Google’s Gemini has launched extensions that integrate with external platforms. Its YouTube extension is live and available for activation in the Gemini settings behind the “Extensions” tab.

The extension lets users search and summarize YouTube videos inside Gemini. Here’s how it can help video marketing.

Find Top Videos

Native YouTube search is highly personalized by viewing history and subscriptions. To see non-personalized results, searchers must log out or use a browser’s privacy (“Incognito”) mode. Even then, the results differ among users.

Moreover, exporting search results from YouTube is burdensome, requiring copying each video’s URL without the parameters.

Gemini’s extension produces the top five search results without personalization. Users can watch any video inside Gemini without impacting their viewing history on YouTube.

Simply prompt:

@YouTube search for

Or create a more advanced prompt:

@YouTube search for and return results as a table containing the video title (linked), number of views, and creator’s name (linked)

My results for the query “how to build a website” are below. I’ve reproduced Gemini’s table for this article.

Produce Video Ideas

Like any generative AI platform, Gemini is a good brainstorming tool. It can suggest helpful video topics to research and refine. The YouTube extension can find competing videos on each proposed topic.

Here’s my prompt:

My target phrase is “how to build a website.” Generate specific video ideas for that audience @YouTube.

Here are two of Gemini’s suggestions:

• Building a website with HTML in 30 minutes or less (targets beginners who want a quick and easy introduction).

• Creating a simple portfolio website using HTML and CSS (targets those who want to showcase their work online).

My next prompt (link):

Search @YouTube for each idea and produce examples.

Gemini’s results are well-ranking videos for each idea.

Improve Video Descriptions

Gemini’s YouTube extension can create detailed descriptions of videos on YouTube and add clickable timestamps.

In my testing, Gemini responds better to one task per prompt. Here’s my first prompt: a YouTube video from Brian Dean, “SEO for Beginners: Rank #1 In Google (2023).”

@YouTube Summarize this video by pulling key takeaways and action items https://www.practicalecommerce.com/geminis-youtube-extension-streamlines-marketing.

Here is Gemini’s response:

The main takeaways and action items:

Find low-competition long-tail keywords that are trending:

• Use Google Autocomplete to find these keywords.
• Look for keywords that are more specific and answer user queries in detail. For example, instead of targeting a broad keyword like “skin care routine”, target a long-tail keyword like “skin care routine for dry skin”.

Create high-quality content that satisfies search intent:

• When creating content, make sure it answers the questions users have behind their search queries.
• Focus on providing value and insights instead of just creating generic content.
• Look at the content structure of websites like Investopedia where they answer user queries directly and in a clear way.

Optimize your content for on-page SEO:

• Include your target keyword in the title, the first couple of hundred words of your content, and a few other times throughout the content.

Optimize your content for user experience:

• Make sure your content is easy to read and navigate.
• Use short paragraphs, subheadings, and bullet points to break up your text.
• Front-load the value by placing the most important information at the beginning of your content.

Build backlinks to your website:

• Create content that is link-worthy, such as content with unique data or insights.
• Target specific statistics in your niche that other websites might want to cite.

My next prompt:

Add a timestamp for each takeaway from the video. The format: mm:ss.

Here’s Gemini’s response with clickable timestamps:

• Target low-competition, long-tail keywords that are trending and not yet mainstream (01:53)
• Create content that satisfies search intent and gives searchers exactly what they’re looking for (02:51)
• Optimize your content for on-page SEO by adding your keyword to your page a couple of times (04:11)
• Optimize your content for user experience by making sure people can get value as soon as possible after landing on your site (04:22)
• Build backlinks by being the source for a specific statistic in your niche (05:04)

Creating quality videos takes time and effort. Ensuring they are discoverable in organic search is often a missing step. Here are three tools to help.

Ahrefs

A free tool in Ahrefs identifies keywords to insert in YouTube video titles, increasing the chances of ranking in YouTube’s and Google’s organic search results.

Ahrefs pulls keywords from YouTube’s autocomplete results. It also shows the organic search volume on YouTube for each keyword, providing the potential clicks if your video ranks for that term.

Click the “Questions” tab to find queries on YouTube associated with the keyword.

Gemini

Video descriptions can make or break rankings and engagement. Google’s Gemini (formerly Bard) streamlines video description composition. To test, I used this video for content marketing on Reddit.

I then entered the following prompt into Gemini:

Generate key takeaways from this video [URL].

Unlike Gemini, ChatGPT won’t work for this task because it cannot access YouTube videos. Here is Gemini’s response:

Here are the key takeaways from the video:

Reddit is a great platform to find content ideas and promote your content because it has a lot of stories, data, and studies shared there. However, it’s important to follow the rules of each subreddit and become part of the community before trying to promote anything.

You should not just drop your link and leave it. Instead, find a cool comment to go with the link, a question, or something to provoke a discussion. It’s also important to mix in sharing other people’s links and content that you found interesting.

Reddit has a voting system, so it’s important to make sure your content is high quality and helpful for people. This will help you get more upvotes and visibility.

By being active on Reddit, you can learn what approaches work best for your target audience and what kind of content gets shared the most.

Michael Cottam’s Embed Code

YouTube’s embed code slows down a page and doesn’t include Schema.org markup, which would help Google generate a video rich-snippet.

A free tool from Michael Cottam, a search engine optimization consultant, fixes both problems. Instead of using the embed code, the tool creates a still image containing a “Play” button. When “Play” is clicked, the tool swaps the image for an iframe that embeds the YouTube video and plays it.

