If you want to know what your competitors are talking about right now, you can read their LinkedIn company pages. That sounds simple, and it is, as long as you settle three things before you start: whose pages count, over which period you read, and what you may conclude from a post. This guide walks through the six steps that turn reading into an observation whose numbers you can pass on.
Every step is shown on a real market: the German market for practice management systems (PVS, from the German Praxisverwaltungssystem), the software that medical, dental and psychotherapy practices use to manage patients, appointments and billing. 33 vendors, 21 LinkedIn company pages, 31 days from 25 August to 24 September 2026. What came out about that market, with every piece of evidence, has a page of its own: The PVS market on LinkedIn, September 2026. This page is about the way there, and that way is the same in every market.
Why LinkedIn, and what the channel costs
A LinkedIn company page has three properties that matter for monitoring: a sender, a timestamp and a retrievable list of its posts. That is why you can say with confidence what it published in a given period. The sentence “vendor X published twelve posts in September, five of them about products” can be checked. With a keyword search across the open web it could not, because nobody knows what the search did not find. Sources, not keywords explains the difference.
The channel has a price. In the PVS market, twelve of the 33 vendors have no LinkedIn company page; they speak through a blog, a newsletter or YouTube. Anyone who reads only LinkedIn sees nothing of a third of the market. And even vendors with a page do not say everything there: tomedo published 17 posts on LinkedIn during the period, alongside around three dozen training videos on YouTube and a newsletter. LinkedIn is one channel of a source, not the source. That does not stop it from being a good first channel. It only changes what you may claim from the results.
The six steps
1. Write the watch list
All vendors in the market, not the five best-known ones. Starting points are approval lists, association memberships and the exhibitor lists of the trade fairs. Every entry gets a segment, so that you can later read by segment. Regulators, associations and trade media join as a second circle with a role of their own: they do not answer “what do the vendors say” but “what is said about the market”.
In the example, 33 vendors from the approval lists of the self-governing bodies were on the list, in three segments (medical, dental, psychotherapy). The second circle of gematik, the KBV, the KZBV, associations and trade media was deliberately left out of the analysis.
2. Record the channels per source
For every source, the list says which channels it uses: LinkedIn company page, newsroom, blog, newsletter, YouTube. If a channel is missing, that goes into the list too. The gap is part of the result, not a fault of the list.
And check every address once. In the example, tomedo’s LinkedIn presence, a showcase page, was entered incorrectly on the first attempt. The result was zero posts, and zero posts look exactly like a vendor that stays silent. Only checking the address brought 17 posts to light. An empty channel is first a question about your own list, and only then a statement about the source.
3. Set a period and collect everything
Not the interesting posts, all of them. Only then can you count and compare later. By hand that means opening every company page once a week and copying every new post into a table, with date, source, link and one sentence of content. The appointment is fixed; its length grows with the number of pages. From around twenty pages, a tool that fetches the pages itself pays off.
In the example it was 31 days, 21 pages and 167 captured entries. Two pages published nothing during the period; they appear in the list with a zero rather than not at all. Silence is an observation.
4. Remove duplicates and assign topics
Before you quote numbers, you need to know how many distinct posts you have. In the example it was 165, not 167: two job ads from a dental vendor had been captured twice. Then every post gets one to three topics from a fixed set. Eight to ten topics are enough: products, personnel, events, customers, partnerships, regulation, company, insights. A post may carry several.
The topic Personnel deserves a look of its own. In the example, 63 of the 165 posts carried it: job ads, company runs, anniversaries, team photos. Anyone using the channel for market monitoring subtracts this share before comparing numbers, or a third of the observation consists of running shoes. At the same time the share is information in itself: at one dental vendor nine of 14 posts carried the topic, at tomedo three of 17. That says what a vendor uses the channel for.
5. Build the matrix and read it
Sources in the columns, topics in the rows, the number of posts in every cell. Then ask three questions: What do several play? What does only one play? What does nobody play?
