How to spot and filter referral spam in Google Analytics

Referral spam is fake traffic that shows up in your Google Analytics without a real visitor ever having been there. You usually recognise it by strange referring domains with a bounce rate of 100% and a session duration of 0 minutes 0 seconds. The short version: filter that traffic out, because as long as it sits in your reports, you are making decisions on polluted data. In this article you will read how to spot referral spam in Google Analytics 4, how to filter it out, and why it costs more than you think for B2B lead gen.

What exactly is referral spam?

Referral spam (sometimes also called ghost spam or crawler spam) is traffic that appears in your referral reports but has no real interaction with your site. With ghost spam, the bot never touches your website at all: it sends data straight to your analytics via your measurement ID. With crawler spam, a bot does visit your pages, but it is not a potential customer, it is software combing the web.

In both cases the result is the same: a line in your report that looks like a referring site, but that will never bring you a lead. The sender usually wants you to get curious and visit the domain, or is simply trying to score backlinks and attention.

For a B2B company steering on leads, that is not harmless noise. It sits in the very reports you look at to decide which channels work and where your budget goes.

How do you recognise spam traffic in your reports?

The clearest signal is a combination of extreme values. Spam traffic often shows up with a bounce rate of 100% and a session duration of exactly 0 minutes 0 seconds (based on hands-on analyses of referral reports). A real visitor, even one who clicks away fast, almost never leaves such a perfect zero pattern behind. If a source combines 100% bounce and 0 seconds with a suspicious number of sessions, you are almost certainly looking at spam.

Watch out for these characteristics as well:

• Strange domain names in your referral report that you do not recognise and that have nothing to do with your market.
• Sudden spikes in referral traffic from countries or sources where you have never been active.
• Not a single conversion from that source, ever, despite sometimes high session counts.
• Illogical language or hostname values, for example referrals arriving on a hostname that is not yours.

The pattern matters more than the exact number. A source that delivers plenty of sessions but zero engagement and zero leads does not belong in your decision data. If you want to dig deeper into signals like these, our data analytics service helps you keep your numbers structurally clean.

Why is referral spam a problem for B2B?

Here is the honest core of it. Polluted data is worse than no data, because you trust it. If spam inflates your session counts, a channel looks more successful than it is. Your bounce rate gets skewed, your average session duration drops, and your conversion rate falls on paper while nothing has actually changed.

At Customer Impact we steer on customers and revenue, not on vanity metrics. Spam traffic is the ultimate vanity metric: it looks like visits, but it never moves anyone towards a quote or a meeting. If you split your marketing budget based on session volume, you risk shifting money to a channel that is in reality full of fake traffic.

For B2B lead gen, where every lead has value and the numbers are smaller than in a webshop, that pollution weighs far heavier. A few hundred spam sessions go unnoticed in an e-commerce account with millions of visitors, but in a B2B account they can tip your entire picture. It directly affects your conversion rate and with it your decisions about CAC. Good conversion tracking therefore starts with a clean dataset.

How do you filter referral spam in GA4?

GA4 differs fundamentally from the old Universal Analytics. There is no classic view filter button anymore like there used to be, so the approach is different. A few layers work together:

1. Use the built-in list of known bots and spiders. Google Analytics 4 excludes traffic from known bots and crawlers by default. This is your first line of defence and is usually already switched on, but it does not catch everything.

2. Define your own internal and unwanted referrals. In the data settings of your data stream you can specify lists of unwanted referrals, so that known spam domains are not counted as referral traffic.

3. Work with explorations and filters at analysis time. In the exploration reports you can build segments and filters that combine hostnames, sources and engagement signals. That keeps suspicious traffic out of your analysis, even when it is already in your dataset.

4. Filter after checking, do not pre-emptively block everything. Only exclude what you genuinely recognise as spam. Filtering too aggressively can cut away real traffic, and that is just as damaging for your decisions.

A common mistake: thinking you set this up once and then you are done. You are not. Spam is a moving target.

How often should you update your filters?

Plan this as recurring maintenance, not a one-off job. A workable rule of thumb is to update your filters and exclusion lists monthly, because new spam domains keep appearing, as Search Engine Journal recommends in its guide on filtering referral spam. A list that was complete last quarter is already missing this month’s newest senders.

Make it a fixed moment: once a month, open your referral report, flag new suspicious sources, and top up your exclusions. That is exactly the kind of small, fast intervention where an agile team makes the difference. You do not need to build a big process, you just need to do it consistently.

Combine this with reliable SEO reporting and a fixed reporting cycle, so that clean numbers become the norm instead of the exception.

Frequently asked questions about referral spam in Google Analytics

Does referral spam hurt my SEO or my rankings?

No. Referral spam sits in your analytics, not in the way Google ranks your site. The problem is purely that it pollutes your reporting data and can tempt you into the wrong decisions. The damage is internal, in how you read your numbers.

Can I remove spam from my data retroactively?

In GA4 you mostly filter forward and at analysis time. Traffic that has already come in stays in your raw dataset, but you can exclude it during analysis via segments and filters in your explorations. For clean historical reports you therefore work with filters at the moment you look at the data.

How can I be sure a source is spam and not a real referrer?

Look at the pattern, not at a single number. A source with 100% bounce, 0 seconds session duration, not a single conversion and a domain that has nothing to do with your market is almost always spam. If you are in doubt, exclude it in a test segment first and see whether your real numbers start making more sense.

Is this only a problem for large sites?

Quite the opposite. Smaller B2B accounts have less traffic, so every hundred spam sessions skews the picture more strongly. For lead gen with limited numbers, a clean dataset is relatively more important than for a busy webshop.

Ready to trust your numbers again?

Referral spam is not a disaster, but it is insidious: as long as it sits in your reports, you are steering on polluted data and shifting budget based on fake visits. Recognise the pattern, filter it in GA4, and update your lists monthly. Do you want your analytics to simply be right, so that your decisions rest on real leads and real revenue? We set up your tracking and reporting so that you can trust your numbers again. Book your free intake.