Lilian Tellez

Work

About me

Resume

Lilian Tellez

LinkedIn • Follower growth

Growing LinkedIn's company page ecosystem

Overview

I led the design strategy for LinkedIn's company page follow ecosystem, coordinating three teams cross growth content experience and feed to run a connector the series of experiments the goal was to help company pages, especially is more ones, the discovered and followed by the right members.

My role

I set the vision and roadmap for this initiative, ran cross-team workshops to align Pages and People Profile teams, and personally designed the highest-impact experiments, recommendation quality improvements, follow loops, and the subscription feature. I kept the work coherent across three teams that each had their own priorities and timelines.

Impact

Increase in company page followers

All our experiments yielded a 6% increase in company page followers

A balanced ecosystem that accounts for SMBs

The increase in followers was mostly for small companies

Increase in engagement

12,000 monthly subscriptions to company pages guaranteeing repeated visits.

Challenge

LinkedIn, a platform designed to foster professional connections and economic opportunities, faced a challenge in its mission to create vibrant communities. While the platform excelled at connecting professionals and facilitating job searches, it struggled to surface meaningful content created by Company Pages, and facilitate follower growth, especially for those with less than 500 followers.

Two teams, Pages and Profile, aimed to improve the "follow" ecosystem. I led research and and data analysis, including a 400-person survey to gain insights to address the challenges, finding the following root causes:

The rise of competitors emphasizing streamlined payment experiences necessitated a critical evaluation of our own checkout flow. Our existing architecture, housing over 30 features, had become dense. A significant portion of the interface was occupied by less critical information, inadvertently obscuring valuable features like PayPal Credit and new flexible payment options. This complexity hindered feature discovery and adoption.

Our challenge was to strategically simplify the information architecture, prioritizing key payment methods and integrating new capabilities intuitively for a broad user base, without disrupting established user behaviors tied to our high conversion rate.

Members valued content, not company names

Members followed companies organically in their feed, motivated by valuable content and conversations. They defined page relevance as "companies in my industry, interest and activity on LinkedIn."

Members follow pages to stay updated

The top reasons for following a page were to stay current in one's area of expertise, while unfollows were primarily due to career path alterations or page inactivity.

A recommendation gap

Algorithms struggled to recommend creators with limited followers, preventing them from gaining visibility and building their audience.

Strategy

I organized a remote workshop with cross-functional teams, including People Profile, to align roadmaps and address shared pain points in the follow experience.
Due to their focus on Creator Profile, we agreed that I would lead initial follow explorations and involve them when needed. The diagram below shows topics for our joint roadmap Vs separate initiatives.

Recommendation quality

Enhance the UX of follow recommendations and related algorithms.

Pages visibility

expand the surfaces where follow recommendations were displayed, ensuring they were prominently featured across the platform

Engagement improvements

Create viral follow loops to help customers discover pages continuously

Content distribution

Distribute content from Pages and people in feeds, notifications and new capabilities like subscriptions.

Exploration and iteration

The recommendation improvements relied on LinkedIn machine learning powered affinity scoring industry, in engagement history influence recommendations, in ways members could act on.

As projects evolved, I maintained a collection of experiments and initiatives in a visual roadmap, paired with a slack channel for questions and daily check-ins.

A visual roadmap helped the pages and people teams zoom in and out on intiviatives

Roadmap execution

Recommendation quality improvements

Hypothesis

If we show social proof + content snippets in follow recommendations, customers will be more likely to be interested and follow Page creators.

Action

Ran 3 approaches through A/B tests to prove/disprove the hypothesis.

Impact

Variant C was the winner, with a 1.4% follower lift, and later implemented across all recommendations.

Experiment progression

Control: Recommendation module

A) Social proof + content (0.5% lift)

B) Content first (1.4 lift)

Trigger-based nudges

Hypotehesis

If we leverage moments when people show interest in a company (liking a post, commenting), to invite them to follow a company, more high-quality follows will be possible.

Action

Ran an experiment where after someone likes or comments on a post from a company, a nudge appears enticing people to follow its page.

Impact

This nudge increased follows for companies and proved to create higher quality followers (less prone to unfollow later).

Follow nudge triggered by a reaction

Discovery carousel triggered by a follow

Social proof enrichment

Hypotehesis

If we standardize the way people experience and see social proof for company pages, we will help people better understand how they are connected to companies, therefore encouraging follows.

Action

Worked on standardizing social proof across multiple instances across LinkedIn, estalbished rules, logic and repeatable patterns to foster a cohesive experience.

Impact

Social proof became one of the highest follow drivers for company pages.

Socual proof on pages and recommendations

Guidelines

Increase engagement and visibility

Follow loops

These were used to enhance the effectiveness of follows, with actions such as suggesting followed creators to a user's network and piggybacking on connection flows. They also improved content distribution by incorporating follow affinity scores.

How loops work:
Members A and B are connected. Member A follows a company, triggering a loop for members B, C, D… in their network - which in turn will trigger the loop back for member A.

Let's imagine two professionals at LinkedIn. Tia and Aarti, who are ex-coworkers, interested in growth marketing, with the same roles and a company in common. What Tia follows will trigger certain recommendations for Aarti, being the person in her circle with more affinities.

Learning about notifications

A way to keep companies present in people’s minds is through notifications.
The closer the notification was to the Member’s direct interests, interactions and connections, the better was at driving engagement.

Company leadership and employer notifications failed. The concept of “coworker” and “leader” can be uncertain and their posts can be often too personal to be relevant.

Subscription to notifications

Hypotehesis

Members want to receive content from Pages they follow, but rely solely on browsing their feed.

Action

Enabled subscription to top posts, delivering notifications and email digests to subscribed members

Impact

Average of 12,000 monthly subscriptions and increased content engagement.

Discovery carousel triggered by a follow

Discovery carousel triggered by a follow

Democratize companies' presence at LinkedIn

Hypotehesis

If members see pages present in their searches and visits for organic discovery, tied to topics they care about, they will be likely to show curiosity for companies and follow.

Action

Worked with multiple teams like search and editorial to add pages to their results and streams.

Impact

Visibility across different touch points increased follows in 4%

Newsletters

My Network

Final thoughts

The insight that changed everything was simple: members follow for content, not for company names. Once we designed around that, putting content quality signals at the front of every recommendation, the numbers moved. The 6% follower lift wasn't from one big feature. It was from a dozen small bets, each one building on what the last one proved.

That's the kind of work I find most interesting: Where the design and the data are in constant conversation.