Within the sphere of technical marketing, the terms client-side and server-side could be described as follows:
A clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. is a machine that communicates with the server. The job of the clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. is to send requests to and then anticipate and process the response from the server. Clients are typically devices and software operated directly by the user, but sometimes a clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. can also be software running on a server itself.
Thus, client-side is used to describe the actions, mechanisms, and capabilities of clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. hardware and software, the requests compiled and sent by the clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers., and the handling of the responses received from the server.
A server is a machine that is designed to intercept and parse the request from the clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers.. There are many different types of servers, such as web servers, email servers, and file servers.
Thus, server-side denotes the activities that happen purely within a server environment. These include intercepting requests, parsing them, initiating new server-side processes (such as data pipeline ingestion), and compiling the response back to the request source.
As a technical marketer, you will frequently encounter these terms especially regarding data-related processes. You will need to understand the differences, the benefits, and the drawbacks of both approaches. There is no one-size-fits-all approach to data collection, and ultimately you will likely end up with a hybrid model facilitated by technologies like server-side taggingNormally, tag references an HTML element (or node). In a marketing context, tags are used to denote HTML elements and JavaScript snippets specifically designed for collecting data to marketing vendors..
Client-side analytics
Client-side analytics typically represents the work of a trackerSoftware that typically runs in the user's web browser or device, designed to collect data from the user to a server..
This trackerSoftware that typically runs in the user's web browser or device, designed to collect data from the user to a server. runs in the web browser or the mobile app, and its job is to collect data about how the user is operating the client-side software.
The trackerSoftware that typically runs in the user's web browser or device, designed to collect data from the user to a server. can access anything that is made available to its code. In a web browser this would be strictly JavaScriptJavaScript is the main language of the dynamic web. The web browser renders the HTML source file into a dynamic document that can be interacted with using JavaScript. and browser APIsStrictly speaking application programming interfaces are methods and protocols in a piece of software that allow other sources to communicate with this software. More broadly they are used to describe any functions and methods that operate how the software works., but in mobile devices the language and the available APIsStrictly speaking application programming interfaces are methods and protocols in a piece of software that allow other sources to communicate with this software. More broadly they are used to describe any functions and methods that operate how the software works. depend on the operating system and which SDKsSoftware Development Kit. Mobile apps use SDKs to run most of their functionality. SDKs are analogous to JavaScript libraries on web pages. have been deployed.
These are typical things available only in a client-side context:
- The content that is visible to the user on the page or screen.
- Information stored behind client-side APIsStrictly speaking application programming interfaces are methods and protocols in a piece of software that allow other sources to communicate with this software. More broadly they are used to describe any functions and methods that operate how the software works., such as those of the web browser or of the device.
- The user’s interactions with the browser or device, such as clicks and taps, mouse movement, scroll actions, time spent with the screen in the foreground, etc.
When an analytics platform is solely concerned with what happens client-side, it is a client-side analytics tool.
In client-side analytics, the user is the centerpiece, and their actions in the client-side environment form the basis of the analytical approach.
Strictly client-side tools include:
- Screen recording software, which observes the user’s visible viewportThe area of the web page or app visible to the user at any given time. It is constrained by the size of the browser window and the resolution of the user's device. and tracks their mouse movements and other interactions on the page.
- Heatmapping and eyetracking software, which present overlays of the page with heat signatures that represent the areas users mostly interacted with or watched.
- Page performance monitors, which measure the time it took to load the page and the elements on the page.
Tools like these are concerned explicitly with the information stored and presented in the user’s device. They would be impossible to replace with a pure server-side solution.
Don’t miss this fact!
There are many things that a client-side trackerSoftware that typically runs in the user's web browser or device, designed to collect data from the user to a server. does that cannot be reliably replaced with a purely server-side solution. Without any tracking code running in the user’s device, it’s impossible to know, for example, the user’s interactions with a piece of content.
Server-side analytics
Server-side analytics is a confusing term.
The most straightforward interpretation of the concept is trackerless analytics. That is, there is no software running to detect the user’s activities in a client-side environment.
Instead, server-side analytics consumes data and information produced purely in a server environment.
Typically, this boils down to log analysis, server-to-server pipelines, and product telemetry.
In “pure” server-side analytics, the focus of data aggregation isn’t necessarily the end user but rather logs stored in a file server or telemetry from other servers and products.
