<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Ben Segni</title>
    <description>The latest articles on DEV Community by Ben Segni (@ben_segni_4c5eacf9af147ab).</description>
    <link>https://dev.to/ben_segni_4c5eacf9af147ab</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F2315840%2F98f1d045-3fd4-4b7b-a6fe-03d6e5ecbaf4.jpg</url>
      <title>DEV Community: Ben Segni</title>
      <link>https://dev.to/ben_segni_4c5eacf9af147ab</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/ben_segni_4c5eacf9af147ab"/>
    <language>en</language>
    <item>
      <title>One GitHub Issue, Six Perspectives: How AI Review Profiles Improve Software Estimation</title>
      <dc:creator>Ben Segni</dc:creator>
      <pubDate>Wed, 11 Nov 2026 11:38:34 +0000</pubDate>
      <link>https://dev.to/ben_segni_4c5eacf9af147ab/one-github-issue-six-perspectives-how-ai-review-profiles-improve-software-estimation-2i9f</link>
      <guid>https://dev.to/ben_segni_4c5eacf9af147ab/one-github-issue-six-perspectives-how-ai-review-profiles-improve-software-estimation-2i9f</guid>
      <description>&lt;h2&gt;
  
  
  One GitHub Issue, Six Perspectives: How AI Review Profiles Improve Software Estimation
&lt;/h2&gt;

&lt;p&gt;Most developers have experienced this during backlog refinement.&lt;/p&gt;

&lt;p&gt;A GitHub issue looks straightforward. The requirements seem reasonable, the implementation appears manageable, and the team agrees on an estimate.&lt;/p&gt;

&lt;p&gt;Then development starts.&lt;/p&gt;

&lt;p&gt;Someone identifies a missing error state. QA raises an edge case. A security concern emerges. The Product Owner clarifies a requirement that changes the scope entirely.&lt;/p&gt;

&lt;p&gt;Suddenly, that five-point issue looks considerably more complicated.&lt;/p&gt;

&lt;p&gt;The problem isn't necessarily poor estimation. It's that &lt;strong&gt;we tend to evaluate requirements through the lens of our own experience.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;What happens when we deliberately look at the same issue from different professional perspectives?&lt;/p&gt;

&lt;h2&gt;
  
  
  One issue, several interpretations
&lt;/h2&gt;

&lt;p&gt;Imagine a GitHub issue with the following requirement:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;Allow users to update their notification preferences, including email notifications and in-app alerts.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;As a developer, my immediate questions might be:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Does an existing API support these preferences?&lt;/li&gt;
&lt;li&gt;Where will the settings be stored?&lt;/li&gt;
&lt;li&gt;Can we reuse the current settings components?&lt;/li&gt;
&lt;li&gt;How should the frontend handle failed requests?&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;These are sensible technical questions.&lt;/p&gt;

&lt;p&gt;But they aren't the only questions worth asking.&lt;/p&gt;

&lt;p&gt;A UX Designer might focus on how users understand the available options.&lt;/p&gt;

&lt;p&gt;A QA Engineer might consider what happens when a user changes preferences across multiple devices.&lt;/p&gt;

&lt;p&gt;A Security Engineer might question whether users can modify another account's preferences.&lt;/p&gt;

&lt;p&gt;All of these perspectives can influence the implementation, testing effort and ultimately the estimate.&lt;/p&gt;

&lt;h2&gt;
  
  
  Introducing review profiles
&lt;/h2&gt;

&lt;p&gt;This is the thinking behind the optional &lt;a href="https://pointwise-estimates.tech/" rel="noopener noreferrer"&gt;AI review profiles in Pointwise&lt;/a&gt;, a GitHub-integrated Planning Poker application.&lt;/p&gt;

&lt;p&gt;Pointwise uses GitHub Copilot to review an issue from a selected professional perspective.&lt;/p&gt;

&lt;p&gt;Rather than asking AI to provide a generic assessment, contributors can choose a profile that influences which aspects of the requirements receive attention.&lt;/p&gt;

&lt;p&gt;There are currently six profiles.&lt;/p&gt;

&lt;h2&gt;
  
  
  1. Engineer: Is the implementation understood?
&lt;/h2&gt;

&lt;p&gt;The Engineer profile focuses on technical implementation.&lt;/p&gt;

&lt;p&gt;For our notification preferences example, useful questions might include:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Are existing services and components reusable?&lt;/li&gt;
&lt;li&gt;Does the change introduce new API contracts?&lt;/li&gt;
&lt;li&gt;Are database changes required?&lt;/li&gt;
&lt;li&gt;What dependencies could affect delivery?&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The objective is to expose technical assumptions and potential implementation complexity.&lt;/p&gt;

&lt;p&gt;This is particularly useful when a requirement appears simple at the UI level but may involve several backend services.&lt;/p&gt;

&lt;h2&gt;
  
  
  2. UX Designer: Does the experience make sense?
&lt;/h2&gt;

&lt;p&gt;Developers often receive requirements describing what a feature should do without fully explaining how users will interact with it.&lt;/p&gt;

&lt;p&gt;The UX Designer perspective encourages questions about usability, accessibility and interaction behaviour.&lt;/p&gt;

&lt;p&gt;For example:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Can users understand the difference between notification types?&lt;/li&gt;
&lt;li&gt;Is it obvious when their preferences have been saved?&lt;/li&gt;
&lt;li&gt;What feedback appears if saving fails?&lt;/li&gt;
&lt;li&gt;Can keyboard and assistive technology users operate the controls?&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;These aren't merely visual design considerations.&lt;/p&gt;

&lt;p&gt;They can introduce additional states, components and acceptance criteria that affect development effort.&lt;/p&gt;

&lt;h2&gt;
  
  
  3. Product Owner: Are we solving the right problem?
&lt;/h2&gt;

&lt;p&gt;A technically complete implementation isn't necessarily a useful product feature.&lt;/p&gt;

&lt;p&gt;The Product Owner profile considers requirements in terms of user value, scope and intended outcomes.&lt;/p&gt;

&lt;p&gt;For our example:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Why do users need notification preferences?&lt;/li&gt;
&lt;li&gt;Which notification types are included in the initial release?&lt;/li&gt;
&lt;li&gt;What should happen to existing users?&lt;/li&gt;
&lt;li&gt;What constitutes successful delivery?&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;This perspective can reveal requirements that are either missing or unnecessarily broad.&lt;/p&gt;

&lt;p&gt;Sometimes the most valuable discovery is that the team can deliver a smaller, more focused solution.&lt;/p&gt;

&lt;h2&gt;
  
  
  4. QA Engineer: What could go wrong?
&lt;/h2&gt;

&lt;p&gt;QA engineers are particularly good at identifying situations that aren't covered by the happy path.&lt;/p&gt;

