3 Ways to use AI for LinkedIn Profiles and Post Analysis

There are such a lot of use instances for generative AI and advertising and marketing. In fact, there are some good ones for social media.

Take any social media asset and there’s most likely a option to plan it, create it or enhance it utilizing AI. So let’s decide one… the LinkedIn profile.

At this time we share 3 ways to make use of generative AI to enhance your LinkedIn profile and the content material you put up on LinkedIn. These are enjoyable examples as a result of they showcase three utterly completely different makes use of for AI: language technology, picture technology and information evaluation.

1. Your LinkedIn Headline and About part (language technology)
2. Your LinkedIn profile image (picture technology)
3. Your LinkedIn content material technique (information evaluation)

These AI strategies additionally vary from easy to superior; we’ll use free instruments and paid instruments; one is a fast test and one other requires time-intensive AI coaching. It’s a mini-tour of what AI can do at the moment.

So even for those who don’t use LinkedIn, this text offers you some new concepts on how you can use AI for social media advertising and marketing and advertising and marketing typically.

1. Let AI counsel adjustments to your LinkedIn Profile

We’ll begin with the quick and free: your LinkedIn headline and About part. When you give your present bio to ChatGPT, Claude or any AI software, and ask it to counsel enhancements, you may even see concepts for fast enhancements.

The immediate wants to incorporate a variety of element about you, so give it your whole profile together with job titles and detailed job descriptions. It’s additionally useful so as to add character limits and pointers. Right here’s it’s:

Check out the response. When you’re not impressed, there are two possible causes.

If the response is off on voice and tone, that you must coach it extra. Begin asking for revisions. It’s not cheap to anticipate wonderful outcomes from a single immediate.

• Rewrite utilizing a [MORE/LESS] formal tone.
• Rewrite utilizing shorter phrases.
• Rewrite making it friendlier, extra approachable and fewer self-promotional.
• Right here is one other LinkedIn profile. Use this as a information for fashion, voice and tone.

If the response is gentle on particulars, that you must give it extra information. Inform it extra about your self or give it your resume. The shorter your present LinkedIn profile, the much less coaching information the AI has to write down a greater LinkedIn profile. As with the opposite use instances on this article, AI usually requires extra information than you assume.

The standard of the output from AI is relative to the amount of the enter.

Plenty of LinkedIn customers don’t replace their profile usually, so the primary response could also be an enormous enchancment. However preserve speaking to the AI and the output ought to preserve getting higher.

As all the time with AI, we aren’t on the lookout for ultimate copy. It is a draft. You’ll wish to polish it up, add particular examples, add formatting and pull again from any overly promotional language. You may additionally wish to spice it up by hand. Plenty of what AI writes tastes like water.

Copy and paste the draft right into a doc and go to city. Or get assist from a LinkedIn or private branding professional, akin to my buddy, writer and private branding skilled, Catherine Morgan.

All set? Make it your headline and LinkedIn abstract. Then take the additional step and replace some other on-line bios you’ve got, together with your web site and different social profiles. It’s vital for personal branding and SEO.

Observe: Premium customers could have observed that LinkedIn has AI inbuilt. It’s a button within the Edit sections for varied elements of the web page. However I don’t advocate it. You may’t give it any directions. You may’t prepare it in any manner. It’s like utilizing a single longshot immediate.

2. Let AI make you a brand new LinkedIn profile image

Shifting on to essentially the most seen aspect in each profile, the LinkedIn profile image. For this, we’ll use an AI headshot generator. I attempted the paid software Secta Labs which prices $50 on the time of this writing.

The very first thing you’ll discover is that this AI-powered software desires a variety of coaching information. It requires a minimal of 25 good high quality images. So be ready to dig by means of social posts and archives. When you begin capturing screenshots out of group images, watch out to not crop them too tightly or they’ll be rejected.

This AI headshot generator rejects images for a number of causes:

• Too small
• Tight cropping
• Low decision
• Hats and sun shades

It’s all of a sudden one other tedious AI coaching mission. However that’s what correct use of AI is all about: gathering giant quantities of coaching information.

Reviewing these was manner more durable than I anticipated. It’s a traditional Paradox of Choice drawback. So many related choices. To get assist, the software enables you to share a subset with others who react to them, providing you with suggestions.

In the long run, I picked one which had a brilliant background shade, which we advocate in our guide to profile pictures. It made my brown eyes a bit blueish, however I’ll repair that later.

To validate my selection with information, I uploaded the file to Photofeeler, which is a freemium service. You may pay $9 and 40 folks will vote in your profile image inside just a few hours. Or you may earn credit by voting on different profiles (enterprise, social or relationship) after which use it without cost.