This is still a popular approach when measuring service uptime and performance, or when analyzing aggregates such as the increase of subscriber counts in the CRMSoftware for managing all your organization's relationships and interactions with customers and potential customers..
There are flavors of pure server-side analytics where the user is the focal point, however. In these cases, user data is collected from the CRMSoftware for managing all your organization's relationships and interactions with customers and potential customers. and from logs, and their behavior is extrapolated from the server-side metadataMetadata is additional data about the data itself. For example, in an analytics system the "event" describes that action the user took, and metadata about the event could contain additional information about the user or the event itself. rather than from directly observing their actions in the client-side context.
However, server-side analytics is these days used to describe a more hybrid approach. The term “server-side” doesn’t necessarily mean that the entire pipeline needs to be located within server contexts.
Instead, “server-side” is used to describe a data process where the request to the analytics server is dispatched from another server but doesn’t necessarily originate there.
Ready for a quick break?
Now’s a good time to do some client-side maintenance. Take a walk and a cup of water to make sure that your data flows are healthy!
Hybrid approaches (server-side tagging, for example)
With the increased availability of server-side taggingNormally, tag references an HTML element (or node). In a marketing context, tags are used to denote HTML elements and JavaScript snippets specifically designed for collecting data to marketing vendors. and proxies, the lines between client-side and server-side analytics have become blurred.
A server-side proxy means that there is a server-side endpoint that consumes the requests from the clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. (or some server-side process) before the request is forwarded to the vendor server. This proxy can be a very simple, transparent forwarding mechanism solely to prevent the clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. from directly communicating with a vendor server.
A common but ethically questionable use case for proxies is to change the URLUniversal Resource Locator, the main method of encoding internet addresses for web browsers to send requests to. of the analytics server from a vendor-owned domainDomain names are stored in the domain name system (DNS) and are used to map a human-readable name (such as www.teamsimmer.com) to a web server. to one that the company controls. Many ad and content blockers target these vendor-owned domainsDomain names are stored in the domain name system (DNS) and are used to map a human-readable name (such as www.teamsimmer.com) to a web server. when determining if the request should be blocked or obstructed. By moving the endpoint to a custom domainDomain names are stored in the domain name system (DNS) and are used to map a human-readable name (such as www.teamsimmer.com) to a web server., blockers will no longer block the traffic. This is ethically questionable because it erodes the user’s agency to choose what data is blocked and what isn’t.
The proxy can also be a more complicated piece of machinery, with full validation and enrichment capabilities. Server-side taggingNormally, tag references an HTML element (or node). In a marketing context, tags are used to denote HTML elements and JavaScript snippets specifically designed for collecting data to marketing vendors. is an example of this type of proxy.
Server-side taggingNormally, tag references an HTML element (or node). In a marketing context, tags are used to denote HTML elements and JavaScript snippets specifically designed for collecting data to marketing vendors. moves the logic of tagNormally, tag references an HTML element (or node). In a marketing context, tags are used to denote HTML elements and JavaScript snippets specifically designed for collecting data to marketing vendors. management to the server. The incoming request from the client-side environment represents the “triggerA tag management system uses triggers (also known as rules) as activation instructions for tags. When trigger conditions are satisfied, any tags that reference that trigger can fire. event”, and the payloadIn an analytics system, the payload is used to describe the data in the request address or the request body, specifically designed to be associated with the analytics system’s schemas at processing time. of this request is what is used to initiate a variety of processes in the server-side taggingNormally, tag references an HTML element (or node). In a marketing context, tags are used to denote HTML elements and JavaScript snippets specifically designed for collecting data to marketing vendors. context, such as:
- Distributing the data in the incoming request to multiple vendors (“tags”) – even those unrelated to the clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. software that sent the incoming request.
- Validating and enriching the incoming request data before it is made available to vendors.
- Initiating data processes, such as by writing the data directly to a data warehouseData warehouse is a repository of data collected by an organization from different sources. The data can then be transformed within the data warehouse before being made available for querying against. or a real-timeReal-time analysis refers to analysis of data that is currently being collected. For example, a publishing media might use real-time data to see how many people are consuming content at any given time. "Real-time" is never really real-time – there's always a latency of at least some milliseconds, usually seconds or even minutes. pipeline.
