SOCIAL MEDIA & DIGITAL MARKETING
Between October 2024 and January 2025, we ran a focused LinkedIn content experiment at Softomate Solutions. Starting from a low baseline of approximately 800 monthly impressions across the founding team's profiles, we implemented a structured 90-day content strategy and tracked the results weekly. By day 90, monthly impressions had increased to over 32,000 across the same profiles, a 300% increase from the baseline month. This is the exact strategy we used, the data from each phase, and what we learned that we would do differently if we started again.
Before the experiment, our LinkedIn posting was inconsistent (two to three posts per month per profile, with gaps of two to three weeks between posts), unfocused (topics ranged from AI and software development to general business advice to personal achievements), and low engagement (average post receiving 15 to 25 impressions, occasional posts reaching 200 to 400 impressions). We were not leveraging LinkedIn as a business development tool. We were using it as an occasional announcement channel.
The two root causes we identified: no clear topic focus meant the algorithm could not categorise our content for interest graph distribution, and inconsistent posting meant we had not built the algorithmic authority that consistent posting creates. Both were addressable without additional budget.
The first two weeks were spent on strategy design before posting began. We defined: the primary topic focus (AI automation and software development for UK professional services businesses), the three post types we would use, the posting schedule (four posts per week per profile), the metrics we would track (impressions, comments, profile views, and inbound connection requests), and the baseline measurements from the previous 30 days.
The three post types we chose: expert insight posts (specific observations from client project work, structured as a claim plus evidence plus implication), case study vignettes (specific client outcomes with enough detail to be credible, short enough to fit a LinkedIn post), and question posts (a specific professional question that invited comments from our target audience of business owners and technology leaders).
We deliberately excluded: general industry news posts (too much competition, no differentiation), motivational or inspirational content (inconsistent with professional positioning), and personal milestone posts (graduation announcements, work anniversaries) that generate reactions from personal connections but not professional engagement from potential clients.
We posted four times per week per profile: two expert insight posts, one case study post, and one question post. All posts were text-only or included native LinkedIn documents (carousels) but no external links in the main text. All links were placed in the first comment on the relevant post.
Weeks three to four results (days 15 to 28): weekly impressions increased from approximately 200 to 800 to 1,200 to 1,800 per profile. The topic focus had an immediate effect on the algorithmic categorisation: posts in the second and third weeks of consistent posting on the same topic reached non-connected professionals in the target industry at a significantly higher rate than the baseline random posts.
The breakthrough in Phase 2: on day 31, a case study post about customer support chatbot performance reached 12,400 impressions over 72 hours. The post shared specific data (71% automation rate, 41% cost reduction, 17 percentage point CSAT improvement) from a real client implementation. The specific numbers generated 34 comments from professionals asking follow-up questions. The comment velocity in the first four hours triggered the LinkedIn algorithm to distribute the post significantly more broadly than our baseline posts. This single post tripled the monthly impression total for the month it was posted in.
With the data from Phase 2, we made three adjustments for Phase 3. First: we increased the proportion of case study posts from one per week to two per week per profile, because the data showed they generated the highest comment-to-impression ratio and the strongest profile visit conversion. Second: we added a systematic comment engagement practice (30 minutes per day visiting the posts of professionals in our target audience and leaving substantive, professional comments) to increase our visibility within the professional communities we were trying to reach. Third: we standardised the expert insight post format around a specific structure: first line as counterintuitive claim, three to four short paragraphs of supporting evidence, one specific actionable implication, one professional question at the end.
Phase 3 results (days 46 to 90): weekly impressions continued to grow consistently, reaching 7,000 to 9,000 per week per profile by week 12. Monthly impressions reached 32,400 combined across profiles in the 30-day period ending on day 90, against a starting baseline of approximately 8,000 monthly impressions combined. The 300% increase was composed of 60% from improved algorithmic distribution due to topic focus and posting consistency, 25% from the breakthrough posts that generated high comment volume, and 15% from the comment engagement practice that increased visibility within target professional communities.
Start the comment engagement practice in week one rather than week four. The 30-minute daily comment practice generates significant organic visibility within professional communities independently of the algorithmic distribution effects. Starting it earlier would have accelerated the Phase 2 growth timeline by two to three weeks.
Produce at least one LinkedIn Article per month in addition to regular posts. Articles are indexed by Google, contributing to SEO visibility for the same professional topics that the posts cover on LinkedIn. We did not produce articles during the experiment and missed the Google search visibility opportunity that consistent article publishing would have created.
Track inbound enquiries separately from impression growth from week one. We tracked impressions rigorously but tracked inbound enquiries only informally. Post-experiment analysis suggested that four to six inbound enquiries were directly attributable to LinkedIn visibility during the 90-day period, but we could not attribute them to specific posts without the tracking infrastructure in place from the start.
Three posts per week consistently produces measurable impression growth for most professional LinkedIn accounts when topic focus and post quality are maintained. Below three posts per week, the algorithm does not develop sufficient topic authority to increase interest graph distribution meaningfully. Four posts per week (the frequency we used in this experiment) accelerates the growth timeline by approximately 20% to 30% over three posts per week, with diminishing returns above five posts per week for most professional account types.
Maintain the posting schedule and topic focus indefinitely. LinkedIn impression growth does not plateau and reverse when consistent posting continues. It slows from the rapid growth rate of the first 90 days (when algorithmic authority is being established from a low base) to a steady growth rate driven by audience accumulation and occasional high-distribution posts. The 90-day experiment produced a new baseline from which continued consistent posting continues to grow impressions, though at a slower percentage rate as the absolute numbers increase.
To learn how AI tools can help you maintain a consistent LinkedIn posting schedule without spending hours writing each post, read our guide on using AI to create a month of social media content in one day.
More from the blog
Let us help
Talk to our London-based team about how we can build the AI software, automation, or bespoke development tailored to your needs.
Softomate Solutions Limited
Deen Dayal Yadav
Online
