Video is no longer the future of commerce, but the price of entry. The increased demand for video was well under way even before the pandemic brought it to a tipping point. Lockdowns and store closures forced brands to find new ways to connect with consumers and video filled the gap. What served as an accelerant has become table stakes. Expectations around immersive, highly visual experiences were reset and there’s no going back.
Making video is the easy part
Video has always required special skills and been laborious and expensive to produce, even before adding on distribution and delivery costs. Digital commerce multiplied the number of contexts in which video needs to exist. A single asset needs to work across multiple channels – brand sites, retail channels and social platforms – and then it needs to be personalised and optimised for millions of individual customer preferences, devices, languages, regions and countless other factors. Prices change, campaigns come and go, stocks run out, events happen – and the emerging demands of agentic commerce add yet another layer of complexity.
Solvable by AI and automation
Given the scale of these challenges, it’s little wonder some brands have been slow to fully commit to video. A human production workflow alone can’t solve the scale problem – practically or economically.
Fortunately AI-powered tools, particularly image-to-video capabilities and other real-time content generation solutions, have arrived to transform not only how video is created, but scaled for the online world. Rather than having to produce multiple static variants, video assets can be dynamically adapted based on device, format, and context at the moment of consumption.
Scale globally, automate locally
One area where brands quickly hit a wall with manual processes is localising videos for global audiences. When dealing with a massive volume of long-form videos for international markets, extracting the original audio to generate and translate multi-language subtitles is just completely unmanageable without AI and automation. It’s an incredibly costly, time-consuming, and painful process.
By tapping into automated workflows, brands can instantly generate those localised subtitles and chapter points, ensuring they comply with accessibility regulations while drastically cutting operational costs. The friction and time of localisation have basically disappeared, allowing businesses to prepare their media for same-day publishing without getting burned by skyrocketing expenses.
Optimising Metadata for SEO and Agentic Search
On the backend, there are also important considerations as brands hit scale. When publishing thousands of videos, its metadata needs to be optimised. Both traditional search engines and emerging LLMs used for agentic shopping rely on video metadata to find products. Optimising might include for example correct tagging, adding titles, chapters, and descriptions. All of these activities can’t simply be done manually at scale. AI tools can help to automatically trigger this metadata and ensure your products are found by customers and bots alike.
Keep direct customer connections and tap into UGC
Platforms such as YouTube make video consumption highly accessible and intuitive, but the downside is that brands lose direct connection with their customers. On YouTube, the relationship is with the platform, not with your brand. This is also a missed opportunity for brands to harness user-generated content, which is regarded as more authentic than traditional brand-produced assets.
But managing and moderating user-generated content – especially at scale – comes with risks. Brands don’t want to just publish everything blindly. They need oversight to ensure the user-generated videos align with brand standards and mitigate potential risks.
While manual reviews may be feasible at small scale, it quickly becomes impractical in today’s environment. AI helps to validate any user-submitted videos for consistency, quality, and compliance. It functions like a dedicated brand reviewer, ensuring every asset meets a brand’s standards before it goes live.
Beyond e-commerce
The opportunities and challenges relating to video aren’t limited to e-commerce and retail. Content consumption patterns are changing across industries. Corporate training, for instance, is moving away from static, text-heavy materials toward concise, interactive video formats. Few workers, especially younger ones, won’t sit down to read through a 200-page handbook; they expect five-minute, interactive videos. Once companies reach a certain scale, automation is no longer optional, but essential.
From table stakes to winning
Despite the dizzying pace of innovation across every stage of the video lifecycle, we are still only just at the beginning. There are countless signs on the horizon that indicate how much the online customer experience is set to develop and drive demand. These include new innovations in AI-assisted shopping, ultra-fast delivery methods, immersive AR/VR shopping, and social commerce evolutions – to list but a few. And while simply offering video is table stakes, brands need to exploit automation and AI fully to achieve scale and win at the game.
Tali Rosman
Serving as leader of Cloudinary’s Video division, Tali is a hands-on executive who is passionate about harnessing technology to drive efficiencies and improve lives.
Before Cloudinary, Tali gained experience across positions at Amdocs, Xerox, and NICE Ltd, while also serving as an advisor and board member for various technology companies. Her global experience spans living and working in six countries across three continents.
Tali’s background features significant experience across strategy, mergers and acquisitions, and technology development . She holds an MBA in Business from INSEAD.