&lt;p&gt;Consider the notification preferences feature.&lt;/p&gt;

&lt;p&gt;What happens if the API request fails?&lt;/p&gt;

&lt;p&gt;What if the user submits changes twice?&lt;/p&gt;

&lt;p&gt;What if preferences are modified from another device?&lt;/p&gt;

&lt;p&gt;What happens when an existing user has no saved preferences?&lt;/p&gt;

&lt;p&gt;The QA Engineer perspective encourages examination of edge cases, validation, error handling and testability.&lt;/p&gt;

&lt;p&gt;These questions matter because unexpected behaviour doesn't become cheaper to address simply because nobody mentioned it during refinement.&lt;/p&gt;

&lt;h2&gt;
  
  
  5. Business Analyst: Are the requirements complete?
&lt;/h2&gt;

&lt;p&gt;The Business Analyst profile looks at requirements, business rules and dependencies.&lt;/p&gt;

&lt;p&gt;It can help identify ambiguity between what stakeholders expect and what developers believe they're implementing.&lt;/p&gt;

&lt;p&gt;For example:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Are notification defaults defined?&lt;/li&gt;
&lt;li&gt;Do different user roles have different permissions?&lt;/li&gt;
&lt;li&gt;Are there business rules governing mandatory notifications?&lt;/li&gt;
&lt;li&gt;Does changing a preference affect other application workflows?&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;This perspective is particularly useful for enterprise applications where seemingly isolated features often interact with wider business processes.&lt;/p&gt;

&lt;h2&gt;
  
  
  6. Security Engineer: What are we exposing?
&lt;/h2&gt;

&lt;p&gt;Security requirements can be overlooked when the discussion focuses primarily on delivering functionality.&lt;/p&gt;

&lt;p&gt;For notification preferences, a security-focused review might ask:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Can users modify preferences belonging to another account?&lt;/li&gt;
&lt;li&gt;Are API endpoints properly authorised?&lt;/li&gt;
&lt;li&gt;Could sensitive information appear in logs?&lt;/li&gt;
&lt;li&gt;Are changes validated on the server?&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Not every issue needs a lengthy security assessment.&lt;/p&gt;

&lt;p&gt;But identifying an authorisation problem before development is considerably better than discovering it after release.&lt;/p&gt;

&lt;h2&gt;
  
  
  AI should challenge assumptions, not make decisions
&lt;/h2&gt;

&lt;p&gt;There's an important distinction between using AI to review requirements and allowing AI to determine how much work something involves.&lt;/p&gt;

&lt;p&gt;AI doesn't necessarily understand the complete codebase, the team's experience or every architectural constraint.&lt;/p&gt;

&lt;p&gt;It can also misunderstand requirements or raise concerns that aren't relevant.&lt;/p&gt;

&lt;p&gt;That's why I see these reviews as &lt;strong&gt;additional perspectives rather than authoritative answers.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;If an AI review suggests that a database migration might be required, the team should verify whether that's true.&lt;/p&gt;

&lt;p&gt;If it identifies an accessibility concern, developers should evaluate the actual interaction.&lt;/p&gt;

&lt;p&gt;The value comes from prompting useful conversations, not blindly accepting generated recommendations.&lt;/p&gt;

&lt;h2&gt;
  
  
  Where review profiles fit into Planning Poker
&lt;/h2&gt;

&lt;p&gt;In &lt;a href="https://pointwise-estimates.tech/" rel="noopener noreferrer"&gt;Pointwise's collaborative Planning Poker workflow&lt;/a&gt;, contributors estimate independently before their scores are revealed.&lt;/p&gt;

&lt;p&gt;The team can then discuss differences and reach consensus.&lt;/p&gt;

&lt;p&gt;An optional Copilot review provides another way to examine the issue, particularly when requirements are unclear or contributors have identified competing assumptions.&lt;/p&gt;

&lt;p&gt;For example, a team might initially estimate five points.&lt;/p&gt;

&lt;p&gt;A QA-focused review could raise several missing error states.&lt;/p&gt;

&lt;p&gt;After investigating those concerns, the team might decide the original estimate remains appropriate, or that the scope needs clarification.&lt;/p&gt;

&lt;p&gt;Either outcome is useful.&lt;/p&gt;

&lt;p&gt;The AI hasn't replaced the team's judgement. It's helped the team ask better questions.&lt;/p&gt;

&lt;h2&gt;
  
  
  See Pointwise in action
&lt;/h2&gt;

&lt;p&gt;I've built these review profiles into &lt;a href="https://pointwise-estimates.tech/" rel="noopener noreferrer"&gt;Pointwise&lt;/a&gt; to make it easier to explore requirements from different perspectives without leaving the GitHub issue and estimation workflow.&lt;/p&gt;

&lt;p&gt;The walkthrough below demonstrates the collaborative estimation process and optional Copilot review.&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/AyxBZgLLidA" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;p&gt;&lt;a href="https://youtu.be/AyxBZgLLidA?si=lAzb_COrcrbwNb_S" rel="noopener noreferrer"&gt;Watch the Pointwise walkthrough on YouTube&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Final thoughts
&lt;/h2&gt;

&lt;p&gt;Software estimation isn't just about understanding how difficult something will be to code.&lt;/p&gt;

&lt;p&gt;It's about understanding the work well enough to make an informed decision.&lt;/p&gt;

&lt;p&gt;Engineers, designers, testers, product owners, business analysts and security specialists all approach requirements differently.&lt;/p&gt;

&lt;p&gt;Those differences are valuable.&lt;/p&gt;

&lt;p&gt;AI review profiles offer a practical way to introduce additional questions into refinement, especially when every specialist isn't available for every discussion.&lt;/p&gt;

&lt;p&gt;But the responsibility for evaluating those questions must remain with the team.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The goal isn't to let AI tell developers what they've missed. It's to help developers discover what might be worth discussing before implementation begins.&lt;/strong&gt;&lt;/p&gt;




&lt;p&gt;&lt;strong&gt;Which perspective tends to uncover the most unexpected complexity in your team's requirements: engineering, UX, QA, product or security?&lt;/strong&gt;&lt;/p&gt;

</description>
      <category>agile</category>
      <category>ai</category>
      <category>github</category>
      <category>productivity</category>
    </item>
    <item>
      <title>Planning Poker: Why Independent Estimates Matter</title>
      <dc:creator>Ben Segni</dc:creator>
      <pubDate>Mon, 09 Nov 2026 10:59:02 +0000</pubDate>
      <link>https://dev.to/ben_segni_4c5eacf9af147ab/planning-poker-why-independent-estimates-matter-19n7</link>
      <guid>https://dev.to/ben_segni_4c5eacf9af147ab/planning-poker-why-independent-estimates-matter-19n7</guid>
      <description>&lt;h2&gt;
  