A row with many occupied cells is a topic the market shares. In the example, four vendors played AI on the phone and in documentation in the same month, and four vendors advertised switching software within two weeks, three of them with a concrete offer. That is the kind of signal you only see when you lay several sources from the same period side by side. A row with a single occupied cell is a unique position, at least in this month and on this channel: in the example, a change of ownership at Doc Cirrus was the only announcement of its kind. A column shows what a source uses the channel for: at CompuGroup Medical two thirds of the posts were product posts, at tomedo six of 17 carried the topic Partnerships.
6. Back every observation with the original and write down its limit
Every statement gets two sentences. The first says what follows from the posts, with a link to the original. The second says what does not follow. The second is the more important one, because it keeps the observation from being passed on as something it is not.
In the example this means: four vendors talk about AI on the phone, so AI is a topic of the market this month. What does not follow is who is further ahead with these functions; a post shows what a vendor wants to say, not what its software can do. Four vendors advertise switching, so switching is a sales argument. What does not follow is that practices switch more often; that takes market data. A vendor with many job ads is advertising many jobs. What does not follow is that it is growing; a job ad is not a hire.
What the result shows and what it does not
After the six steps you have, for every vendor with a page, a verifiable statement about what it talked about during the period, and for the market a matrix that separates shared topics from unique positions. You do not have: the reach and effect of the posts, the communication of vendors without a page, posts on the personal profiles of managing directors, or any statement about product quality. Anyone who wants to close one of these four gaps needs more channels or more sources, not more LinkedIn.
And then: repeat
The first analysis is the most laborious one, because the list has to be built. From the second one on, the comparison becomes possible that makes the method worth anything: which topic is growing, which is fading, who is newly loud, who has gone quiet. The period should therefore be chosen from the start so that it can be repeated. A calendar month is a good rhythm. What the analysis of the example month looks like, with six observations, the topic matrix and the complete list of the 33 vendors, is under The PVS market on LinkedIn, September 2026.
Frequently asked questions
How do you track competitors on LinkedIn?
You make a list of the company pages that count, open them on a fixed schedule and record every new post with date, link and topic. The important part is to capture all posts, not just the interesting ones, otherwise nothing can be compared later. Public personal profiles of managing directors are added as sources of their own.
Is LinkedIn enough to monitor a market?
No. LinkedIn has the advantage that for every company page you can say with confidence what it published. But not every vendor is active there, and some say the most important things in a newsletter or on their own website. In the German PVS market, twelve of 33 vendors had no LinkedIn company page. LinkedIn is one channel of a source, not the source.
How many topics does the matrix need?
Eight to ten. Fewer, and everything lands under “products”; more, and the cells become so thin that nothing can be compared. The set is fixed before collecting starts and does not change during the period, otherwise the weeks are not comparable.
Why count posts and not mentions?
Because the question is what a specific market participant says itself, not where its name appears. The number of mentions measures the attention of third parties; the number and topics of a source’s own posts show what a vendor uses its channel for. The comparison Sources, Not Keywords explains the difference.
References
The figures in the example come from the analysis The PVS market on LinkedIn, September 2026, which links every observation to the original post. The posts mentioned here (all in German):
- tomedo GmbH: Interface between tomedo and medflex.de, 3 September 2026. https://www.linkedin.com/feed/update/urn%3Ali%3AugcPost%3A7501209917161185281 (opens in a new tab)
- CompuGroup Medical Deutschland AG: Why practices do not switch their software, 14 September 2026. https://www.linkedin.com/feed/update/urn%3Ali%3AugcPost%3A7505234358656585729 (opens in a new tab)
- Doc Cirrus GmbH: Doc Cirrus becomes part of the ETERNO Group, 21 September 2026. https://www.linkedin.com/feed/update/urn%3Ali%3AugcPost%3A7507712630514601984 (opens in a new tab)
- tomedo GmbH: Partner meeting 2026 in Jena, 16 September 2026. https://www.linkedin.com/feed/update/urn%3Ali%3AugcPost%3A7505902599955693569 (opens in a new tab)