GenAI for Video Marketing & Sales

GenAI Transforming Video Content Creation

Producing videos used to be costly and lengthy. The ability of smartphones to easily capture video, along with a decrease in viewers' expectations for top-quality output, led to a sharp decline in video production costs and time. Now, generative artificial intelligence has further reduced these requirements by transforming the entire creation process.

Video is still a useful tool for capturing attention and generating interest from target audiences. Companies intend to increase their investments in the technology. The Content Marketing Institute (CMI) recently discovered that 61 percent of marketers believe their organizations will invest more in videos. Use is also increasing. According to CMI, 42 percent of marketing organizations currently use AI to create videos, up from 18 percent last year.

GenAI Beyond Video Production

Chris Lavigne, head of production at video marketing platform provider Wistia, says genAI is also used in preproduction processes. He adds that genAI's planning, scripting, storyboarding, and ideation capabilities are more developed than text-to-video tools. The CMI report supports this. According to the report, most marketers use AI in preproduction (scripting and brainstorming) or postproduction editing (voice dubbing and generating visuals). Additionally, 80 percent of these users think AI helps streamline video production, enabling quicker turnaround times and better content.

GenAI for video production involves various applications. Quint Boa, founder of creative video agency Synima, explains that text-to-video tools use diffusion models trained on millions of video clips to generate visuals from prompts. Voice cloning employs neural networks trained on voice samples to re-create speech patterns, intonation, and emotional cues. Lip-sync avatars combine speech-to-text with facial landmark animation to sync voice with realistic face movement, Boa adds.

Lavigne notes that the technology is rapidly evolving, especially for inanimate objects. Human expressions, however, remain more difficult. Krish Mantripragada, chief product officer of Seismic, a sales enablement platform provider, says that context is crucial, especially in B2B environments. B2B companies should create content specific to their company and situation. Mantripragada says that AI must be grounded in enterprise content for B2B companies. This ensures that AI-generated content adheres to brand tone, company policies, industry regulations, and approved messaging. The technology can also incorporate business context to create relevant video content.

Lavigne recalls that transcribing video for documentaries used to require sending the video to a human editor and sifting through a text document to find the best quotes. Now, time codes are automatically linked to transcripts, making it easy to find specific moments. AI can analyze video transcripts and suggest edits automatically. Lavigne expects programmatic video editing to be available soon.

Mantripragada adds that genAI excels at creating multimodal learning content, accommodating different learning needs and preferences. He says that creating multimodal content based on learner preferences is a significant advantage of AI, as it was historically resource-intensive. GenAI tools can create audio and video lessons, podcasts, and deep-dive study guides from source content for specific situations. Users can interact with AI to ask probing questions, ensuring that it doesn’t provide generic information.

Limitations and Oversight

GenAI-created video is not perfect. Lavigne says that genAI video often defies the laws of physics. Additionally, genAI struggles to accurately portray human emotion and can cause face distortions. However, he adds that these limitations may be temporary.

Establishing guardrails is important for ensuring the appropriate use of genAI tools in video creation. Lavigne emphasizes the need to notify talent and recommends talent release forms to specify permission to synthesize their likeness. Human review of the content produced is also essential for ensuring accuracy and quality. Boa points out that AI cannot guarantee copyright safety, so clear policies on copyright checking are important for all AI-generated assets. Basic security protocols should also be established. CMI’s research shows that 66 percent of marketing organizations have security measures in place specific to the use of genAI.

Lavigne believes that AI is a tool that empowers rather than replaces content creators. He says that video producers must understand how to use these tools to improve their workflow. Mantripragada agrees, emphasizing that AI allows marketing, enablement, content creation, and training teams to quickly and cost-effectively create various content formats. He adds that users of AI video tools should have a basic level of comfort with the technology and understand prompt engineering. Content creators will still need to validate the quality of AI-generated content.

Mantripragada describes the current period as an exciting time for video creation, noting that genAI is revolutionizing every part of our jobs. Boa, Lavigne, and Mantripragada suggest that the technology is rapidly evolving. Linda Pophal is a freelance business journalist and content marketer. There are many genAI tools available for video production, with more emerging continually. Boa lists some major players in each tool category.