffic from known, non-threatening sources

Lunio’s Head of Demand Generation, Leo Pizarro, defined GIVT in our recent invalid traffic webinar:

General invalid traffic is basic invalid traffic that’s generated by (mostly) non-malicious activity. As part of this group you have good bots like web crawlers. You also have basic automation rules that come from open source tools. They can run basic automated scripts, but are limited to ISP-assigned IP addresses and can’t often execute JavaScript or identify CAPTCHAs. They’re very easily identifiable by most ad platforms and legacy click fraud solutions.

Leo Pizarro

Head of Demand Generation, Lunio

While GIVT can still have an impact on your ad spend, it’s not always a serious problem. Google and other ad networks have a pretty good handle on recognising and eliminating GIVT from your ads and analytics.

Understanding Sophisticated Invalid Traffic (SIVT) 

Sophisticated invalid traffic is a much more significant issue for advertisers. It’s harder to detect, often malicious, and can come from bots or humans. SIVT includes:

  • Bots that pretend to be real human users
  • Botnets designed to skew data and analytics
  • Malware
  • Cookie stuffing bots
  • Accidental clicks from human users
  • Deliberate malicious clicks from human users

Leo defines SIVT:

Sophisticated invalid traffic is what you might commonly call ad fraud. It’s a type of activity that’s much harder to detect. It uses methods that emulate human behaviour (such as using legitimate IP addresses and rotating them to bypass static methods of blocking bots). It can store cookies and emulate click patterns to mimic humans. One of the most frustrating things about SIVT is that it blends in with your real website traffic in the form of a low-and-slow attack, making it difficult to pinpoint suspicious peaks in activity.

Leo Pizarro

Head of Demand Generation, Lunio

Fraudsters use SIVT to boost their own earnings or spoof domains for phishing and other illegal activities. And ultimately this kind of activity ends up affecting all marketers, by wasting ad budgets and contaminating analytics.

SIVT is a problem — and unlike GIVT, most ad networks can’t seem to stay on top of it.

How IVT Is Handled by Ad Platforms

Advertisers have long hoped ad platforms would get better at recognising and removing IVT before it impacts their revenue. But while they can detect and filter out GIVT, the problem seems to be getting worse with SIVT.

Platforms vary widely in their ability to detect and filter SIVT. Our research shows that 5.5% of traffic coming from Google channels was invalid, compared with 17.5% of traffic from other platforms (such as LinkedIn, Meta, and TikTok). Here’s the breakdown of IVT by channel:

cost of invalid traffic

Among ad platforms, Google is the market leader in dealing with IVT. Here’s an extract from Google’s IVT management policy:

“When our system detects invalid traffic with your ads, we automatically filter these out of your reports. You won’t be charged for them and you’ll still have the option to view the data. You’ll be credited for any invalid interaction that escaped our automated detection in the past two months.”

This refund system is clunky, but it’s miles better than other platforms. Take this extract from Meta’s advertising terms:

“Facebook shall have no liability for click fraud or other improper actions, or for invalid clicks or other technological issues, each of which may affect the cost of advertising.”

Leo explains why there’s such a difference between ad platform approaches to IVT:

Google is processing a lot more data than most other networks. Last year a report showed that Google was processing around 8.5bn searches a day. Meta sees roughly 22bn visits each month. To add to that, Google has an entire knowledge base on invalid traffic, showing advertisers how to prevent it and educating them on the problem. But keep in mind that most of this is dealing with GIVT. It’s SIVT that’s becoming the problem nowadays. Most social platforms like Meta and LinkedIn are not equipped to deal with sophisticated invalid traffic, whereas Google might be in a slightly better position.

Leo Pizarro

Head of Demand Generation, Lunio

This problem gets worse around the peak shopping season. Facebook’s rate of invalid traffic increased by 8% in October-December 2022, meaning more than a quarter of all traffic coming from Meta was invalid. 

Note: Download our Meta Retail Peak Season Report to learn more about how IVT affects Facebook advertising in the build-up to Christmas.

Ultimately, there’s not much financial incentive for ad networks to clamp down aggressively on SIVT. Analysing real-time traffic is expensive, and the costs ramp up the longer it takes their machine learning models to make a decision. And because SIVT is becoming increasingly hard to detect, the difference between seconds and minutes can lead to a tenfold increase in associated costs.

The Increasing Prevalence of SIVT

Thanks to increases in computing power and the advent of AI tools like ChatGPT, sophisticated invalid traffic is on the rise.

The 2023 Imperva Bad Bot Report found that in 2022, 47.4% of all internet traffic came from bots, a 5.1% increase on the previous year. And this trend is likely to continue as a result of Moore’s Law. David Triger, Lunio’s Chief Technical Officer, explains:

According to Moore’s Law, computing power nearly doubles every two years (with some adjustments). This means that the power of hardware is growing much faster than the population, or even the population that has access to the internet. This trend will continue for quite a while. If no preventative measures are taken, it’s quite likely we’re going to be in the single digits of real activity across the web in ten years time.

David Triger

CTO, Lunio

Both GIVT and SIVT are affected by Moore’s Law. But the introduction of ChatGPT and other AI tools now means people with very limited coding experience can create aggressive bots, further contributing to the SIVT problem.

Automated Campaign Types & SIVT

Advertisers using automated campaign types like Google’s Performance Max and Meta’s Advantage+ have seen increased IVT issues.

There are lots of known issues with PPC automation, including limited control over targeting, and less visibility in analytics. This creates conditions that make it much harder for advertisers to detect and mitigate IVT.