The AI generated headshot is on the left. It scored about the identical because the images I examined years in the past. Possibly I simply look extra competent as I grow old.

Now that you’ve got your new LinkedIn headshot, attempt utilizing AI to generate a LinkedIn background picture. This concept comes from veteran LinkedIn person and generative AI-pro, Nicole Leffer.

Nicole Leffer, CMO AI Advisor “An awesome cowl picture may also help your profile stand out on LinkedIn and it’s really easy to create one that basically represents you now with Dall-E-3 in ChatGPT! Simply inform it what you do and the vibe you wish to share with the world in your profile, and ask it to create you the proper horizontal picture to make use of as your cowl picture on LinkedIn. You may even present it your profile picture (I personally shade over my face earlier than importing an image of me), and ask it to create your cowl picture with colours that can look nice together with your garments and the prevailing background. Make just a few choices and check them out!”

3. Let AI analyze your LinkedIn put up efficiency

Subsequent we’ll transcend LinkedIn profile optimization and have LinkedIn analyze our social media advertising and marketing information. Right here’s the place AI may also help enhance you social media advertising and marketing technique.

This technique works for LinkedIn Premium customers who put up frequently and have Inventive Mode turned on. You’ll additionally want a ChatGPT Plus account to add spreadsheets for information evaluation.

Social media is a data-rich atmosphere the place each motion triggers information reactions. Create a put up, video, ballot, or article and also you’ll quickly have impressions, engagements, feedback and perhaps some reposts.

Give this information to AI. AI offers you insights into your social media technique.

The issue is that the info from LinkedIn isn’t structured. It could actually’t be exported right into a clear spreadsheet, even for those who’re a premium person. So our first step is the very tedious and handbook strategy of constructing a CSV file that has all of the LinkedIn information we will get, organized properly into columns and rows.

Listed here are the steps for turning unstructured LinkedIn information into an AI-friendly CSV fie:

• From your individual profile, click on on the “Analytics & instruments” bar on the high
• From the Analytics menu, click on on Impressions
• Set the date vary to 12 months. You need as a lot information as doable.
• Click on the blue “Export” button within the high proper nook to obtain the info as an Excel spreadsheet. The file could have a number of tabs. The “TOP POSTS” tab is an important.
• Clear it up and add extra information

You’ll must scroll by means of the highest posts from the Analytics > Posts report, click on on every and transfer the info into the spreadsheet one cell at a time. Pour a cup of espresso and get comfy. This may occasionally take some time. Or have interaction a digital assistant.

In the long run, you desire a column for “Submit Textual content” which incorporates the textual content from every LinkedIn put up. Sure, meaning visiting all of them and copying and pasting the textual content from every right into a cell on this new column.

If the posts are on varied matters, it’s helpful to categorize all 50 posts manually. The AI can do that, however not in addition to nicely as you may. So add “Subject Class” as one other column. The classes ought to align together with your content material technique and the classes in your weblog. And for those who put up movies and polls, create a “Kind” column and tag these.

These columns are essential. With out them, the AI can’t give enter into the matters and textual content. It’s all simply information with out context, apart from timing.

When you’ve accomplished this brute pressure, information entry mission, you’re information ought to look one thing like this:

Bought your LinkedIn put up information into a pleasant, structured format? Nice! Let the prompting start.

Begin with a easy role-and-goal immediate and add the info. Then begin asking for insights. There are such a lot of doable kinds of evaluation, I’m going to listing many prompts, displaying instance outputs for just some.

• You’re an skilled social media advertising and marketing information analyst, expert at discovering insights from information about social media posts. Your purpose is to seek out insights from advertising and marketing information that may drive choices about social media advertising and marketing technique. I’m providing you with information displaying the efficiency of the highest LinkedIn posts posted to a private LinkedIn account over one 12 months. Are you able to analyze?

It ought to verify that it might entry the file and summarize what it finds. Subsequent we’ll start evaluation. I’ve damaged up the prompts into three teams: subject evaluation, put up evaluation and strategic evaluation.

Subject evaluation prompts:

• Record the common impressions, common engagements, common feedback, common reposts for every subject class.
• Which subject classes get essentially the most feedback, engagements and reposts per impression?
• Which matters have the best affect on follower development? Follower information is within the FOLLOWERS tab?
• Create a scatter plot chart displaying the common efficiency of matters for impressions and engagement. Use a dot for every of the subject classes.

• Draw a heatmap matrix displaying the normalized common impressions, common engagements, common variety of feedback, common variety of reposts and common variety of new followers for every subject class.

The insights from the responses could also be apparent. You’ll see which matters set off essentially the most engagement and dialog together with your viewers. Give attention to these. That is just like the strategies for using AI to analyze GA4 data.