To use, upload the video to YouTube and enter into the tool its:

• URL,
• Title and description,
• Duration,
• Upload date.

The embed code doesn’t load until the image is clicked, preventing slow page speeds. The tool also includes structured data markup for the Google rich snippet.

Google Ads has invested heavily in video. Recent examples include new YouTube advertising formats and Performance Max video campaigns. Video isn’t solely a brand awareness play. Compelling product videos can generate sales while promoting the brand. Google refers to this method as “leveling up for action” using ABCD principles:

• Attention,
• Branding,
• Connection,
• Direction.

Google Ads offers seven YouTube campaign subtypes for video. They range from reaching users while browsing to driving conversions. The subtypes I will review are:

• Video views,
• Drive conversions,
• Ad sequence.

With all subtypes, audience targeting segments are in-market, affinity, demographics, and advertiser-created. The right targeting is critical. Certain campaigns are for prospecting, while others are down-funnel. Audience targeting and associated videos must align with a campaign’s goals.

Video Views

The “video views” campaign subtype is new and targets users likely to consider the product or brand. The advantage is the ads are responsive. Advertisers provide a long headline and description along with the video and landing page URL. Google then shows the video across the different ad formats, including skippable in-stream, in-feed, and through YouTube Shorts.

Bidding is set at a target cost per view (CPV), and advertisers must include a total budget with an end date. For example, a two-week campaign may have a $2,000 budget with a CPV of $0.05.

A couple of beginning steps are necessary to avoid showing ads on irrelevant YouTube channels and other sites. First, confirm the option in the Networks section to target video partners on Google’s Display Network is unchecked.

Then, once the campaign is live, add topic, placement, and keyword exclusions. Similar to negative keywords in Search campaigns, these exclusions stop ads from showing on the wrong sites and YouTube channels. Unfortunately, without exclusions videos will show on irrelevant sites even with precise targeting.

Drive Conversions

Advertisers using the “drive conversions” subtype can optimize for specific actions and use goal-based bid strategies such as target cost-per-acquisition and return-on-ad-spend. Ecommerce merchants can attach their Google Merchant Center feed to show products with the ads.

Test a variety of videos and lengths. For example, test three videos, each at 10, 20, and 30 seconds, for nine altogether. As with video views, “drive conversions” ads are responsive.

Here are a couple of examples, below. The first is a skippable in-stream ad on computers, and the second is an in-feed video on phones.

Ad Sequence

Perhaps the most interesting campaign subtype, ad sequence tells a story through a video series. Consumers must view a video before they can see the next. Google offers many sequences. The most common is “custom” (advertisers create their own sequence) and “automatic” (Google optimizes).

As with the other campaign subtypes, ad sequence audiences can be as general or as specific as necessary. Bidding is target cost-per-thousand-impressions (CPM) or maximum cost per view. Budgets are set with end dates. Ads can show only on YouTube for 7 or 30 days — viewers see an entire sequence just once during that time.

A sample sequence can include three videos, such as:

• First video: details a problem.
• Second video: shows how the product solves the problem.
• Third video: explains how to buy the product.

Videos can be “skippable,” “non-skippable,” or “bumper.” A bumper ad is six seconds or less. Non-skippable videos are seven to 15 seconds. Skippable videos are any length but can be skipped after five seconds. Advertisers should keep video lengths in mind when crafting a story.

The transition type from one video to the next depends on the bidding strategy. The options are “impression” (ad was shown to the viewer), “view” (viewer engaged or watched at least 30 seconds), or “skip” (viewer skipped the ad). “Impression” is the most common — anyone served the video will see the next one.

Regardless, a well-aligned sequence is critical.

YouTube, the video streaming social platform, released new community guidelines relating to the disclosure of artificial intelligence (AI) used in content. 

The platform published a blog on Nov. 14 saying that the updates will have creators on its platform inform their viewers if the content that is being shown is “synthetic.”

“We’ll require creators to disclose when they’ve created altered or synthetic content that is realistic, including using AI tools.”

An example given in the update was an ai-generated video that “realistically depicts” something that never happened or content of a person saying or doing something that they did not.

This information will be displayed for viewers in two ways, according to YouTube, with the first being a new label added to the description panel and if the content is about “sensitive topics” a more prominent label to the video player.

Sensitive topics according to YouTube include political elections, “ongoing conflicts,” public health crisis and public officials. 

YouTube says it will work with creators to help its community better understand the new guidelines, however it said for anyone who does not abide by the rules their content is subject to removal, “suspension from the YouTube Partner Program, or other penalties.”

Related: Google sues scammers over creation of fake Bard AI chatbot

The platform also touched on the topic of AI-generated deep fakes, which have become both increasingly common and realistic. It said they’re integrating a new feature that will allow users to request the removal of a synthetic video that “simulates an identifiable individual, including their face or voice, using our privacy request process.”

Recently, multiple celebrities and public figures such as Tom Hanks, Mr. Beast, Gayle King, Jennifer Aniston and more have battled with deep fake videos of themselves endorsing products.

AI-generated content has also been a thorn in the side of the music industry over the last year, as many deep fakes of artists using illegal vocal or track samples have also plagued the internet.

In its updated community guidelines YouTube says it will also remove AI-generated music or content that mimics an artist’s unique singing or rapping voice as requested by its “music partners.”

Over the summer YouTube began working on its principles for working with the music industry on AI technology. Alongside the community guidelines, YouTube recently released new experimental AI chatbots that chat with viewers while watching a video.

Magazine: ‘AI has killed the industry’: EasyTranslate boss on adapting to change