AI for video editing - a viable use case for your B2B content strategy?

---

B2B content strategy is something of an enterprise oxymoron. The strategy is there, but often the content is not.

Video is a promising area for reaching the Informed B2B Buyer, but too often the content is stilted and overscripted. Or, the content is quite good, but it gets buried in an event archive, hard to find after the show.

Brent Leary of the Playaz Production Network knows a thing or two about B2B video that engages audiences. I've talked to Leary extensively about the pressing question: can "AI" make video content production easier?

In a recent live video show on LinkedIn, Leary and I broke down the ingredients of quality B2B video content - and whether AI for video editing can help. (You can see the full video replay on LinkedIn, or check the audio podcast below). During the video, Leary vented out on the plight of B2B video content, and how we can do better:

The best part of these conversations are the organic parts that aren't guided.

Indeed. The full quote:

It's because there's a need for folks to feel like they're in control of the content. And because of that, they don't trust the people in front of the camera to deliver compelling, interesting on-point content. They feel like they have to try to guide it. And the best part of these conversations are the organic parts that aren't guided. They just come out from the give and take the normal interaction.

For the video highlight clips, I demoed a bunch of "AI" related services. I produced a couple highlight clips with OpusClip, including the aforementioned Leary quote:

(OpusClip was originally launched for social influencers, so it just has a vertical phone-style clip format, but I'[m told a landscape clip option will come in the September timeframe).

I could devote an entire post to  "what makes B2B video great." But today, I'm going to hone in on the AI for video editing topic. It's a huge pain point; it also sheds light on the potential (and limits) of what AI can do today.

The business case for AI video editing

I do a fair amount of live, interactive B2B video events. The attendee chats are vigorous and the events run long, often a bit more than an hour. Unpredictable, unscripted events keep people engaged, but: watching the full replay is a different matter. Yes, if you're deep into ESG, you'll probably want to watch my full ESG research reveal show with Brian Sommer, but if you're not, watching a full hour video replay, where you are not part of the live discussion, is a big ask.

In the past, I handled that mostly with my audio podcast option (aka my Busting the Omnichannel series). Audio versions lack video cues, but at least it doesn't tie you to watching the replay on a computer or phone. You can putter or listen while commuting, etc. But I've wanted to bear down on highlight clips, something Leary does a terrific job of. But there are two differences: Leary is a better video editor than I am, and he seems to get something useful out of the video editing process - at least more than I do.

Could you outsource video editing to a human expert? Sure - but as Leary and I discussed, even most human editors won't necessarily know what the most important B2B topic is on replay. Then you have the inevitable delays of handing work off to the human editor, and the tedium of transferring files.

When would I use a human/expert video editor? When I wanted a super-slick, 2-3 minute highlight video, such as an event reel or keynote highlight package. Sometimes these highlight reel clips involve splicing 20+ pull quotes together, in a super fast-paced format. Back in the day, the now-retired Den Howlett produced an event video where we asked each attendee to sum up an event in one word. The final product went down like hot cakes, but the editing was a beast.

There is a more modest/typical B2B editing use case: B2B companies tend to produce longer videos like webinars, interviews, or event content. This editing use case is less about a super spiffy highlight reel, it's about:

Pulling a few 30 - 120 second clips from a longer conversation, perhaps splicing a few of them together, and getting them out on social channels.

Doesn't that sound like something today's "AI" could do? For the purposes of this show, I tried several AI for video editing services. There are a slew of them out there. The ones I am interested in are not really editing helpers, as much as social clip producers. To get the most out of these services, I believe you still need a fallback video editing environment (and someone who can use it where needed). In my case, I use Vimeo's online video editor, which allows me to do things such as:

• Quickly combine a few shorter social clips into one video
• Iron out a rough edge, such as clipping a longer segment
• Add an intro and outro if I want (not all AI social clips services give you this option)
• Tweak the layout format (example: for vertical OpusClips, I add a black background, such as this clip with Leary on what makes B2B video entertaining)

The pros and cons of clips services: OpusClip, Wisecut and Pictory.ai

When evaluating AI services, be wary of the what I call the "AI overreach." This is when the AI system is making too many decisions without your input, leaving you with flawed output. All the social clips service are a work in progress; they are still refining and adding features. I tried one clips service that just dumped a bunch of 30-60 second social clips for my use, but without any ability to influence the output. (Some of these services limit either the video file size or video duration upload length, either on free/entry plans or mid-tier plans, which adds friction).