Look nearer and you might also see that some matters could also be higher for development and visibility (impressions, new reposts, followers) whereas others are higher for dialog and thought management (engagements, feedback). Now you may align matters with advertising and marketing objectives.

Submit evaluation prompts:

• How does phrase rely, use of numbers and emojis correlate with put up efficiency?
• How does video carry out in comparison with non-video posts? Create a visualization displaying the distinction in common efficiency for these two kinds of posts.
• Utilizing TF-IDF evaluation, which particular phrases and which three-word trigrams are over-represented within the high performing posts?

Once more, the insights could also be apparent. Posts with particular attributes could align carefully with engagement. The info could also be providing you with ideas for content material creation, telling you to make use of particular language or social put up size. Sadly, the usage of particular kinds of photos isn’t doable with this strategy.

Strategic evaluation prompts:

• Utilizing semantic distance evaluation, which subtopics should not represented right here however prone to carry out nicely on this LinkedIn account?
• How would you summarize the LinkedIn content material technique represented on this dataset?

That final one is particularly fascinating. You could have a documented LinkedIn content material technique. Does it match with what AI thinks of your LinkedIn content material and its efficiency?

This type of evaluation takes time. So does this laborious, handbook course of make sense for you?

• When you’re from a model validating your content material technique, this occasional batch-process AI-reporting might be fantastic and can prevent the price of LinkedIn reporting software program.
• When you’re at a social media company, doing common reporting for purchasers, a real-time course of displaying ongoing efficiency utilizing paid providers (akin to SHIELD Analytics, Inlytics and AuthoredUp) are far more environment friendly.

However after all, the distinction between AI and LinkedIn reporting instruments is that AI can take the following step, past evaluation. Listed here are the prompts for those who’d prefer to take the following step:

Content material technology prompts:

• Generate a listing of 10 instance headlines which may be used to create LinkedIn posts that drive excessive ranges of engagement based mostly on this information.
• Utilizing the matters and tone of the posts on this dataset, create an instance LinkedIn social media put up that’s prone to drive excessive engagement, feedback and reposts.

These prompts are just like the idea generation prompts used by 43% of bloggers. This type of concept technology is the commonest use case for advertising and marketing synthetic intelligence.

• Create a single immediate that may work as a mode information that may be given to ChatGPT so the AI can successfully write future posts on this similar voice, tone and elegance.

You may then take that fashion information, edit it, then share it with any social media entrepreneurs in your group, or use it to tell changes to your model pointers.

Scott Aaron, Author, Podcast Host, and LinkedIn Expert “It’s past vital to know what content material is touchdown and what content material you might be posting is lacking the mark. On the finish of every week, I have a look at my put up analytics to see what my viewers loved so I can use that information to create extra content material that speaks to sure ache factors that my viewers could have of their companies. Many individuals battle with what to put up on LinkedIn, and I all the time inform them to “ask ChatGPT”. I’m not suggesting that you simply copy and paste what you plug into it, what I’m suggesting is that you simply use the output because the “bread” of your content material sandwich. The meat of your sandwich is your factors of view that you simply convey to the put up itself. Now let’s return to your top-performing content material from the week. For instance, let’s fake you wrote a easy put up that had 3 recommendations on “how you can enhance XYZ in your small business.” You may plug that into ChatGPT and ask it to “rewrite this right into a long-form e-newsletter article for LinkedIn that’s partaking, entertaining, and informative.” Take the output and edit it so it’s in your fashion of wording, phrases, and punctuation. Doing this on a weekly foundation will assist you to by no means run out of repurposed content material. Bear in mind, give them extra of what they need and fewer of what they don’t need.”

Coaching information is the important thing to high quality

All three of those strategies had one factor in widespread: all of them depend on numerous high-quality coaching information. That’s the key to AI outcomes. It’s additionally what led to the breakthrough of the large gamers who created these AI instruments.

Positive, for quite simple AI use instances, no coaching is required. However here’s what all advertising and marketing AI practitioners ultimately uncover: the extra information (and the extra structured information) we give the AI, the higher the insights will likely be.

• Give it extra information, it offers you higher insights.
• Give it extra on-brand content material samples, it offers you higher draft copy.

It’s a coaching problem. So we have to make investments time feeding it information. This takes time for those who’re beginning with unstructured information. So right here is one final commentary based mostly on private expertise utilizing AI this final 12 months:

• Gathering coaching information requires little ability, however numerous time.
• The prompting course of requires extra ability, however little or no time!

FYI, There have been a number of fascinating use instances that didn’t make it into this text, together with ideas for job seekers, LinkedIn Gross sales Navigator and networking. Keep tuned for a doable Half 2 of AI for LinkedIn!