Website Statistics

Monthly Human Traffic Data

The RP Photonics website consists of four parts:

We carefully monitor the evolution of human web traffic of our website — which has become more difficult due to increasing activities of robots, some of which try to appear like human users (for example, scaping our content for unauthorized training of AI models). We apply the best available methods to generate reliable and trustworthy data.

In 12/2025, we recognized a substantial traffic surge, which we found to be caused mainly by some Chinese bot activity, and which could not be fully filtered out based on the recorded data. This trend continued strongly in 01/2026. Therefore, we decided in the beginning of 2026 to apply a new type of more aggressive traffic data filtering based on additional recordings of various activities (with Javascript) as evidence for real humans. For example, we statistically analyze mouse movements and can recognize “unnatural” movements, e.g. with constant speed in an exact direction. Other signals are obtained from scrolling and clicking events. With great effort, we developed a rather sophisticated filtering system and believe that the resulting data are quite realistic. There might be a few bot activities still falsely categorized as human, while others are wrongly discarded, but this is estimate to introduce only a small error (at most a few percent in one or the other direction). We surely do not include any traffic from known AI bots like ChatGPT-User, on which we comment below; these clearly signal their identity.

Applying this new filtering lead to significantly lower page view numbers. Below, we now display page views only from 01/2026 on, based on the new system. It includes only traffic on Encyclopedia and Buyer's Guide, not from other parts of our website.

Why more traffic on the Encyclopedia: Typical visitors are seeking technical information on most days, and look up suppliers only on some days. Therefore, they naturally cause more traffic on the Encyclopedia than in the Buyer's Guide. However, as regular users of the Encyclopedia, many will remember where to find the suppliers when these are needed. Also, there are enhanced entries of advertisers which even allow displays on the pages of Encyclopedia articles.

Traffic by AI Systems

AI systems such as ChatGPT and Perplexity have (mostly since 2025) begun to access RP Photonics pages at a substantial scale (many thousand times per day) — primarily the Encyclopedia, but also the Buyer’s Guide with substantial volumes. This reflects the recognized quality and authority of RP Photonics, which has become a primary knowledge source for AI systems when serving photonics professionals worldwide.

This development provides a significant additional benefit for our advertisers: Many further users are exposed to RP Photonics content — including advertiser product descriptions — indirectly via AI-assisted research workflows. Rather than primarily referring users to our website, AI systems mostly integrate RP Photonics content directly into their responses, that way also extending the reach and influence of both editorial content and advertising placements.

An essential aspect of the advertising value is that the displayed product descriptions form part of the substantive content that AI systems actively retrieve when users inquire about suppliers, technologies or components. In contrast to conventional display advertising formats, which are typically ignored by AI systems, these content-based listings are directly integrated into AI-assisted research and supplier discovery workflows. Here, the RP Photonics concept is very successful: We do advertising by informing users, i.e., by serving them what they are searching for, rather than by inserting ads where they only disturb.

The following graphic shows only page loads of ChatGPT-User, the agent used in dialogs of users with ChatGPT:

The most relevant traffic for our advertisers is the referral traffic to their own websites — people sent there through links on the RP Photonics web pages. The following diagrams show how many referrals are obtained in one month by (a) customers with our Advertising Package and (b) suppliers having only free entries in our Buyer's Guide.

You can see that our Advertising Package strongly increases the referral traffic, and that the larger suppliers (in terms of the number of registered products) can obtain many hundreds of referrals (eventually leading to sales leads) per month, i.e., thousands per year — which also leads to a very favorable outcome in terms of value for money.

That referral traffic is also the kind of traffic which advertisers can verify by comparing the data with their own server statistics. When that is done, one can and should also compare with traffic from other photonics websites. Our advertising customers usually report

• that traffic from RP Photonics is far stronger than that from any other photonics resource (even when they pay more for advertising there), and
• that traffic numbers reported by us are compatible with their own statistics, while statistical data presented by others often turn out to be highly inflated.

