Chapter 7
What do cyber firmswriteabout, and what do audiences respond to?
One in five LinkedIn posts by cybersecurity firms is a general tip, and tips are the topic least likely of all to get more reactions than the same firm's typical post. Posts about people and events get the most reactions, but reactions measure reach, not credibility. Content about the firm's own work, such as case studies, shows expertise best.
- 19%of posts are tips and education
- 32%of tips get more reactions than the firm's typical post
- 85%of posts about team members beat the typical post
- 40%of firms that post regularly get almost no reactions
- 30% of cybersecurity firms post on LinkedIn at least once a month. Active firms publish a median of 4 posts per month.
- The most common topics are tips and education (19% of posts), service promotion (13%) and events (13%). Tips and service promotion beat the same firm's typical post less than half the time: tips in 32% of cases, promotion in 40%.
- The posts most likely to beat the typical post are about team members (85% of them), proof of the firm's own work such as case studies, awards and certifications (74%), partnerships (69%) and events (61%). Together they make up 23% of posts. Reactions measure reach, however, and posts about people do not show competence. Among expert content, proof of the firm's own work performs best, expert opinion sits in the middle (42%), and general tips come last.
- Images and video beat the typical post more than half the time (58% and 54%), while article links and text-only posts do so much less often (38% and 30%).
- 53% of LinkedIn ads are brand ads or promoted posts that do not offer the viewer any next step. 26% include a concrete offer, such as a guide, a free assessment or a consultation.
- Firm websites use words that buyers do not type into Google ("comprehensive", "trusted") and rarely use the phrases buyers search with.
How many firms post?
30% of cybersecurity firms post on their LinkedIn company page at least once a month. The share grows with firm size: it is 20% among firms with up to 10 employees, 45% among firms with 11-50 employees and 60% among firms with 51-200 employees. Firms that post regularly do so a median of 4 times a month.
Posting regularly does not mean that posts reach anyone. Among firms that post at least twice a month, 40% get an average of fewer than three reactions per post. Their median follower count is 216, so some of them post to a small audience, but even that audience barely reacts.
posts in 90 days from one firm, with an average of one reaction per post.
That is more than one and a half posts a day. This frequency takes steady work, but without a change in topics and format it does not translate into reach.
What do firms write about?
I analyzed 12,976 posts from the company pages of 685 firms across all 22 states and assigned one topic to each post. Firms most often publish tips and educational content, for example on how to spot phishing or how to secure passwords. Next come events, promotion of the firm's own services, and news about attacks and breaches.
Post topics of cybersecurity firms
% of 12,976 posts from 685 firms
Source: LinkedIn company page posts, all states in the study; a topic assigned to each post. October-November 2026.
These three most common topics have one thing in common: they can be written without input from people at the firm and without information about specific clients. A tip about phishing or a news item about an attack looks much the same at every firm. Topics that require knowledge from inside the firm, meaning posts about the team, projects, events and results of the work, make up 23% of posts combined.
Which topics get reactions?
Firms differ in follower count, so raw reaction counts are hard to compare. I therefore compared each post with the number of reactions the same firm's posts usually get (the median of its posts, which I call "the firm's typical post") and calculated what share of posts on a given topic got more reactions than the typical post. If topic made no difference, the result would be about half. The comparison covers 331 firms whose typical post gets at least 3 reactions; with smaller numbers, the differences are random.
Post performance vs. the same firm's typical post, by topic
% of posts that got more reactions than the same firm's typical post
line = half of posts
Reactions: likes + 2 × comments (+ shares where available). Source: 6,193 posts from 331 firms, October-November 2026.
Posts about team members, such as new hires, work anniversaries, promotions, certifications earned by employees or the team at an event, beat the same firm's typical post in 85% of cases. A typical post of this kind gets 3.6 times as many reactions as usual. Proof of the firm's own work, partnerships and events also often perform well. Tips and news about attacks, among the most common topics, are the least likely to beat the typical post.
What separates the best posts from the weakest?
A second measure confirms this picture. For each firm, I took its five best and five weakest posts and compared how often each topic appears in them. Posts about people, events and case studies more often land among the best. Tips and news about attacks more often land among the weakest.
Topic share among a firm's 5 best posts minus its 5 weakest posts
percentage points
Source: 331 firms, 5 best and 5 weakest posts each. October-November 2026.
Reactions are not everything
Reactions show how many people noticed a post, not whether they came to see the firm as an expert or whether they will ever buy from it. Reactions to posts about a new team member or a work anniversary probably come largely from coworkers and friends. That extends reach, but it does not show a buyer what the firm can do.
For an expert brand, another finding matters more: content about the firm's own work performs well, and generic content performs poorly. Posts with proof of the firm's own work, meaning case studies, awards and certifications, beat the firm's typical post in 74% of cases. Expert opinion (42%) and regulatory content (39%) sit in the middle, and general tips (32%) come last. So the problem is not expert knowledge, but knowledge that any firm could publish.
Stetson Cybergroup of New York published a recap of its expert's appearance at a regional technology summit. The post got 80 times more reactions than the firm's typical post.