For this type of auto-generated clips service, I really like OpusClip, enough that I've become a paid subscriber. OpusClip allows users to include keywords from the clips as well (before the AI gets to work), and specify the desired clip length. I also produced this Leary clip on what makes B2B content entertaining with OpusClip:

Now, OpusClip is typical of the social clips services: the "AI" is going to make some decisions on which aspects of your video to highlight. Therefore, while these clips are super convenient, they may not contain the exact part of the video you were thinking of capturing (nor can you combine two clips, but that can easily be done in a basic video editor). OpusClip takes this further, and has the AI actually rank the quality of the clips for engagement value. No surprises - I find the rankings a bit flawed. However, I'm seriously impressed by the effectiveness of their proposed video titles, as well as the AI's explanation as to why the clip is engaging. Take this example from the ESG show:

The attempt to rank, title, and explain why clips are engaging really impresses me. It's not perfect - I don't like the word "revolutionary" in this context, nor do I think the social media angle is relevant whatsoever - but it's still useful to see what the AI puts out. From here, I typically download these clips, tweak them a bit in Vimeo, and they are ready for upload to YouTube and social channels (I add them to my Enterprise hits/misses video highlights playlist). The time saved is enormous - perhaps 80 percent (it takes me about 15 minutes to finalize one of these clips, if I need to tweak it). It's the difference between clips and no clips. I don't have the bandwidth to make clips from scratch.

But, as I've said, OpusClip has social clips limitations: what if I want more control over what I can produce? I found that Wisecut offers a bit more comprehensive video collections than other clips services. And, for the most part, their interface is pretty easy - not as easy as OpusClip, but pretty good.

The only problem I found with Wisecut: I often wanted to bring the end of one clip into the next one. They are working on a way to do that, but for now, that has to be done manually. Wisecut also offers intro/outro clip options, though I struggled to make those work. But in theory, you can put intro/outro on Wisecut clips without needing an external editor. Wisecut can output in landscape mode, which is often better for B2B social sharing.

Finally, there is a different approach to try at Pictory.ai. There is more to Pictory.ai than just AI video editing; Pictory also offers blog to video, and script to video (I haven't tried those). For AI video editing, Pictory.ai has a learning curve. It took me a while to get setup with my intro/outro clip templates. But the nifty thing about Pictory is being able to edit the video by editing the transcript itself. Or, you can highlight text from the transcript and turn only that highlighted text into video. That highlight option can be useful if you have a particular theme in mind. During my show with Leary, he did some great show and tell of his equipment setup. I was able to capture that in Pictory in an 11 minute highlight video, pulling several disparate segments together:

The only problem with the highlight option? Once you make your highlight text selections, you can't re-arrange them much. I hope they enhance that. You can also ask Pictory's AI to pick the most important parts of the text for the video. I haven't tried that yet, only because my use case with Pictory.ai was more about the text editing this time around. I probably saved 80 percent of my time using Pictory's text-based editing to create that highlight clip. I like that Pictory.ai is approaching this problem differently.

My take - video editors aren't threatened at this time

AI editing tools are still immature, but they have improved in the last year - to the point I suspect most people (or teams) with a similar use case to mine can find a good option. I know that these clips services have helped to seriously extend the replay reach of my videos, upwards of 75 percent more replay views of all clips than I was getting before.

Will these tools take work away from video editors? Perhaps, but I'm not sure that the current crop of tools is sophisticated enough to make much of a dent. I can see video professionals turning the tables, and taking advantage of these tools in their own workflows, not unlike how I use tools like Auphonic to streamline podcast production and distribution. Also, experienced video editors can help set up the templates to get these services humming and "on brand" (Pictory.ai is one of the services that enables automatic logo insertion). They can also help tweak/edit the output.

To me, the shorter term impact is to free up content and marketing teams to share (and repurpose) more stockpiledcontent, in a citizen developer type of way. AI doesn't replace video editors yet, but it can make those editors more productive, and, to some extent, democratize video content production. Of course, none of this matters if the raw material of good video footage isn't there. Caliber of content assets is still the biggest problem for most. But for those with good footage on the shelf, these services can inject new life into the video pursuit.