  
  Planning Poker: Why Independent Estimates Matter
&lt;/h2&gt;

&lt;p&gt;You've probably seen this happen during a refinement session.&lt;/p&gt;

&lt;p&gt;A developer reads through a ticket, considers the requirements and says, "Looks like a five to me."&lt;/p&gt;

&lt;p&gt;They're experienced. They know the codebase. They've worked on similar features before.&lt;/p&gt;

&lt;p&gt;A few moments later, the rest of the team agrees.&lt;/p&gt;

&lt;p&gt;Five points it is.&lt;/p&gt;

&lt;p&gt;The meeting moves on, everyone feels productive, and nobody particularly questions the outcome.&lt;/p&gt;

&lt;p&gt;But did the team genuinely agree? Or did everyone simply accept the first reasonable number?&lt;/p&gt;

&lt;p&gt;This is one of the problems Planning Poker is designed to address.&lt;/p&gt;

&lt;h2&gt;
  
  
  The first estimate changes the conversation
&lt;/h2&gt;

&lt;p&gt;Imagine four developers discussing a GitHub issue to add notification preferences to an application.&lt;/p&gt;

&lt;p&gt;The most experienced developer estimates three points.&lt;/p&gt;

&lt;p&gt;Another developer initially thinks eight because the change might require updates to an existing API and database schema.&lt;/p&gt;

&lt;p&gt;But after hearing three, they reconsider.&lt;/p&gt;

&lt;p&gt;Perhaps they've overcomplicated it. Perhaps the senior developer knows something they don't.&lt;/p&gt;

&lt;p&gt;They settle on five.&lt;/p&gt;

&lt;p&gt;The remaining developers follow with three and five.&lt;/p&gt;

&lt;p&gt;The team quickly reaches a consensus.&lt;/p&gt;

&lt;p&gt;Except the original concern about the API and database changes never gets discussed.&lt;/p&gt;

&lt;p&gt;This is an example of anchoring: the tendency for an initial piece of information to influence subsequent judgements.&lt;/p&gt;

&lt;p&gt;It doesn't require anyone to behave badly. The senior developer might have offered their estimate with the best intentions.&lt;/p&gt;

&lt;p&gt;But once a number is in the room, it can influence how everyone else evaluates the work.&lt;/p&gt;

&lt;p&gt;And in a team with different levels of experience, that influence can be considerable.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why Planning Poker hides estimates
&lt;/h2&gt;

&lt;p&gt;The central idea behind Planning Poker is simple.&lt;/p&gt;

&lt;p&gt;Everyone considers the same requirement and selects an estimate independently.&lt;/p&gt;

&lt;p&gt;The estimates remain hidden until everyone has made their choice.&lt;/p&gt;

&lt;p&gt;Then they're revealed together.&lt;/p&gt;

&lt;p&gt;This small change removes the opportunity for the first estimate to become the reference point for everyone else.&lt;/p&gt;

&lt;p&gt;It also encourages developers to think about the work before seeing what their colleagues believe.&lt;/p&gt;

&lt;p&gt;That matters because different developers bring different knowledge.&lt;/p&gt;

&lt;p&gt;A frontend engineer might recognise complexity in component state management.&lt;/p&gt;

&lt;p&gt;A backend engineer might identify an API compatibility issue.&lt;/p&gt;

&lt;p&gt;A QA engineer might notice that the acceptance criteria don't describe what happens when a request fails.&lt;/p&gt;

&lt;p&gt;None of these perspectives should automatically outweigh the others.&lt;/p&gt;

&lt;p&gt;Independent estimation gives them an opportunity to surface.&lt;/p&gt;

&lt;h2&gt;
  
  
  Disagreement is useful information
&lt;/h2&gt;

&lt;p&gt;Let's return to the notification preferences issue.&lt;/p&gt;

&lt;p&gt;This time, the team estimates independently.&lt;/p&gt;

&lt;p&gt;The scores are revealed:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Contributor&lt;/th&gt;
&lt;th&gt;Estimate&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Developer A&lt;/td&gt;
&lt;td&gt;3&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Developer B&lt;/td&gt;
&lt;td&gt;5&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Developer C&lt;/td&gt;
&lt;td&gt;8&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Developer D&lt;/td&gt;
&lt;td&gt;13&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;At first glance, that might look like a failed estimation round.&lt;/p&gt;

&lt;p&gt;Four people, four different estimates.&lt;/p&gt;

&lt;p&gt;But it's potentially much more useful than four people agreeing on five without discussion.&lt;/p&gt;

&lt;p&gt;The next question shouldn't be, "How do we get everyone to agree?"&lt;/p&gt;

&lt;p&gt;It should be, "What is everyone seeing differently?"&lt;/p&gt;

&lt;p&gt;Developer A explains that the application already has a settings component that could be reused.&lt;/p&gt;

&lt;p&gt;Developer D points out that notification preferences are currently stored differently across two services.&lt;/p&gt;

&lt;p&gt;Developer C raises a concern about how existing users will receive default settings.&lt;/p&gt;

&lt;p&gt;Suddenly, the conversation isn't about numbers.&lt;/p&gt;

&lt;p&gt;It's about architecture, dependencies and behaviour.&lt;/p&gt;

&lt;p&gt;The estimates have exposed assumptions that might otherwise have remained hidden until implementation.&lt;/p&gt;

&lt;p&gt;That's valuable.&lt;/p&gt;

&lt;h2&gt;
  
  
  Don't automatically average the estimates
&lt;/h2&gt;

&lt;p&gt;A tempting response to disagreement is to calculate an average.&lt;/p&gt;

&lt;p&gt;Three, five, eight and thirteen produce an average of 7.25.&lt;/p&gt;

&lt;p&gt;Round it to eight and move on.&lt;/p&gt;

&lt;p&gt;Convenient, but largely pointless.&lt;/p&gt;

&lt;p&gt;Story points aren't precise measurements, and averaging them doesn't resolve the underlying differences in understanding.&lt;/p&gt;

&lt;p&gt;If one developer believes a database migration is necessary and another believes it isn't, choosing eight points doesn't answer that question.&lt;/p&gt;

&lt;p&gt;The team needs to investigate the disagreement.&lt;/p&gt;

&lt;p&gt;Perhaps the migration is unnecessary.&lt;/p&gt;

&lt;p&gt;Perhaps the requirements need updating.&lt;/p&gt;

&lt;p&gt;Perhaps the issue should be split into smaller pieces of work.&lt;/p&gt;

&lt;p&gt;Only after resolving the important assumptions should the team consider estimating again.&lt;/p&gt;

&lt;h2&gt;
  
  
  Senior developers still have an important role
&lt;/h2&gt;

&lt;p&gt;None of this means experienced engineers should remain silent.&lt;/p&gt;