Anjolen of New York regularly posts about its employees: work anniversaries, volunteering, support for student cyber competitions. A post about one employee's tenth work anniversary got 22 times more reactions than the firm's typical post.
Trail of Bits of New York describes the results of its own research, for example a test in which an AI model escaped an isolated environment three times. Posts like these get more than a thousand interactions. Original research is content that cannot be copied from another firm.
Does format matter?
More than half of posts (52%) are an image or photo with text. 20% of posts include a link to an article, usually on the firm's blog, 17% are text only, and 8% are video.
Post performance vs. the firm's typical post, by format
% of posts that got more reactions than the same firm's typical post
line = half of posts
Images or photos (58%) and video (54%) beat the typical post most often. Posts with an article link (38%) and text-only posts (30%) perform noticeably worse. A link that leads off LinkedIn and the lack of an image are two features that go hand in hand with lower performance in the data.
How do posts open?
Almost one in five posts opens with a bold claim ("Cyberattacks are no longer a question of if…", 18% of posts), and one in ten with a question (10%). Posts that open with an announcement, such as introducing a new team member, an award or a partnership, beat the typical post most often (64%). Questions (41%) and bold claims (40%) come in below half.
Post performance by opening sentence
% of posts that got more reactions than the same firm's typical post
83% of posts have no call to action at the start. Among those that do, the most common is a link ("read more", "learn more"), and posts with a link beat the typical post in only 36% of cases.
What do ads offer?
An ad is content too: in a few sentences, the firm says what it offers and what the viewer should do next. 13% of the firms in the study run ads on LinkedIn or Google (chapter 6 shows how many and where). I analyzed 430 ads for what they offer the viewer: 319 from the LinkedIn Ad Library and 111 ads in Google search results. On LinkedIn, 53% of ads are brand ads or promoted posts without a concrete offer, and 18% are invitations to a webinar or event. 26% of ads include a concrete offer for buyers, such as a guide, a free assessment or a consultation.
What cybersecurity firms' LinkedIn ads offer
% of 319 ads
Excluding job ads. Source: LinkedIn Ad Library, October-November 2026.
- Brand, promoted post
- "Small business owners already handle finance, sales and HR. With us, they don't have to add cybersecurity to that list." Or: "How do you know you waited too long on cybersecurity? When your customers find the gap first." The ad reminds viewers of the firm or of a problem, but does not offer any next step, such as downloading a guide or booking a call.
- Service promotion
- "The biggest threat to your practice may be vendors you don't even know about." An ad for a specific service, but with no next step for the buyer.
- Webinar or event
- "Heading to RSAC? Join us before the conference." An invitation with a registration link.
- Guide, e-book, checklist
- A downloadable e-book on the difference between offensive and defensive security.
- Free assessment
- "A cybersecurity assessment built specifically for financial firms" with a button to book a meeting.
- Report, study
- "Every other cyberattack targets a small business." An ad promoting a report or study.
- Consultation
- "Implementing Zero Trust takes more than buying another license" with a button to book a call.
- Regulatory deadline
- "NYDFS deadlines are no longer just approaching." An ad that refers to the date by which a firm must meet a requirement.
Google ads look different: they are short text ads above the search results. 51% of them describe only the service (e.g. "Startup SOC 2 Services"), and 44% offer a concrete step, most often a free assessment or a call.
A concrete offer is something the viewer can use right away: a free assessment, a guide or checklist, a report, a consultation or a demo. A brand ad makes sense when a firm wants buyers to remember its name. With a small budget, however, an ad with a concrete offer makes results easier to measure, because the viewer takes a specific action.
What language do websites use?
I compared the words used on the websites of 1,488 firms with the phrases buyers use to search Google for cybersecurity services. Websites most often describe the firm with words meant to build trust: "comprehensive", "trusted", "tailored". None of these words appear in the phrases buyers type into the search engine.
Seller words and buyer phrases
% of websites that use them
Website words (0 searches)
- comprehensive52.2%
- trusted52.6%
- tailored51.7%
- proactive43.3%
- safeguard41.4%
- robust24.2%
- peace of mind17.9%
Buyer phrases
- penetration testing services8.8%
- cmmc level 29.1%
- soc 2 readiness3.9%
- soc 2 audit3.9%
- hipaa risk assessment2.0%
- free assessment3.4%
- near me0.3%
Source: website copy of 1,488 firms; buyer phrases with Google search demand. October-November 2026.
Buyer phrases are specific: they name the service, the standard and the stage ("soc 2 readiness", "hipaa risk assessment", "cmmc level 2"). They rarely appear on firm websites. The phrase "soc 2 readiness" appears on 3.9% of websites, even though far more firms offer SOC 2 readiness assessments. A website that does not use the buyer's language has a lower chance in search, and, according to my analysis of ChatGPT answers, also in AI tools, which describe firms mainly based on their own websites.
Do websites show proof?
Buyers of cybersecurity services worry most about paying for an automated scan presented as a full service. Proof of work, such as case studies, client logos and testimonials, helps reduce that worry. On the websites of firms in the region, however, it is rare.
