Feedback
Feedback is information or evaluation about an action, system, or outcome that can inform assessment or subsequent decisions and behavior.
Lead Developer Career Guide
Information about performance, contributions, or progress that is communicated to reinforce strengths or guide improvement.
- How it is used
- Guides individual growth, team performance, documentation revisions, and decisions about development processes and products.
- How deep it goes
- Develops practical tactics for giving, receiving, and planning feedback conversations, including specificity, setting, listening, and cultural differences.
- Examples
- Uses missed deadlines, specific praise, presentation criticism, and documentation-reader questions to make feedback concrete.
- What you can do after
- Practice/perform specific feedback conversations, solicit perspectives from varied stakeholders, and use the responses to guide improvement.
Think Like a Software Engineering Manager
A feedback mechanism is a process or structure for collecting information about experiences, performance, or conditions.
- How it is used
- Uses feedback to support employee growth, manage performance, improve leadership, and respond to team concerns.
- How deep it goes
- Develops practical channels and conversation habits, from check-ins and surveys to data-backed reviews and follow-up, without prescribing one universal script.
- Examples
- Uses a working-hours policy, a missed project milestone, and periodic employee surveys to show different feedback loops.
- What you can do after
- Practice/perform two-way feedback by gathering specific input, discussing it constructively, acting on it, and checking progress.
Read how Think Like a Software Engineering Manager explains feedback →
"Looks Good to Me"
Feedback is information or opinions shared to help another person or team evaluate and improve work.
- How it is used
- Improves code reviews by making concerns actionable while supporting trust, shared learning, and maintainable code.
- How deep it goes
- Explains comment quality, team expectations, feedback culture, real-time collaboration, and AI assistance without reducing judgment to automation.
- Examples
- Uses vague and constructive PR comments, inconsistent reviewer detail, pairing sessions, mob sessions, and AI review tools as examples.
- What you can do after
- Apply it via code review by writing focused comments, agreeing on response expectations, and choosing human or automated feedback deliberately.
The Art of AI Product Development
User feedback is information provided by users about their needs, experiences, problems, preferences, or reactions to a product.
- How it is used
- Uses feedback to discover needs and guide ongoing improvements to AI product experience and performance.
- How deep it goes
- Distinguishes implicit and explicit signals and outlines collection, incentives, review, and communication in a continuous feedback loop.
- Examples
- Illustrates stated versus observed language-learning habits and a demand-forecasting tool that collects reasons and comments.
- What you can do after
- Apply feedback collection and interpretation to identify product friction and inform model improvements.
Read how The Art of AI Product Development explains feedback →
Software Architecture Decision-Making
Feedback is information from results, operation, stakeholders, or changing conditions that informs subsequent decisions and adaptations.
- How it is used
- Tests architectural decisions against delivery, operational, organizational, and changing-context evidence.
- How deep it goes
- Explains feedback as an architectural decision loop, with revisit triggers, early validation, incident learning, and collaborative conditions.
- Examples
- Uses an uncertain greenfield decision log, early delivery slices, incident reviews, and concerns about painful boundaries or permanent workarounds.
- What you can do after
- Evaluate/understand when architectural assumptions need revisiting and preserve options until stronger evidence appears.
Read how Software Architecture Decision-Making explains feedback →
Design for Developers
Feedback is information communicated to users about the effects or status of their interactions with a product.
- How it is used
- Uses feedback both to communicate interaction status to users and to guide design improvements from their responses.
- How deep it goes
- Works through interface cues, loading states, interviews, and prioritization without specifying animation code or an interview-analysis tool.
- Examples
- Illustrates feedback with a shaking login form, loading and skeleton states, and interviews about visuals and user flows.
- What you can do after
- Practice/perform interaction-state checks and user interviews, then prioritize recurring problems that hinder use.
Enterprise RAG
User feedback is information provided by users about their experience, observations, problems, expectations, or assessment of a system.
- How it is used
- Drives iterative improvement of RAG chatbot answers through user ratings, comments, diagnosis, and follow-up.
- How deep it goes
- Elaborates capture, serverless storage, log-based retrieval-versus-prompt diagnosis, fixes, verification, and user notification.
- Examples
- Uses thumbs ratings and comments, a failed product query, log searches by email, and a thank-you-it’s-fixed message.
- What you can do after
- Operate/configure a feedback loop that captures complaints, diagnoses answer failures, validates fixes, and closes the loop with users.
Learn Go with Pocket-Sized Projects
Feedback is the book-specific slice-based representation of one hint for each character position in a word guess.