&lt;p&gt;Their knowledge is often essential to understanding the work.&lt;/p&gt;

&lt;p&gt;They may recognise architectural constraints, identify opportunities to reuse existing code or explain why an apparently simple change has wider consequences.&lt;/p&gt;

&lt;p&gt;The distinction is between sharing technical context and establishing the estimate before everyone else has considered it.&lt;/p&gt;

&lt;p&gt;A senior developer saying, "There's an existing service we can reuse here," adds useful information.&lt;/p&gt;

&lt;p&gt;A senior developer opening with, "This is definitely three points," may unintentionally narrow the discussion.&lt;/p&gt;

&lt;p&gt;The objective isn't to remove experience from the process.&lt;/p&gt;

&lt;p&gt;It's to ensure that experience informs the team's judgement rather than replacing it.&lt;/p&gt;

&lt;h2&gt;
  
  
  What happens after the reveal?
&lt;/h2&gt;

&lt;p&gt;A productive Planning Poker session doesn't need endless rounds of voting.&lt;/p&gt;

&lt;p&gt;A practical approach is:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Give everyone time to review the requirement.&lt;/li&gt;
&lt;li&gt;Ask contributors to choose estimates independently.&lt;/li&gt;
&lt;li&gt;Reveal all estimates simultaneously.&lt;/li&gt;
&lt;li&gt;Invite people with different estimates to explain their reasoning.&lt;/li&gt;
&lt;li&gt;Identify any missing requirements, dependencies or risks.&lt;/li&gt;
&lt;li&gt;Estimate again if the discussion has materially changed the team's understanding.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Sometimes the second round produces agreement.&lt;/p&gt;

&lt;p&gt;Sometimes it doesn't.&lt;/p&gt;

&lt;p&gt;That's not necessarily a problem.&lt;/p&gt;

&lt;p&gt;If significant uncertainty remains, the team might need to investigate the work before committing to an estimate.&lt;/p&gt;

&lt;p&gt;Forcing consensus simply to complete refinement can create false confidence.&lt;/p&gt;

&lt;p&gt;And false confidence has a habit of becoming somebody else's problem halfway through a sprint.&lt;/p&gt;

&lt;h2&gt;
  
  
  Consensus doesn't mean identical thinking
&lt;/h2&gt;

&lt;p&gt;A team doesn't need everyone to approach a problem in exactly the same way.&lt;/p&gt;

&lt;p&gt;In fact, that would remove much of the value of collaborative estimation.&lt;/p&gt;

&lt;p&gt;What matters is that everyone understands the important assumptions behind the agreed estimate.&lt;/p&gt;

&lt;p&gt;A developer might still personally consider the work closer to eight points than five.&lt;/p&gt;

&lt;p&gt;But if the team has discussed the risks, clarified the scope and reached a reasoned decision, that's a much stronger outcome than simply following the first number suggested.&lt;/p&gt;

&lt;p&gt;Consensus should be the result of shared understanding, not social pressure.&lt;/p&gt;

&lt;h2&gt;
  
  
  See Pointwise in action
&lt;/h2&gt;

&lt;p&gt;This principle of independent estimates is central to &lt;a href="https://github.com/marketplace/pointwise-estimation" rel="noopener noreferrer"&gt;Pointwise&lt;/a&gt;, the GitHub-integrated Planning Poker application I've been building.&lt;/p&gt;

&lt;p&gt;Contributors select estimates independently, scores remain hidden until the session leader reveals them, and the GitHub issue stays available throughout the discussion.&lt;/p&gt;

&lt;p&gt;Here's a short walkthrough of the workflow:&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/AyxBZgLLidA" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;p&gt;&lt;a href="https://youtu.be/AyxBZgLLidA?si=lAzb_COrcrbwNb_S" rel="noopener noreferrer"&gt;Watch the Pointwise walkthrough on YouTube&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Final thoughts
&lt;/h2&gt;

&lt;p&gt;Planning Poker isn't valuable because it produces a number.&lt;/p&gt;

&lt;p&gt;It's valuable because it creates space for different perspectives before the team reaches a decision.&lt;/p&gt;

&lt;p&gt;Independent estimates reduce the influence of the first opinion.&lt;/p&gt;

&lt;p&gt;Simultaneous reveals make disagreement visible.&lt;/p&gt;

&lt;p&gt;And the discussion that follows helps the team understand the work more clearly.&lt;/p&gt;

&lt;p&gt;The next time everyone immediately agrees with the most experienced developer, it might be worth asking whether you've reached genuine consensus or simply skipped the conversation.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;How does your team handle large differences in estimates? Do you explore the reasoning or try to reach agreement quickly?&lt;/strong&gt;&lt;/p&gt;

</description>
      <category>agile</category>
      <category>programming</category>
      <category>productivity</category>
      <category>teamwork</category>
    </item>
    <item>
      <title>Why Story Point Estimates Go Wrong Before Planning Poker Even Starts</title>
      <dc:creator>Ben Segni</dc:creator>
      <pubDate>Mon, 02 Nov 2026 12:45:40 +0000</pubDate>
      <link>https://dev.to/ben_segni_4c5eacf9af147ab/why-story-point-estimates-go-wrong-before-planning-poker-even-starts-20cp</link>
      <guid>https://dev.to/ben_segni_4c5eacf9af147ab/why-story-point-estimates-go-wrong-before-planning-poker-even-starts-20cp</guid>
      <description>&lt;h1&gt;
  
  
  Why Story Point Estimates Go Wrong Before Planning Poker Even Starts
&lt;/h1&gt;

&lt;p&gt;You've probably been in this refinement session.&lt;/p&gt;

&lt;p&gt;A ticket appears on screen. Someone reads the title, there's a brief discussion, and the team starts estimating.&lt;/p&gt;

&lt;p&gt;One developer says three points. Another says eight. Someone else is convinced it's a thirteen.&lt;/p&gt;

&lt;p&gt;Twenty minutes later, everyone's still debating the number.&lt;/p&gt;

&lt;p&gt;But here's the uncomfortable question: &lt;strong&gt;did everyone actually understand the same piece of work?&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;In my experience, many estimation problems begin long before anyone chooses a story point value.&lt;/p&gt;

&lt;h2&gt;
  
  
  The problem isn't always the estimate
&lt;/h2&gt;

&lt;p&gt;Imagine a GitHub issue titled:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;Allow users to manage their notification preferences.&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;Seems straightforward.&lt;/p&gt;

&lt;p&gt;One developer assumes it means adding three checkboxes to an existing settings page.&lt;/p&gt;

&lt;p&gt;Another assumes it requires a new API endpoint, database changes and validation.&lt;/p&gt;

&lt;p&gt;A third is thinking about backwards compatibility, default preferences and whether existing users need migrating.&lt;/p&gt;

