In our increasingly visual online economy, the once hyperbolic saying that “a picture is worth a thousand words” is, if anything, now an understatement. There’s plenty of research on how images and videos to boost online engagement and lead to increased sales conversion. However, even the most visually stunning website won’t drive these outcomes if images and videos load too slowly or, worse, not at all. In fact, when load time increases from 1 to 3 seconds, the number of people leaving a page, or bounce rate, increases by 32%.
It’s for these reasons that Google’s relatively new “Core Web Vitals” metrics assess a website’s load time as part of establishing its search ranking. The metric, Largest Contentful Paint (LCP) refers to how long it takes for the largest “above-the-fold” (visible on a website without having to scroll down) element to appear. According to the HTTP Archive, in 70.6% of mobile pages and 79.4% of desktop pages, this is an image. To perform well, Google recommends keeping the LCP under 2.5 seconds.
This creates a dilemma: websites must contain beautiful, authentic imagery and videos to attract visitors and sell products. However, if those high-quality (often large) visual elements don’t load fast, Google penalises your search ranking. Furthermore, you could also lose customers. Solving this means optimising visual media quality, loading speed and website speed – variables that often conflict. Fortunately, new tools for optimising visual content along with a multi-CDN approach can help enterprises to strike the right balance.
Image and video optimisation
According to the HTTP Archive, unoptimised images are the chief cause of web page bloat. Images load slowly because they are not compressed enough (or at all) or are in an old, heavy file format. Developers tend to err on the side of protecting image quality when compressing files and often assign higher quality values than necessary. This is one area where machines excel over human instinct. AI-based tools are now available that can automatically set the optimal file size and visual quality for an image ‘on-the-fly,’ depending on a web visitor’s device, browser, and connection speed.
Another easy way to reduce page bloat is to use lighter-weight image formats and video codecs. It may surprise some to know that the trusty and reliable JPEG and PNG formats are more than 30 years old when websites were dominated by text. Newer image formats like WebP, AVIF, JP2, HEIC and JPEG XL were built with the modern web in mind. These formats cut page load times and bandwidth requirements significantly. Using JPEG XL over JPEG, for example, reduces bandwidth requirements by 25 to 30%.
The downside of newer formats is that not all browsers and devices support them yet. So, you need to make sure that the file formats you choose display well on all your users’ devices. Again, AI-based optimisation tools automate this process, always selecting the format that best matches a user’s requirements.
The same is true for video codecs. For example, the new format AV1 was designed specifically to improve video transmissions over the Internet. It also offers a tremendous advantage for image file compression. AV1 enables as much as 50% or higher bitrate savings benefit compared to the current Internet codec standard, H.264.
Under the hood: content delivery
Slow content delivery and loading times still happen due to the scale of the Internet itself and its underlying architecture.
Why is that? The Internet is a decentralised ‘network of networks’ linked together by TCP/ IP protocols. You never know by which route your content will travel through it and whether it’s the fastest. That’s why Content Delivery Networks (CDNs) – distributed server systems across multiple data centres, with an origin server at the centre – were developed. This setup forms a protective layer, guaranteeing that the origin server is protected from attacks. As a result, they ensure that content doesn’t get stuck in “traffic jams.”
There are a range of robust, proven CDN services on the market, each with advantages and disadvantages. A key feature to consider when devising your CDN strategy is to make sure all your target geographies are served well. Unfortunately, it’s difficult to find a single CDN covering all regions. Even if you find one provider that offers generic coverage, it is unlikely to have enough servers situated close to all your target audiences. This means you wouldn’t be able to ensure a consistent content delivery experience.
In 2021, the consequences of relying on a single CDN service were all too apparent. A series of high-profile outages from major CDN providers took down sites like Amazon, Spotify, Southwest Airlines, and many others. This shows the inherent risk of a single CDN approach.
The Multi-CDN approach ensures content is delivered using the server closest to your user or the one best suited for your content needs. And if an outage hits one provider, it automatically switches to the next best CDN.
Don’t attempt, however, to manage multiple CDNs manually. The configuration complexities, ongoing maintenance overheads, and costs can easily spiral out of control. Until recently, a multi-CDN strategy was only available to online brands like Facebook and Netflix, with massive in-house development teams.
Fortunately, now there’s a solution for smaller companies needing to optimise visual media for the web – automation. Our customers needed us to solve the problem by giving our premium customers access to dynamic multi-CDN switching. This capability uses real-time data to auto-select the best CDN for every context. It directs users to an optimal CDN edge server, avoiding outages or low availability. It’s effectively a huge, globally distributed network that delivers top performance as efficiently as a slime mould. This means you no longer have to choose between a beautiful website that attracts visitors and a high performing one that keeps them engaged.
A best practice put to work: Vorwerk
Germany-based Vorwerk needed to deliver high-quality images and videos quickly and flawlessly to millions of customers of its Thermomix smart kitchen appliance. This was managed via its “Cookidoo” recipe software platform – which hosts more than 66,000 digital recipes. Crucially, Vorwerk needed to deliver a consistent user experience for viewing images and videos across the 42 countries it serves. This posed a particular challenge in the critical Chinese market, where content must be delivered through a local CDN. Vorwerk reports that by working with Cloudinary, it’s delivering exceptional customer experience across all devices and served markets, including China. Cloudinary enables Vorwerk to deliver image and video content fast, without interruptions, on all its devices.
Cloudinary’s mission is to empower companies to deliver visual experiences that inspire and connect by unleashing the full potential of their media. With 60 billion assets under management and nearly 10,000 customers worldwide, Cloudinary is the industry standard for developers, creators and marketers looking to manage, transform, and deliver images and videos online. As a result, leading brands like Atlassian, Bleacher Report, Bombas, Grubhub, Hinge, NBC, Mediavine, Minted, Peloton and Petco are seeing significant business value in using Cloudinary, including faster time to market, higher user satisfaction and increased engagement and conversions. For more information, visit www.cloudinary.com.