Describe the customer
Start with one honest sentence about the company, person, problem, and outcome you believe matter most.
ICP & positioning / 01
Stop writing personas from memory.
Know who the product is for—and why they care.
A living ideal customer profile with the evidence, uncertainty, and language still attached.
Specky reads the customer signals already in your workspace, drafts a practical ICP, and carries it into research, prioritization, validation, and every feature launch.
- Real signals stay linked to every segment
- Assumptions remain visibly unconfirmed
- Per-product ICP with reusable customer segments
An evidence-backed ICP connecting customer signals to product and marketing decisions
The job
Turn scattered customer knowledge into a customer definition the team can actually use.
Most ICP documents become a polished collection of guesses. Specky keeps the customer job, buying trigger, pains, desired outcomes, and exact language connected to the sources that support them.
Start with a question
“Who gets the most value from this product today—and what evidence would change our mind?”
InterviewsFeedbackCallsAnalytics
Start with your current customer belief and the real interviews, calls, feedback, support, or behavior that could sharpen—or challenge—it.
Specky returns
A living ICP the whole product loop can use
- Primary, secondary, and experimental segments
- Jobs, pains, triggers, outcomes, objections, and customer language
- Clear separation between evidence and assumption
01
Collect
Bring in the conversations and behaviour that reveal the real job.
02
Draft
Turn repeated patterns into a specific working profile.
03
Review
Inspect citations, contradictions, and missing evidence.
04
Activate
Use the profile in research, planning, experiments, and launches.
What goes in / what comes out
Clear inputs. Useful work back.
The feature should be understandable before anyone connects a source or runs an AI generation.
You give Specky
- Your current belief about the best-fit customer
- Customer interviews, feedback, calls, support, CRM, or analytics
- The product or workspace this ICP belongs to
- Human judgment about which segment the team will prioritize
Specky gives back
- A concise ICP snapshot with company and buying context
- Personas, jobs to be done, pains, trigger events, and desired outcomes
- Primary, secondary, and experimental segments with fit and evidence strength
- Customer language, objections, anti-ICP, and open evidence gaps
- Direct actions for interviews, experiments, planning, and feature marketing
The workflow
From a fuzzy belief to work the team can use.
Each step leaves a reviewable artifact. People stay responsible for the claim, the decision, and what ships.
Read the evidence
Specky checks real workspace signals and only links source IDs that exist in the Product Graph.
Separate fit from certainty
A strategically attractive segment can still have weak evidence. Both are visible so the team knows what to learn next.
Use it everywhere
Carry the selected segment and customer language into the next interview, priority decision, experiment, or launch.
A concrete example
From “B2B SaaS teams” to a segment you can find and help.
A broad audience does not tell product, sales, or marketing what to do. A useful ICP makes the buying situation and customer job specific.
Starting point
- 1Interview notes from six growth PMs
- 2PostHog questions about activation
- 3Support threads about fragmented context
- 4A founder's belief about 20–200 person SaaS companies
Useful result
- Primary segment: growth PMs at scaling B2B SaaS teams
- Trigger: activation changes with no shared explanation
- Job: connect behaviour and customer language before choosing a test
- Evidence gap: buying authority and willingness to switch
Questions
The practical details
What is an evidence-backed ICP?
It is a working definition of your best-fit customer where pains, jobs, triggers, language, and segment claims can be traced to real product evidence. Assumptions remain visible instead of being rewritten as facts.
What data can Specky use?
Specky can use customer interviews, feedback, support conversations, sales and call transcripts, product analytics, feature requests, documents, and other real signals already connected to the Product Graph.
Does Specky decide our ICP for us?
No. Specky drafts the profile, shows the supporting evidence and gaps, and lets the product team edit or validate it. The team remains responsible for strategic choices.
Can one product have multiple segments?
Yes. Mark segments as primary, secondary, or experimental. Each segment has its own jobs, pains, desired outcomes, fit score, and evidence strength.
How does the ICP connect to product work?
Use it to design interviews, evaluate opportunities, target experiments, prioritize work, and create feature marketing for a specific audience rather than writing for everyone.
Make the customer definition useful again.
Build the first profile from what you know, then let every new signal make it sharper.