Some notes on traffic quality:

• Traffic quality from various photonics buyer's guides is probably quite similar (on a high level), as they are basically used by the same type of audience: photonics professionals searching for products. Traffic from search engines is usually of far lower quality.
• Caution: When measuring “engagement” in the form of how many pages users look at, you may be misled. For non-advertisers, we lead users to the supplier homepages, from where they need to find the product offer. For advertisers, we lead them directly to exactly the right target pages. As a result of the latter, users often stay there, rather than move on to other pages — which may look like lower engagement — but in reality, they are just already on the page they need!

By the way, you may check the referral traffic on your own server.

The RP Photonics website is probably the most popular photonics website worldwide. We are very transparent on our web statistics, while our major competitors don't publish their data and only give vague indications. This strongly suggests already that transparency would not be in their favor.

The strongest indication for us are the reports of various advertising customers that they get many times more referral traffic to their website from us than from any of our competitors. (We hear that consistently; nobody ever reported the opposite.)

While a few other platforms such as photonics.com have even more website content than us — typically, by offering thousands of articles which appeared in trade journals in recent years –, our content has the crucial advantage of being strongly linked with our Buyer's Guide. In addition, the quality of trade journal articles — often written by advertisers — is clearly lower than ours with encyclopedic quality. Therefore, the traffic on our pages is far more relevant than any traffic on some only loosely related lower-quality articles.

If you want to know how we achieved the amazing popularity of our resource:

As we do not require our users to register with us, we have no detailed knowledge of them. (We also highly value data privacy.) However, it is perfectly clear that basically all users are experts working in the area of photonics — in industry or in scientific research. Due to the very technical nature of our website and its devotion to the area of photonics only, nobody else would spend time there.

In May 2015, we did a survey among the readers of the RP Photonics Encyclopedia. Within 24 hours (from Wednesday morning to Thursday morning), we obtained 1456 answers. The obtained results:

This confirmed our expectation that the vast majority of readers are professionals in the field of photonics, working either in scientific research or in the industry.

In June 2015, we did a similar survey among the users of the RP Photonics Buyer's Guide, asking what word best describes their function. We obtained the following results:

This shows that our buyer's guide obtains traffic not primarily from marketing people who simply want to check their own entries, but from those user groups which are most likely to buy photonics items.

Geographically, users are spread worldwide. Due to the use of the English language, the resource is accessible for a large proportion of the worldwide population, particularly for technical experts.

Obtaining reliable web traffic data is not a trivial thing; we do that in a very careful way, which may lead to lower traffic numbers and respects the privacy of the users.

Standard server statistics can be spoiled by the activity of various kinds of robots as used by search engines and spammers. To discard anything in the statistical data which is not related to real traffic, since 05/2013 we are using a statistics software which applies very strict filtering rules:

• Only when Javascript is executed in the browser, a log entry is created. Therefore, most search engine robots are not counted. The few others who deal with Javascript are detected, and the results of their activity are ignored.
• If a single machine (more precisely, a single IP combined with a user agent) is doing too many page views within a short time interval, that whole traffic is discarded, as it might be some robot activity.
• If a user loads the same page which he or she has just loaded before, only a single view is counted.
• Views from any computers within RP Photonics are also discarded.
• Of course, we would never engage with “click farms”, paying people for generating fake clicks.

Note that we operate strictly in accordance with the strict German data protection laws. In particular, we do not record the users' IP addresses, but only store encrypted data, from which the IP values cannot be reconstructed (at least not without excessive computation power).

If we used Google Analytics instead of our own software, we would presumably obtain significantly higher traffic numbers, as Google Analytics would not filter out false hits (e.g., multiple hits to the same page) similarly strictly. We do not use Google Analytics due to concerns about privacy matters and related legal uncertainties.

The previously used AWStats software also reported substantially higher numbers than our new system. We conclude that our system provides a fairly conservative estimate of the web traffic.