&lt;p&gt;They could reasonably arrive at three, eight and thirteen points.&lt;/p&gt;

&lt;p&gt;And none of them is necessarily wrong.&lt;/p&gt;

&lt;p&gt;They're estimating different interpretations of the same requirement.&lt;/p&gt;

&lt;p&gt;Planning Poker might expose that disagreement, but it cannot magically supply the missing information.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;An estimate is only as useful as the shared understanding behind it.&lt;/strong&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Story points don't fix unclear requirements
&lt;/h2&gt;

&lt;p&gt;Story points are generally used to express relative effort, complexity and uncertainty rather than a precise number of hours.&lt;/p&gt;

&lt;p&gt;But teams sometimes treat them as though they're a measurement instrument.&lt;/p&gt;

&lt;p&gt;Three points becomes a small task. Five becomes medium. Eight becomes difficult.&lt;/p&gt;

&lt;p&gt;Eventually, somebody asks why an eight-point issue took three days when the last eight-point issue took six.&lt;/p&gt;

&lt;p&gt;That's when things get messy.&lt;/p&gt;

&lt;p&gt;The problem is that two pieces of work can receive the same estimate for entirely different reasons.&lt;/p&gt;

&lt;p&gt;One might be technically complicated but well understood.&lt;/p&gt;

&lt;p&gt;Another might be straightforward to implement, but surrounded by unanswered questions.&lt;/p&gt;

&lt;p&gt;Both could reasonably receive eight points.&lt;/p&gt;

&lt;p&gt;What matters is understanding &lt;em&gt;why&lt;/em&gt;.&lt;/p&gt;

&lt;p&gt;If that reasoning never enters the discussion, the final estimate hides the very information the team needs.&lt;/p&gt;

&lt;h2&gt;
  
  
  Context gets lost between tools
&lt;/h2&gt;

&lt;p&gt;Most development teams don't keep every relevant detail in one place.&lt;/p&gt;

&lt;p&gt;The requirements might be in GitHub.&lt;/p&gt;

&lt;p&gt;The design could be in Figma.&lt;/p&gt;

&lt;p&gt;An important architectural decision might be buried in a document.&lt;/p&gt;

&lt;p&gt;And the one person who understands an existing integration might be on annual leave.&lt;/p&gt;

&lt;p&gt;During estimation, developers end up switching between tabs, searching for comments and trying to reconstruct decisions made weeks earlier.&lt;/p&gt;

&lt;p&gt;Sometimes they simply estimate based on what they can see.&lt;/p&gt;

&lt;p&gt;This creates a subtle problem: &lt;strong&gt;everyone may be looking at the same ticket, but not everyone has the same context.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;A developer familiar with the existing implementation may see a small change.&lt;/p&gt;

&lt;p&gt;Someone unfamiliar with it may anticipate considerable risk.&lt;/p&gt;

&lt;p&gt;The useful conversation isn't necessarily about whose estimate is correct.&lt;/p&gt;

&lt;p&gt;It's about what each person knows that the others don't.&lt;/p&gt;

&lt;h2&gt;
  
  
  Planning Poker works best when it reveals assumptions
&lt;/h2&gt;

&lt;p&gt;Planning Poker has a valuable mechanism: participants choose estimates independently before revealing them together.&lt;/p&gt;

&lt;p&gt;This reduces the opportunity for the first estimate to influence everyone else's thinking.&lt;/p&gt;

&lt;p&gt;But the reveal isn't the finish line.&lt;/p&gt;

&lt;p&gt;It's where the useful discussion begins.&lt;/p&gt;

&lt;p&gt;Suppose the team reveals:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Developer&lt;/th&gt;
&lt;th&gt;Estimate&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Developer A&lt;/td&gt;
&lt;td&gt;3&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Developer B&lt;/td&gt;
&lt;td&gt;5&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Developer C&lt;/td&gt;
&lt;td&gt;13&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Developer D&lt;/td&gt;
&lt;td&gt;5&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;The instinct might be to ask Developer C why they've estimated thirteen.&lt;/p&gt;

&lt;p&gt;That's reasonable, but the three-point estimate deserves attention too.&lt;/p&gt;

&lt;p&gt;Perhaps Developer A knows an existing component can be reused.&lt;/p&gt;

&lt;p&gt;Perhaps Developer C has spotted a security requirement nobody else considered.&lt;/p&gt;

&lt;p&gt;Both perspectives could change the implementation approach.&lt;/p&gt;

&lt;p&gt;The purpose of the discussion shouldn't be to persuade the outliers to choose five.&lt;/p&gt;

&lt;p&gt;It should be to discover why the estimates differ.&lt;/p&gt;

&lt;p&gt;Once the assumptions are understood, the team can estimate again with better information.&lt;/p&gt;

&lt;h2&gt;
  
  
  Five things to establish before estimating
&lt;/h2&gt;

&lt;p&gt;I've found it useful to treat estimation as the final part of understanding a requirement, rather than the first.&lt;/p&gt;

&lt;p&gt;Before asking for story points, establish a few basics.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. What outcome are we trying to achieve?
&lt;/h3&gt;

&lt;p&gt;A ticket should describe the intended behaviour, not merely prescribe an implementation.&lt;/p&gt;

&lt;p&gt;"Add three checkboxes" tells developers what to build.&lt;/p&gt;

&lt;p&gt;"Allow users to control which notifications they receive" explains the outcome.&lt;/p&gt;

&lt;p&gt;That distinction leaves room for better technical decisions.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. What's included and excluded?
&lt;/h3&gt;

&lt;p&gt;Does the notification preferences change affect emails, in-app notifications or both?&lt;/p&gt;

&lt;p&gt;Does it apply to existing users?&lt;/p&gt;

&lt;p&gt;Are default settings changing?&lt;/p&gt;

&lt;p&gt;Clear boundaries prevent developers from inventing their own scope.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. What dependencies exist?
&lt;/h3&gt;

&lt;p&gt;Does the work require another team, an API change, a database migration or a third-party service?&lt;/p&gt;

&lt;p&gt;Dependencies can introduce uncertainty even when the implementation itself is simple.&lt;/p&gt;

&lt;h3&gt;
  
  
  4. What are the acceptance criteria?
&lt;/h3&gt;

&lt;p&gt;The team should understand what successful delivery looks like.&lt;/p&gt;

&lt;p&gt;For example:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Users can update their notification preferences.&lt;/li&gt;
&lt;li&gt;Changes persist after refreshing the page.&lt;/li&gt;
&lt;li&gt;Existing preferences remain unchanged unless explicitly updated.&lt;/li&gt;
&lt;li&gt;Failed requests display useful feedback.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;These criteria give everyone a common reference point.&lt;/p&gt;

&lt;h3&gt;
  