- How it is used
- Guides word-game players by reporting the status of each character in a guess.
- How deep it goes
- Defines three statuses and implements duplicate-aware feedback computation with tests.
- Examples
- Uses SMALL versus HELLO and repeated-character test cases to examine accurate hints.
- What you can do after
- Build feedback generation that distinguishes exact matches from unused misplaced characters.
Read how Learn Go with Pocket-Sized Projects explains feedback →
Everyday Data Visualization
Feedback is information or evaluation provided about a product or process to guide its improvement.
- How it is used
- Confirms chart interactions for users and guides revisions to requirements, designs, and prototypes.
- How deep it goes
- Explains visible confirmation and the plan–build–feedback cycle, with guidance on timing, placement, unfinished work, and distracting motion.
- Examples
- Uses changing chart labels and animated bars for interaction feedback, plus stakeholder review of requirements and design artifacts.
- What you can do after
- Practice/perform timely interaction confirmation and seek stakeholder or user input before a visualization is finished.
Shipping Go: Develop, deliver, discuss, design, and go again
Feedback is information about a product or process obtained from evaluation, use, or other responses that can guide subsequent changes.
- How it is used
- Guides the next small product change after customers or teammates respond to a delivered increment.
- How deep it goes
- Explains feedback through delivery cycles and code-review habits, without specifying a collection or measurement protocol.
- Examples
- Uses a one-day hello API demo, README release milestones, customer responses, and code-review conversations.
- What you can do after
- Practice/perform short release-and-response cycles, then use customer and review feedback to choose improvements.
Read how Shipping Go: Develop, deliver, discuss, design, and go again explains feedback →
BDD in Action, Second Edition
Feedback is information about observed results or understanding that guides subsequent decisions, actions, or improvements.
- How it is used
- Guides requirements discovery, feature prioritization, stakeholder collaboration, and assessment of automated acceptance-test health.
- How deep it goes
- Explains feedback through early demonstrations, conversations, examples, and reliable acceptance tests, without specifying one universal feedback method.
- Examples
- Uses early feature demonstrations, stakeholder conversations, password-strength examples, and automated acceptance-test failures.
- What you can do after
- Apply feedback to revise requirements, prioritize valuable features, refine acceptance criteria, and investigate unreliable automated tests.
Delivering Machine Learning Projects: From design to deployment
Feedback is information, opinions, or evaluations provided about work, behavior, or outcomes to support assessment and improvement.
- How it is used
- Uses feedback to expose production problems and guide post-project improvement.
- How deep it goes
- Explains collection channels, participation methods, review forms, presentation, and investigation, but stops short of formal feedback analytics.
- Examples
- Uses confidential user complaints about login timeouts and a rated team post-project review form.
- What you can do after
- Operate feedback channels and review practices to diagnose production defects and select improvements.
Read how Delivering Machine Learning Projects: From design to deployment explains feedback →
AI for Everyday IT
Feedback is information or evaluation about a person's performance, behavior, or work intended to support improvement.
- How it is used
- Uses feedback to support employee development, constructive management conversations, and communication practice.
- How deep it goes
- Provides prompt templates and follow-up practices for constructive feedback, plus a repeatable interview exercise.
- Examples
- Uses missed deadlines, project successes, presentation recordings, and mock job interviews as feedback scenarios.
- What you can do after
- Apply it via an AI assistant to draft constructive messages, plan follow-up, and practice interview answers.
Interview Speak
Information about a person's work or behavior that can guide improvement.
- How it is used
- Uses feedback to assess whether someone accepts criticism, learns from it, and changes behavior.
- How deep it goes
- Explains selection, neutral description, nondefensive reaction, behavioral change, and present-day application.
- Examples
- Uses a perfectionism criticism, reflective answer revision, and a manager acting on team requests.
- What you can do after
- Apply feedback by describing criticism clearly, responding openly, and demonstrating specific changes in future behavior.
Coding with AI
User feedback is information, opinions, or evaluations provided by users about a product, service, or interaction.
- How it is used
- Uses acceptance, review, and ratings to evaluate AI-generated coding suggestions and improve future assistance.
- How deep it goes
- Explains feedback through an evaluation loop and demonstrates review, thumbs-up, and thumbs-down controls.
- Examples
- Uses musician applause or tomatoes, accepted Copilot edits, and a Cursor result modifying river_run.py.
- What you can do after
- Apply feedback by reviewing proposed changes and rating AI coding suggestions before retaining or copying them.
Think Like a CTO
Information about a person's behavior or performance provided to support understanding or improvement.