- Removing sensitive information from the incoming request, and sanitizing and normalizing the parsed data before it is forwarded to vendors.
Even without server-side taggingNormally, tag references an HTML element (or node). In a marketing context, tags are used to denote HTML elements and JavaScript snippets specifically designed for collecting data to marketing vendors., the concepts of client-side analytics and server-side analytics overlap greatly in modern analytics stacks.
Data collected from client-side software is often enriched in the vendor servers by integrating it with information from other sources. This is what Google does when joining analytics data with ads data, for example.
Data collected directly from server-side processes is rarely analyzed in isolation – instead, it can be used to enrich existing client-side flows. For example, a common use case is to collect business critical data with server-to-vendor flows, so that it is impervious to client-side mechanisms designed to restrict data collection (ad and content blockers, consentData protection laws often defer to positive consent for collecting or processing data from the user. Requesting consent usually involves a consent banner or a consent pop-up where the user is asked whether the site or app can collect data from them. mechanisms, network failures).
Deep Dive
Different types of data flows
These days, analytics data can be collected through a wide variety of different combinations of client-side and server-side measurement.
In fact, just a plain old client-to-vendor stream, where data is collected directly from the user’s device and processed directly into the vendor servers, is becoming quite a rarity.
Although tools like Google Analytics 4 certainly allow you to collect data directly to Google servers, there’s a lot of enriching happening at processing time, which can expand the data that you collected from the clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. device.
Here are some of the most popular types of data flows utilized these days. But these are greatly simplified – in reality, many complex data pipelines incorporate multiple “stops” between different clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. and server machines.
- Client-to-vendor: the traditional “client-side analytics” approach, where data is collected by a client-side trackerSoftware that typically runs in the user's web browser or device, designed to collect data from the user to a server. and sent directly to the vendor servers.
- Client-to-server-to-vendor: here, the bulk of data originates in the clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers., but it is first forwarded to a server (typically owned by the same company that runs the clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. environment), where it is enriched and validated before forwarded to the vendor servers. This is an example of server-side taggingNormally, tag references an HTML element (or node). In a marketing context, tags are used to denote HTML elements and JavaScript snippets specifically designed for collecting data to marketing vendors..
- Server-to-vendor: traditional “server-side analytics” approach, where the data originates in a server context, from where it is sent directly to vendor servers.
- Server-to-client-to-vendor: in some rare cases, the server-side environment needs the clientA machine connected to the internet that sends requests to a server. A web browser would be a typical example of a client, as it sends requests to web servers. to do some work. Typically, this is to enrich the data payloadIn an analytics system, the payload is used to describe the data in the request address or the request body, specifically designed to be associated with the analytics system’s schemas at processing time. with information from third-party cookiesCookies are a way to persist information on the web from one page to the next and from one browsing session to the next. They are small bits of information always stored on a specific domain, and they can be set to expire (self-delete) after a given amount of time., which are only available in a client-side context.
Key takeaway #1: Client-side represents the data that is collected
A “client” is a machine that sends requests or dispatches to a server. As such, it can mean many different things from devices to smart appliances to server machines, too. In digital marketing, “client-side” is usually restricted to mean what happens in the user’s device and specifically the web browser. Client-side represents the user who visits a site or uses an app. If technology runs client-side, it runs in the user’s device.
Key takeaway #2: Server-side represents machines talking to each other
The term “server-side” is usually used to describe anything and everything that happens in a server environment without involving a client-side component. For example, dispatching requests directly from a CRMSoftware for managing all your organization's relationships and interactions with customers and potential customers. to an analytics system would represent a server-side flow. Similarly, parsing application logs and collecting them to a databaseStructured storage for data that usually serves a singular purpose. For example, a company's financial records would be stored in a database. would also be an example of a server-side process.
Key takeaway #3: Server-side tagging muddies the waters
Hybrid approaches like reverse proxies and server-side taggingNormally, tag references an HTML element (or node). In a marketing context, tags are used to denote HTML elements and JavaScript snippets specifically designed for collecting data to marketing vendors. add ambiguity to the terms “client-side” and “server-side”. These setups typically involve a client-side component as the dispatch system, but the data is then funneled through a server-side network, where it’s enriched, consolidated, normalized, transformed, and validated before it’s forwarded to vendor servers. So it’s a combination of client-side and server-side data processing.