Free sample lesson ยท How LinkedIn Unified AI Products Around a Target Audience Graph
Why fragmented products often mirror fragmented org charts
Lesson 1 of the Wondering course How LinkedIn Unified AI Products Around a Target Audience Graph
The LinkedIn Puzzle
A small business owner logs into LinkedIn. She wants to hire, market, and sell.
She finds three separate products, each with its own login, its own interface, its own logic.
Nothing about her needs is fragmented. So why is the product?
- The owner
The answer is not on the screen. It is in the org chart.
Separate Owners, Separate Products
LinkedIn's sales, marketing, and hiring tools lived in separate products.
The reason was simple. They were owned by separate divisions.
Each division had its own leadership, its own roadmap, its own targets.
Each shipped what it controlled. Nobody owned the space between them.
How Boundaries Leak In
Organizational lines do not stay inside the company.
A division owns a budget, so it builds its own thing. It owns a roadmap, so it sets its own priorities.
It is measured on its own numbers, so shared work loses.
- Separate division: A tool is assigned to one division to own.
- Own roadmap: That division plans and prioritizes independently.
- Own metrics: Success is measured inside the division, not across it.
- Separate product: The tool ships as its own product with its own experience.
- Visible seam: The customer feels the boundary as fragmentation.
The result reaches the customer as a seam in the product.
Nobody Chose Fragmentation
Here is the common misconception. People assume fragmented products come from bad product decisions.
Usually no one decided anything. Fragmentation is the default output when ownership is split.
Three teams doing good work in three lanes still produce three products.
Good intentions do not cross an org boundary on their own.
Reading Products Backwards
This gives you a diagnostic tool.
When a product feels stitched together, ask who owns each piece.
Inconsistent logins, duplicated data, mismatched interfaces. These are usually reporting lines showing through.
- Separate divisions: Each piece of the product has a different owner.
- Separate logins: Each tool carries its own entry point.
- Mismatched interfaces: Each product looks and behaves differently.
- Duplicated data: The same customer information sits in several places.
- Org chart: Reporting lines can be inferred from the product itself.
You can often sketch a company's structure by using its product alone.
Where Fixes Have To Land
If the cause is structural, the fix has to be structural.
Redesigning three interfaces to look alike does not merge three roadmaps.
Unified products usually follow unified ownership: one leader, one budget, one measure of success.
- Surface fix
- Structural fix
Change the boundary, and the product can finally follow the customer instead of the org chart.
Check yourself
LinkedIn's separate sales, marketing, and hiring products were associated with separate owning divisions.
- True
- False
Answer: True
Each product belonged to a different division with its own leadership, roadmap, and targets. The separation in ownership appeared as separation in the product.
A division controls its _____, sets its _____, and is judged by its _____.
Answer: A division controls its budget, sets its roadmap, and is judged by its targets.
These three forms of control reinforce organizational separation: budgets determine resources, roadmaps determine priorities, and targets determine how performance is evaluated.
A product has three login systems, duplicated customer data, and interfaces that do not match. What is the most useful first diagnostic question?
- Who owns each part of the product?
- Which interface uses the most modern colors?
- Which team released its feature most recently?
- Which login system has the shortest password form?
Answer: Who owns each part of the product?
The recommended diagnostic is to identify who owns each piece. These product seams may reflect separate reporting lines, budgets, roadmaps, or success measures.
Keep going
This is lesson 1 of 9 in How LinkedIn Unified AI Products Around a Target Audience Graph. Next up:
- The cost of shipping parallel AI features instead of shared foundations
- Cold start: inferring audience with LLMs
- Warm signals: time-decayed activity weighting
- Thin markets: fusing buyer-side and seller-side signals
- Embedding retrieval + multi-stage ranking
Wondering turns any topic into short lessons you can read, watch, or listen to, with practice questions and an AI tutor.