  
  5. What remains unknown?
&lt;/h3&gt;

&lt;p&gt;Not every question needs answering before estimation.&lt;/p&gt;

&lt;p&gt;But unresolved questions should be visible.&lt;/p&gt;

&lt;p&gt;Sometimes the right outcome of refinement is a short technical investigation rather than an increasingly speculative estimate.&lt;/p&gt;

&lt;p&gt;That's still progress.&lt;/p&gt;

&lt;h2&gt;
  
  
  Don't turn estimation into a negotiation
&lt;/h2&gt;

&lt;p&gt;One particularly unhelpful pattern is treating estimates as targets.&lt;/p&gt;

&lt;p&gt;Someone estimates thirteen points.&lt;/p&gt;

&lt;p&gt;The team discusses it.&lt;/p&gt;

&lt;p&gt;A manager asks whether it could be eight.&lt;/p&gt;

&lt;p&gt;After enough discussion, eight somehow becomes the agreed number.&lt;/p&gt;

&lt;p&gt;But unless the scope, implementation or uncertainty has changed, the work hasn't become smaller.&lt;/p&gt;

&lt;p&gt;Only the number has.&lt;/p&gt;

&lt;p&gt;A useful estimation session should produce shared understanding and expose risks. It shouldn't manufacture confidence.&lt;/p&gt;

&lt;p&gt;And sometimes the team should simply acknowledge that a requirement isn't ready to estimate.&lt;/p&gt;

&lt;h2&gt;
  
  
  Better context leads to better conversations
&lt;/h2&gt;

&lt;p&gt;This is one of the reasons I built &lt;a href="https://github.com/marketplace/pointwise-estimation" rel="noopener noreferrer"&gt;Pointwise&lt;/a&gt;, a collaborative Planning Poker application integrated with GitHub.&lt;/p&gt;

&lt;p&gt;The idea is straightforward: keep the GitHub issue visible while contributors estimate independently, reveal their scores together and discuss differences.&lt;/p&gt;

&lt;p&gt;There's also an optional GitHub Copilot review to help examine requirements from different perspectives.&lt;/p&gt;

&lt;p&gt;AI can help identify questions worth asking, but it shouldn't replace the team's technical judgement.&lt;/p&gt;

&lt;p&gt;The important part remains the conversation between the people responsible for delivering the work.&lt;/p&gt;

&lt;h2&gt;
  
  
  See Pointwise in action
&lt;/h2&gt;

&lt;p&gt;Here's a short walkthrough showing how Pointwise brings GitHub issues, independent estimates, simultaneous reveals and optional Copilot reviews into one workflow.&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/AyxBZgLLidA" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;p&gt;&lt;a href="https://youtu.be/AyxBZgLLidA?si=lAzb_COrcrbwNb_S" rel="noopener noreferrer"&gt;Watch the Pointwise walkthrough on YouTube&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Final thoughts
&lt;/h2&gt;

&lt;p&gt;Good estimates don't start with choosing a number.&lt;/p&gt;

&lt;p&gt;They start with understanding the problem.&lt;/p&gt;

&lt;p&gt;If a team disagrees on story points, that disagreement might be telling you something valuable about missing context, hidden dependencies or competing assumptions.&lt;/p&gt;

&lt;p&gt;Don't rush to eliminate the difference.&lt;/p&gt;

&lt;p&gt;Explore it.&lt;/p&gt;

&lt;p&gt;Because sometimes the most useful outcome of Planning Poker isn't the estimate you agree on.&lt;/p&gt;

&lt;p&gt;It's the problem you discover before writing a single line of code.&lt;/p&gt;




&lt;p&gt;&lt;strong&gt;How does your team handle requirements that aren't ready to estimate? Do you investigate them first, or estimate with uncertainty included?&lt;/strong&gt;&lt;/p&gt;

</description>
      <category>agile</category>
      <category>programming</category>
      <category>productivity</category>
      <category>planningpoker</category>
    </item>
    <item>
      <title>The Problem With Story Point Estimation Isn't Always the Numbers</title>
      <dc:creator>Ben Segni</dc:creator>
      <pubDate>Mon, 26 Oct 2026 22:19:28 +0000</pubDate>
      <link>https://dev.to/ben_segni_4c5eacf9af147ab/the-problem-with-story-point-estimation-isnt-always-the-numbers-2oe9</link>
      <guid>https://dev.to/ben_segni_4c5eacf9af147ab/the-problem-with-story-point-estimation-isnt-always-the-numbers-2oe9</guid>
      <description>&lt;p&gt;&lt;em&gt;Why I built &lt;a href="https://pointwise-estimates.tech" rel="noopener noreferrer"&gt;Pointwise&lt;/a&gt; to help development teams make better-informed decisions during estimation.&lt;/em&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  We've all been in these estimation meetings
&lt;/h2&gt;

&lt;p&gt;Someone presents a user story or GitHub issue.&lt;/p&gt;

&lt;p&gt;There's a brief explanation of what needs to be delivered, a few questions, and then everyone is asked to estimate the work.&lt;/p&gt;

&lt;p&gt;One developer thinks it's a three. Another thinks it's an eight.&lt;/p&gt;

&lt;p&gt;A senior developer confidently explains why it should be a five.&lt;/p&gt;

&lt;p&gt;After a bit of discussion, the team agrees on a number. Funnily enough, it's a five.&lt;/p&gt;

&lt;p&gt;And everyone moves on.&lt;/p&gt;

&lt;p&gt;But did we actually improve our understanding of the work, or did we just reach an agreement?&lt;/p&gt;

&lt;p&gt;I've been involved in plenty of estimation sessions over the years, and I've started to question whether the way we approach them always produces the most useful outcomes.&lt;/p&gt;

&lt;p&gt;Not because planning poker is inherently flawed, but because the conversations surrounding it can be.&lt;/p&gt;

&lt;h2&gt;
  
  
  The influence of a strong opinion
&lt;/h2&gt;

&lt;p&gt;Experience is incredibly valuable in &lt;a href="https://pointwise-estimates.tech" rel="noopener noreferrer"&gt;software development&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Senior engineers often recognise implementation patterns, understand architectural constraints and can identify potential problems before anyone else.&lt;/p&gt;

&lt;p&gt;That's exactly the kind of knowledge we want in an estimation session.&lt;/p&gt;

&lt;p&gt;But there's a subtle problem.&lt;/p&gt;

&lt;p&gt;When someone with considerable experience expresses a strong opinion, that opinion can influence the rest of the team, even when other participants have legitimate concerns.&lt;/p&gt;

&lt;p&gt;A less experienced developer might think an issue is more complicated but struggle to articulate why.&lt;/p&gt;

&lt;p&gt;Someone else might have encountered a similar problem before and know about a dependency that hasn't been discussed.&lt;/p&gt;

