Skip to content

Generate Ideas workflow


megaphone

NORTH STAR - INCREASING THE GENERATED IDEAS → POST RATIO



Problem Alignment


Problem

The current idea generation workflow is overly simplistic, it relies on only 1–2 inputs, which leads to outputs that are generic, repetitive, and difficult to act on.

CURRENT STATE
image.png
Lacks personalization
As a result:
Ideas lack context (audience, tone, platform, intent)
Outputs feel surface-level rather than insightful
Creators still need to heavily edit or rethink the idea before they finalize an idea

This creates a critical gap:
Instead of accelerating content creation, the system adds another layer of work.
For a platform like Buffer, whose core value lies in enabling consistent publishing, this is a fundamental issue.
If ideas are not high-quality and ready-to-use, they don’t convert into posts.

High-level Approach

Transform ideas from generic suggestions into context-aware starting points by tailoring them to:
The creator's voice and positioning
Their target audience
Platform-specific dynamics (LinkedIn ≠ Twitter ≠ Instagram)
Content intent (educate, entertain, promote)

Validation for v0

This direction is grounded in well-established creator behavior and platform dynamics.
It is widely understood that what works on one social platform does not necessarily work on another. Content that performs well on LinkedIn may fail on Twitter or Instagram due to differences in:
Audience expectations
Content formats
Attention spans
Platform algorithms
At the same time, audiences themselves behave differently across platforms, they perceive, consume, and engage with content in unique ways depending on context.
This creates a clear need:
Creators don’t just need ideas, they need platform-aware ideas.


Core Problems

1. Idea generation is still treated as a manual task

The current system positions idea generation as optional support:
“If you need help, you can use this.”
But for Buffer to drive retention and become central to the creator workflow, this needs to shift to:
“When I want to create, I start with Buffer.”
Today, creators don’t naturally think of Buffer when they want to write.

2. Low quality and generic ideas

Ideas generated are often:
Too broad
Not actionable
Lacking creator or audience context
This leads to:
Extra effort to refine ideas
Lower trust in the system
Reduced likelihood of publishing

3. No personalization

The system does not adapt to:
Creator’s voice or niche
Audience preferences
Platform-specific behavior
Creator setup
image.png
.
image.png

For something like Mother’s Day, it’s reasonable to expect creators to come up with a prompt or idea themselves, that’s part of their job.
But when we look at the current idea suggestions, many of them assume the creator will step outside their usual setup. That doesn’t always work. Some creators only produce content from their desk, and that’s exactly what their audience expects from them.
While there are cases where creators may want to experiment or try something new, we should first focus on supporting their daily workflow, what they consistently do.
Once that’s strong, we can then start solving for those occasional, more creative or out-of-routine scenarios with a feature something like SUPRISE ME

4. Missing context of real-world moments

Content creation is often driven by timely events and cultural moments, such as:
New Year’s Day
Christmas
Mother’s Day
Important events across the world
However, the current idea workflow does not account for these triggers at all.
This results in missed opportunities for:
Relevant content
Higher engagement
Timely posting

Solution Alignment

I want to approach this in two phases: V0 and V1.

V0 (Baseline Solution)

In V0, we focus on building a solid foundation by considering the key factors that define what makes a good contuantity
Users can later adjust how many ideas they want to see
ent idea for a creator.
Core parameters:
Tone & voice
Niche / industry
Primary audience
Recording setup (e.g., desk-based, outdoor, etc.)
Platform preference
Calendar events (e.g., Mother’s Day, festivals)
Auto-generation enabled by default
Idea generation approach:
Automatically generate content ideas
Default to 1 idea per day
This avoids overwhelming the user
Focus is on quality over q

Auto generation enabled

With ideas being generated automatically, users will be greeted with a relevant thought or idea every time they open their Buffer dashboard.
Why This Matters
Occupies valuable mindshare for creators
Positions Buffer as a proactive partner in content creation, not just a publishing tool
Strengthens daily engagement and habit formation

V1 (Intelligence Layer)

In V1, we make the system more intelligent by leveraging existing data from Buffer:
Past content and its performance
Creator’s channel persona
This allows for more personalized and higher-quality suggestions over time.

What This Solves

When a user opens their dashboard, they immediately have a content idea to evaluate
Removes the need for multiple steps and reduces friction in the workflow
Supports the creator’s daily content process, rather than interrupting it

High Fidelity Wireframes

This is just a concept design for the flow, users will use to personalize their content once.
image.png










Want to print your doc?
This is not the way.
Try clicking the ··· in the right corner or using a keyboard shortcut (
CtrlP
) instead.