Influencer intelligence

Weekly intelligence on the people who shape your market

Orbelos follows your own lists of people on X and LinkedIn and turns each week of their posts into a briefing: the subjects, who drives them, the brands they name and what changed. Every number and every sentence opens the posts behind it.

Weekly briefing

Example

Cloud & AI infrastructure · 62 people · week to 4 October

  • Sovereign cloud overtook FinOps as the most discussed subject+62%38 posts
  • Halden’s key custody model drew favourable comment from 6 membersIndicative11 posts
  • Two people outside the list were quoted by 9 members14 posts

Invented people and brands, shown to illustrate the layout.

Every week, for each list

What you get

Briefing

Headlines, a timeline of the week and what changed since the last edition, written from the posts of your list.

Topics

What your list is talking about, and whether each subject is emerging, rising, steady or fading.

People

Who drives a conversation, who amplifies it and who links one group to another. New voices your list has started to quote.

Brands

Mentions and share of voice for the brands on your watchlist, with the stance taken toward each.

Evidence

Open the exact posts behind any number, chart point or sentence. Ask a question and get an answer with numbered citations.

Reports

Collect any block into a report while you browse. Edit the text, add your commentary, then export to PDF or PowerPoint or share a read-only link.

From a list to a report

How it works

Four steps

  1. Add the people you want to follow, or start from a published list.
  2. Orbelos collects their public posts on X and LinkedIn every week.
  3. Each list gets a weekly briefing: topics, people, brands and what changed.
  4. Open the posts behind anything you read, and build reports from what you find.

Channels

XLinkedIn

Both from the start. A person’s X and LinkedIn accounts are read as one profile.

Evidence first

Built so you can check it

  • Numbers are countedCounts, shares, ranks and changes are calculated. A language model never produces a number.
  • Sentences cite their postsGenerated text carries citations. Text that cites nothing is rejected before it is stored.
  • Judgements are labelledA model judgement such as stance is marked Indicative wherever it appears.
  • Your work stays yoursLists, reports and everything derived from them belong to one client and are visible to nobody else.

Who uses it

Made for agencies and in-house teams

Agencies

One subscription and one invoice. A separate client for each case you manage, with nothing shared between them.

In-house teams

Communications, analyst relations and marketing teams who need to know what a known set of people is saying about their market.

Try it before you decide

The demo opens the published lists and their weekly briefings.