&lt;p&gt;Yet the conversation can gravitate towards the most confident explanation.&lt;/p&gt;

&lt;p&gt;This isn't necessarily intentional. It's a natural consequence of people having different levels of experience, confidence and familiarity with the work.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The challenge is making sure we're evaluating the reasoning behind an estimate, rather than the authority of the person giving it.&lt;/strong&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Not everyone has the same context
&lt;/h2&gt;

&lt;p&gt;Another problem I've noticed is how unevenly knowledge is distributed across a development team.&lt;/p&gt;

&lt;p&gt;When an issue is introduced, some people already understand it.&lt;/p&gt;

&lt;p&gt;Perhaps they helped define the requirements, worked on a related feature or know the area of the codebase particularly well.&lt;/p&gt;

&lt;p&gt;Others are seeing it properly for the first time.&lt;/p&gt;

&lt;p&gt;They might understand the general objective, but not the implementation details, existing dependencies or potential complications.&lt;/p&gt;

&lt;p&gt;And that's perfectly normal.&lt;/p&gt;

&lt;p&gt;We can't expect everyone to have the same knowledge of every part of a product.&lt;/p&gt;

&lt;p&gt;The difficulty comes when we ask everyone to estimate the work as though they share the same understanding.&lt;/p&gt;

&lt;p&gt;Some participants are estimating based on detailed knowledge.&lt;/p&gt;

&lt;p&gt;Others are estimating based on a high-level description and a handful of assumptions.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;We're asking everyone the same question, but we're not necessarily giving them the same information.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;That difference matters.&lt;/p&gt;

&lt;h2&gt;
  
  
  The constant search for context
&lt;/h2&gt;

&lt;p&gt;Then there's the practical reality of an estimation meeting.&lt;/p&gt;

&lt;p&gt;Someone raises a question about the requirements.&lt;/p&gt;

&lt;p&gt;We switch back to GitHub to read the issue.&lt;/p&gt;

&lt;p&gt;Someone mentions a related pull request, so we open another tab.&lt;/p&gt;

&lt;p&gt;Perhaps there's a dependency on another issue, or a previous implementation that might help explain the complexity.&lt;/p&gt;

&lt;p&gt;We investigate, discuss what we've found, and eventually return to the planning poker application.&lt;/p&gt;

&lt;p&gt;Then someone asks another question.&lt;/p&gt;

&lt;p&gt;And off we go again.&lt;/p&gt;

&lt;p&gt;The problem isn't that we're looking for information. That's exactly what we should be doing.&lt;/p&gt;

&lt;p&gt;It's that the information is scattered across different places, and the estimation tool itself often contributes very little beyond allowing everyone to select a number.&lt;/p&gt;

&lt;p&gt;The team is doing the hard work of understanding the issue somewhere else.&lt;/p&gt;

&lt;h2&gt;
  
  
  We don't have unlimited time to investigate
&lt;/h2&gt;

&lt;p&gt;Ideally, we'd thoroughly investigate every issue before estimating it.&lt;/p&gt;

&lt;p&gt;We'd examine the requirements, review the architecture, identify dependencies, consider testing scenarios and resolve every ambiguity.&lt;/p&gt;

&lt;p&gt;But that's rarely practical.&lt;/p&gt;

&lt;p&gt;Development teams have limited time for refinement and estimation.&lt;/p&gt;

&lt;p&gt;There might be ten issues to discuss, several developers involved and other work waiting to be completed.&lt;/p&gt;

&lt;p&gt;We need enough understanding to make a reasonable assessment, but we can't turn every estimation session into a detailed technical investigation.&lt;/p&gt;

&lt;p&gt;So we make assumptions.&lt;/p&gt;

&lt;p&gt;We rely on previous experience.&lt;/p&gt;

&lt;p&gt;We use the information available and try to move the conversation forward.&lt;/p&gt;

&lt;p&gt;Sometimes that works well.&lt;/p&gt;

&lt;p&gt;Other times, important considerations only become apparent once development has started.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The question I became interested in was whether we could help teams uncover more useful information without significantly increasing the time spent estimating.&lt;/strong&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Why I built Pointwise
&lt;/h2&gt;

&lt;p&gt;These frustrations led me to build &lt;strong&gt;&lt;a href="https://pointwise-estimates.tech" rel="noopener noreferrer"&gt;Pointwise&lt;/a&gt;&lt;/strong&gt;, a collaborative planning poker application integrated with GitHub Issues and GitHub Copilot.&lt;/p&gt;

&lt;p&gt;I wanted to address four related problems:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Opinion bias:&lt;/strong&gt; Help participants form independent estimates before being influenced by others.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Uneven understanding:&lt;/strong&gt; Give the team additional perspectives on requirements, complexity and risks.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Fragmented context:&lt;/strong&gt; Keep GitHub Issues and estimation within the same workflow.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Limited time:&lt;/strong&gt; Make it easier to identify questions worth discussing without requiring a full technical investigation.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;The aim isn't to remove uncertainty from software estimation. That would be a fairly ambitious claim.&lt;/p&gt;

&lt;p&gt;It's to help teams make better use of the time they already spend discussing work.&lt;/p&gt;

&lt;h2&gt;
  
  
  Independent thinking before consensus
&lt;/h2&gt;

&lt;p&gt;Pointwise uses hidden estimation to help reduce the influence of early opinions.&lt;/p&gt;

&lt;p&gt;Participants select their scores independently, and the estimates remain hidden until the session leader reveals them.&lt;/p&gt;

&lt;p&gt;That means the senior engineer's three doesn't immediately become everyone else's three.&lt;/p&gt;

&lt;p&gt;When the scores are revealed, differences become the starting point for discussion.&lt;/p&gt;

&lt;p&gt;If one participant selects a three and another selects an eight, the team can explore why.&lt;/p&gt;

&lt;p&gt;Perhaps someone knows about an existing implementation that makes the work straightforward.&lt;/p&gt;

&lt;p&gt;Perhaps someone else has identified a dependency or testing challenge.&lt;/p&gt;

&lt;p&gt;Both perspectives deserve consideration.&lt;/p&gt;

&lt;p&gt;The objective isn't to eliminate disagreement.&lt;/p&gt;

&lt;p&gt;It's to make disagreement useful.&lt;/p&gt;

&lt;h2&gt;
  
  
  Giving the team more to work with
&lt;/h2&gt;

&lt;p&gt;This is where GitHub Copilot comes into the picture.&lt;/p&gt;

&lt;p&gt;Pointwise integrates with the GitHub Copilot SDK to review issues from different professional perspectives:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Software Engineer&lt;/li&gt;
&lt;li&gt;UX Designer&lt;/li&gt;
&lt;li&gt;Product Owner&lt;/li&gt;
&lt;li&gt;QA Engineer&lt;/li&gt;
&lt;li&gt;Business Analyst&lt;/li&gt;
&lt;li&gt;Security Engineer&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Each perspective examines the issue with different considerations in mind.&lt;/p&gt;