The lack of logos can be partly explained by confidentiality: clients of cybersecurity firms often do not want their name linked to security testing. A case study can still be written without the client's name, by giving the industry, company size, problem and result. Posts with proof of the firm's own work, including case studies, get more reactions on LinkedIn than the firm's typical post, so the same content can work both on the website and on social media. Chapter 3 shows how many websites answer the specific questions buyers of a given service ask, for example about a sample report.
What does the website offer the buyer?
The main call to action is most often an invitation to talk ("Schedule a consultation", "Book a call", a free assessment), found on 36% of websites, or a general "Contact us" (24%). Both require the buyer to be ready for a conversation. A smaller step, a lead magnet offered in exchange for contact details (a checklist, a readiness assessment, a report, a webinar), appears on 48% of websites.
The results in this chapter show a typical trap of B2B content marketing. Firms publish what is easiest to produce: general tips and commentary on the news. This content looks professional, but it does not set a firm apart from its competitors, because most firms in the industry publish similar posts.
It helps to separate two jobs that content does. Posts about people and events extend reach and show that there are real people behind the firm. An expert brand is built by something else: knowledge that no one else has, such as a project outcome, a finding from the firm's own data or an opinion backed by experience. In an industry where clients give a firm access to their systems, this kind of content answers the question buyers ask themselves most often: who exactly will be working for me, and have they done something like this before?
The second takeaway is about language. Today, a firm's website and posts are read not only by people, but also by search engines and AI models. If the website says "comprehensive protection" and the buyer searches for "soc 2 readiness assessment", the firm is less likely to appear in the results and in ChatGPT's answer, no matter how well it delivers the service.
Michał Holub, elevate41
What to do
Below, I set out three first steps for each firm size group. All of them are based on the findings in this chapter and require no additional ad budget.
Fewer general tips, more of the firm's own work
Publish a weekly post about what the firm actually does
Instead of general tips, it is worth describing the firm's own work: what problem the team solved this week, what the last project taught them, who on the team is responsible for what. Lessons from projects show competence, and posts about team members extend reach.
Add a photo instead of a link
Posts with an article link and text-only posts beat the firm's typical post less often than others. It is worth summarizing the article in the post itself, adding a photo or graphic, and putting the link in a comment.
Use buyer phrases on the website
It is worth checking whether the website names services the way buyers search for them, for example "SOC 2 readiness assessment" instead of "comprehensive compliance solutions".
A calendar for reach and expertise
Split the content calendar between reach and expertise
Posts about people, events and partnerships build reach, while case studies and lessons from projects build credibility. General tips and news that any firm could publish are worth keeping to a minimum.
Publish one case study a month, even without the client's name
Industry, company size, problem, process and result. One write-up can work as a website page, a post, material for salespeople and an answer to the question "have you done something like this before?"
Record short videos with an expert
In the data, video, along with images, is one of the formats that most often beat the typical post. A minute with a penetration tester or an auditor commenting on one specific problem links the visual to a real person and their expertise.
Content no one else has
Publish original research and data
Firms of this size have data from hundreds of projects. A summary of the most common vulnerabilities or audit findings, without client data, is content that cannot be copied and that builds an expert position.
Review topics every quarter
Each quarter, it is worth comparing post performance with the firm's typical post, as in this chapter, and shifting the calendar toward the topics that work.
Use the same language on the website and in posts
Service names on the website, in posts and in ads should match the phrases buyers search with. This makes the firm easier to find on Google and in AI tool answers.
Questions for your own firm
Six questions that help assess a firm's content against the findings in this chapter.
- Do our ads offer something concrete, or do they only promote posts?
- What share of our posts from the last three months are tips and news, and what share are about people, projects and events?
- How many reactions does our typical post get, and which posts did noticeably better?
- Do posts with a link perform worse than posts with a photo?
- Does our website name services the way buyers search for them, or with words like "comprehensive" and "trusted"?
- Is there at least one case study on our website, even without the client's name?
Methodology
How I analyzed content
- Post sample
- 12,976 company page posts from 685 firms across all states. In the three pilot states (New York, New Jersey, Massachusetts): the 10 best, the 10 weakest and up to 20 random posts for each firm. In the other states: the 5 best and the 15 most recent posts of each firm that had at least 5 posts. I base topics, formats and opening sentences on this sample.
- Other figures
- posting cadence, proof on websites and website language: all 1,844 firms in the study from 22 states in the Eastern Time Zone
- Post performance
- likes + 2 × comments (+ shares in the pilot states), compared with the same firm's median; outside the pilot states, the median is calculated from the 15 most recent posts. Comparisons include firms whose typical post gets at least 3 reactions (331 firms, 6,193 posts)
- Topic, format, opening
- assigned to each post based on its content
- Ads
- 319 firm ads from the LinkedIn Ad Library and 111 ads above Google results, excluding job ads; I assigned each one to a type of offer based on its content
- Website language
- website copy of 1,488 firms compared with buyer phrases and their Google search demand
The sample deliberately includes each firm's best and weakest posts, so topic shares approximate the real content mix. Post performance shows an association between topic and reactions, not proof that the topic itself causes them.