This is seen in the rise of fake or spam lead submissions stemming from automated campaigns. It’s common to get high volumes of fake leads when using Performance Max, in particular, because it forces B2B brands to use Google channels they wouldn’t normally use. This has led some advertisers to avoid PMax for the time being:

Meta Advantage+ users experience similar problems. Leo shared the results of an experimental lead gen campaign he ran on Advantage+:

We decided to run an experiment without Lunio to see how much Advantage+ had changed since its launch. It was a lead gen experiment, so we set a target cost-per-lead. Within 24 hours, we generated 200 sign-ups. We were amazed. But then, when looking at the cost-per-lead, something seemed off. Our cost-per-lead was £0.07. So we downloaded the list, and found 200 fake email addresses, all in the same format and with a similar timestamp.

Leo Pizarro

Head of Demand Generation, Lunio

Every lead generated was fake. Personal email addresses are especially hard to validate, leading to wasted time and energy in the follow-up. So there are still significant issues with automated audience targeting.

In addition, fake leads are usually much cheaper than real ones. This can create a negative feedback loop within your automated campaign, whereby the algorithm continues to seek out more junk conversions to make your acquisition costs lower.

4 Key Ways Invalid Traffic Affects Marketers 

Invalid traffic can have a serious impact on your revenue and business growth. Here are four ways IVT affects marketers.

1. The Direct Financial Cost of IVT

Every invalid click results in wasted ad spend. According to our research:

  • $12.35 billion was wasted on Google Ads in 2022.
  • $16.59 billion is forecasted to be wasted on Google Ads in 2024.
  • $42.28 billion was wasted on non-Google ad channels in 2022.
  • $54.78 billion is forecasted to be wasted on non-Google channels in 2024.

In total, we anticipate a 33% increase in wasted ad spend due to IVT from 2022 to 2024:

cost of invalid traffic

You may be able to claim some of this back from Google, but it’s an arduous process, and other ad networks don’t tend to offer refunds.

2. Lost Revenue Opportunity

If all invalid clicks converted at normal rates, revenue would be much higher. So it’s not just the actual wasted ad spend you have to account for — it’s the amount of lost revenue, too.

A study conducted by Nielsen revealed that the average return on ad spend across all industries is 2.87:1. When applied to our wasted ad spend forecast for 2024, this equates to $204.83bn in lost revenue opportunity for brands and advertisers in 2024.

3. Skewed Marketing Data

Advertisers need accurate marketing data to make good decisions about where to spend their budget. Unfiltered invalid traffic — whether it’s GIVT or SIVT — can invalidate your data, so you can’t rely on your analytics.

For example, say you see a jump in conversions following a recent campaign, and you want to scale up your ad investment based on this. Do you know that all these conversions are from real users? And if not, do you know what proportion of the traffic/leads is invalid? If not, can you be certain that increasing spend will give you good results?

Making bad decisions often leads to further wasted ad spend and stifled growth opportunities. One estimate suggests bad data costs the US $3 trillion every year.

Make sure to track conversion volume and actual revenue as well as the metrics in your ad analytics platform. This can give you a more accurate picture of the effectiveness of your ad campaigns.

4. Fake Lead Submissions 

Generating quality leads is a big concern for performance marketers right now:

cost of invalid traffic

Our recent Wasted Ad Spend Report found that more than two-thirds (69.1%) of performance marketers report fake leads from paid media campaigns. And this problem isn’t isolated to automated campaign types like PMax; all ad campaigns are vulnerable to lead gen fraud, with Display networks particularly affected.

The amount you stand to lose from fake leads is dictated by:

  • Time spent validating leads — Sales teams must validate and follow up on each lead, losing time they could have spent focusing on genuine prospects. This leads to lost productivity and lower revenue.
  • IVT scale and sophistication — Large, frequent, or sophisticated IVT attacks are more likely to cause serious or sustained losses.
  • How quickly you act — The longer it takes to discover and block the source of fake lead submissions, the more money, time, and energy you lose.

How Lunio Detects IVT Ad Platforms Don’t Catch

Google won’t fix your invalid traffic problem anytime soon. But Lunio definitely will. 

Through real-time analysis of every click on your paid ads, Lunio assesses the validity of every traffic source hitting your website. Clicks that come from known blacklisted IPs or exhibit multiple obvious signs of automated on-page behaviour (e.g. bouncing in less than a second or filling out forms faster than a human possibly could) are immediately marked as invalid and blocked from interacting with your client’s ads again in future. 

But in certain cases, it’s difficult to determine based on a single interaction whether the source is a bot or just an indecisive human being. And that’s where Lunio’s “suspicious” category comes into play. Rather than immediately blocking the source of a suspicious click, there’s a requirement for further investigation before a decision is made. 

Lunio will then assess hundreds of data points stemming from a few additional clicks from the same source until it has enough information to accurately categorise it as invalid or legitimate. This process virtually eliminates the risk of flagging real interactions as invalid. And over time this translates into greater spend efficiency and profitability across all campaigns. 

See How Much of Your PCC Traffic is Invalid

Lunio’s 14-day free trial includes a full traffic audit, which shows you exactly how much invalid traffic is affecting your PPC success (and how much you could save by eliminating it). There’s no obligation to buy when the trial period ends; getting a clear picture of your traffic health can help you decide on the best way to fix invalid traffic on your site.

Get a Free Invalid Traffic Audit

Ads are for humans, not bots. Get a demo and save up to 25% of your advertising budget by automatically eliminating fake ad engagements across all paid channels.

Related Posts

You may also like...

Stop All Advertising Fraud in Seconds