&lt;p&gt;For example, a QA review might identify missing acceptance criteria, edge cases or testing requirements.&lt;/p&gt;

&lt;p&gt;A Security Engineer review might raise questions about permissions, authentication or data handling.&lt;/p&gt;

&lt;p&gt;A Product Owner review might highlight unclear requirements or assumptions about expected behaviour.&lt;/p&gt;

&lt;p&gt;These reviews can give the team additional questions and considerations to discuss, particularly when some participants are less familiar with the issue.&lt;/p&gt;

&lt;p&gt;The intention isn't for AI to tell everyone what the estimate should be.&lt;/p&gt;

&lt;p&gt;And it certainly isn't to suggest that AI-generated feedback is automatically correct.&lt;/p&gt;

&lt;p&gt;Copilot can misunderstand requirements, overlook important context or raise concerns that aren't relevant.&lt;/p&gt;

&lt;p&gt;Its observations still need to be assessed by the people doing the work.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;The value is in providing additional perspectives that help the team ask better questions, rather than treating AI as another authoritative voice in the room.&lt;/strong&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Keeping everything closer to the issue
&lt;/h2&gt;

&lt;p&gt;I also wanted to reduce the amount of switching between applications.&lt;/p&gt;

&lt;p&gt;Pointwise connects directly to GitHub, allowing teams to find and read issues during an estimation session.&lt;/p&gt;

&lt;p&gt;Participants can review the issue, submit their estimates, reveal scores and request Copilot reviews within the same experience.&lt;/p&gt;

&lt;p&gt;That doesn't mean every piece of information is suddenly available in one place.&lt;/p&gt;

&lt;p&gt;Some questions will always require further investigation.&lt;/p&gt;

&lt;p&gt;But it removes some of the friction involved in moving between GitHub and a separate estimation tool, while keeping the issue itself central to the discussion.&lt;/p&gt;

&lt;h2&gt;
  
  
  Making better use of limited time
&lt;/h2&gt;

&lt;p&gt;For me, this is probably the most important part of the idea.&lt;/p&gt;

&lt;p&gt;I don't expect Pointwise to replace proper refinement, technical investigation or conversations between developers.&lt;/p&gt;

&lt;p&gt;And I don't think AI can reliably turn an incomplete issue into a fully understood piece of work.&lt;/p&gt;

&lt;p&gt;What I'm interested in is whether it can help teams make better use of a limited estimation window.&lt;/p&gt;

&lt;p&gt;If a review highlights a missing acceptance criterion, the team can discuss it.&lt;/p&gt;

&lt;p&gt;If it raises a potential dependency, someone can confirm whether that dependency exists.&lt;/p&gt;

&lt;p&gt;If it identifies uncertainty that can't be resolved during the session, the team can acknowledge that uncertainty rather than burying it inside a story point estimate.&lt;/p&gt;

&lt;p&gt;Sometimes the most valuable outcome of an estimation session isn't agreeing that something is a five.&lt;/p&gt;

&lt;p&gt;It's discovering that we don't yet know enough to estimate it confidently.&lt;/p&gt;

&lt;h2&gt;
  
  
  Building the application
&lt;/h2&gt;

&lt;p&gt;From an engineering perspective, Pointwise has also been an interesting project to develop.&lt;/p&gt;

&lt;p&gt;The frontend uses React and TypeScript, with backend APIs deployed on Vercel.&lt;/p&gt;

&lt;p&gt;Supabase provides persistent storage, while Upstash Redis manages collaborative estimation session state.&lt;/p&gt;

&lt;p&gt;I use Redis Lua scripts to handle important state transitions atomically, helping prevent simultaneous participant actions from creating inconsistent session states.&lt;/p&gt;

&lt;p&gt;GitHub OAuth handles user authentication, while the GitHub App manages repository access.&lt;/p&gt;

&lt;p&gt;The application also integrates with the GitHub Copilot SDK for issue reviews and Stripe for subscription management.&lt;/p&gt;

&lt;p&gt;One of the more challenging architectural aspects was keeping authentication, repository permissions and subscription entitlements separate while ensuring they worked together consistently.&lt;/p&gt;

&lt;p&gt;It's been a valuable exercise in building a complete SaaS product, but the technical decisions have ultimately been driven by the experience I wanted to create.&lt;/p&gt;

&lt;h2&gt;
  
  
  Better estimation starts with better understanding
&lt;/h2&gt;

&lt;p&gt;There are already plenty of planning poker applications.&lt;/p&gt;

&lt;p&gt;If all Pointwise did was let developers select cards and reveal numbers, I don't think it would be particularly interesting.&lt;/p&gt;

&lt;p&gt;What I wanted to explore was whether we could improve the quality of the conversation around those numbers.&lt;/p&gt;

&lt;p&gt;Reduce the influence of strong opinions.&lt;/p&gt;

&lt;p&gt;Give people with less context a better starting point.&lt;/p&gt;

&lt;p&gt;Make relevant information easier to access.&lt;/p&gt;

&lt;p&gt;And use AI to surface questions that might otherwise be overlooked, without expecting it to make decisions for the team.&lt;/p&gt;

&lt;p&gt;I don't have evidence yet that Pointwise produces more accurate estimates. That's something that would need to be evaluated through real team usage.&lt;/p&gt;

&lt;p&gt;But I believe improving shared understanding is a worthwhile objective in its own right.&lt;/p&gt;

&lt;p&gt;Because ultimately, the value of estimation isn't just the number we agree on.&lt;/p&gt;

&lt;p&gt;It's what we learn about the work before we start building it.&lt;/p&gt;

&lt;h2&gt;
  
  
  I'd be interested in your experience
&lt;/h2&gt;

&lt;p&gt;Pointwise is now available through the GitHub Marketplace, with a free plan for teams that want to try it.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;&lt;a href="https://pointwise-estimates.tech" rel="noopener noreferrer"&gt;Explore Pointwise Estimates&lt;/a&gt;&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;I'm continuing to develop the application and would welcome feedback from teams that manage their work through GitHub Issues.&lt;/p&gt;

&lt;p&gt;I'm particularly interested in how other developers experience estimation sessions.&lt;/p&gt;

&lt;p&gt;Does everyone genuinely understand the work before voting? Do senior opinions sometimes influence the outcome? And how often do you find yourselves switching between applications just to gather enough context?&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Are we using estimation to understand the work, or are we sometimes just trying to agree on a number?&lt;/strong&gt;&lt;/p&gt;

</description>
      <category>agile</category>
      <category>github</category>
      <category>productivity</category>
      <category>webdev</category>
    </item>
  </channel>
</rss>