- How it is used
- Uses feedback to improve executive performance, align technology with business needs, and guide software release and product decisions.
- How deep it goes
- Explains direct questions, team pulse checks, and managed user input, but stops short of defining a formal measurement framework.
- Examples
- Uses client and user input, CEO conversations, team pulse checks, and The Office’s out-of-touch boss as teaching examples.
- What you can do after
- Practice/perform it by requesting candid reviews, setting improvement goals, checking team needs, and managing external input.
Human-in-the-Loop Machine Learning: Active learning and annotation for human-centered AI
A response from an interface that indicates whether a user's action was recorded or had an effect.
- How it is used
- Confirms whether an annotation action was recorded, helping users tell a working control from a misleading one.
- How deep it goes
- Defines feedback through interface responses and relates it to affordance, without detailing implementation.
- Examples
- Uses a click followed by animation or a message, contrasted with a button that does nothing.
- What you can do after
- Evaluate annotation-interface controls by checking whether actions produce a clear indication that they were recorded.
AI-Powered Search
User feedback is information supplied directly or indirectly by a user about the quality, usefulness, or correctness of a system response.
- How it is used
- Improves personalized search and recommendations by adapting ranking or relevance to explicit judgments, behavior, and live interaction.
- How deep it goes
- Explains explicit and implicit collection, profile updates, feedback loops, and safeguards, but stops short of implementation algorithms.
- Examples
- Uses thumbs-up or thumbs-down judgments, clicks, purchases, newer interactions, and live conversational response refinement.
- What you can do after
- Apply feedback through explicit controls or behavioral signals to update personalization while limiting errors and respecting user choice.
LLMs in Production
Human feedback is an evaluation, preference, or correction supplied by people about a system's behavior or output.
- How it is used
- Uses human preferences to steer language-model outputs toward acceptable behavior.
- How deep it goes
- Explains response ranking, reward modeling, PPO, costs, privacy, bias, and degradation risks without detailing implementation.
- Examples
- Uses preferred answers to prompts, backflip attempts, and changing prime-number performance as examples.
- What you can do after
- Evaluate generated responses with human preferences and recognize the operational risks of using that feedback for updates.
Effective Conversational AI
User feedback is information provided by users about their experiences, needs, or evaluations of a software system.
- How it is used
- Use direct reports to uncover chatbot pain points and measure whether answers or experiences meet user needs.
- How deep it goes
- Collect and categorize survey responses, pair them with transcripts, and scope ratings, while cautioning against acting on isolated reports.
- Examples
- Uses survey reports about vaccine eligibility and appointment scheduling, plus in-product thumbs-up or thumbs-down ratings.
- What you can do after
- Evaluate chatbot understanding by gathering purpose-specific ratings and investigating recurring user-reported problems.
Build an AI Agent (From Scratch)
End-user feedback is information about system quality or satisfaction provided by people who interact with the deployed system.
- How it is used
- Uses information from deployed-system users as one input for evaluating agent quality and satisfaction.
- How deep it goes
- Defines feedback as an evaluator category but does not specify collection procedures, metrics, or implementation details.
- What you can do after
- Evaluate deployed-system quality or satisfaction using information supplied by interacting users.
Read how Build an AI Agent (From Scratch) explains feedback →
Build High Value Software Systems
Information about the effects or results of work that can guide subsequent decisions or changes.
- How it is used
- Uses feedback to guide changes, improve learning, and keep engineering work moving through validation.
- How deep it goes
- Explains feedback through CI/CD delays and PR coaching, without presenting a general feedback framework.
- Examples
- Uses a 12-to-42-minute pipeline and a PR comment about API failures and error boundaries.
- What you can do after
- Apply feedback by reducing validation waits and writing review comments that explain failure modes and reusable solutions.
Read how Build High Value Software Systems explains feedback →
Effective Platform Engineering
Customer feedback is information provided by customers about their needs, experiences, problems, or responses to a product or service.
- How it is used
- Guides platform prioritization, continuous improvement, roadmap decisions, and validation of developer experience.
- How deep it goes
- Explains capture, triage, behavioral comparison, roadmap prioritization, incremental delivery, and measurement without detailing collection instruments.
- Examples
- Uses request queues, Feature Request boards, observed-versus-stated usage, roadmaps, feedback loops, and developer satisfaction measures.
- What you can do after
- Apply feedback through prioritized roadmaps, validated increments, and measurable platform improvements.
Building LLM Applications with DSPy
Text returned by an evaluation metric that explains why a response received its score.
- How it is used
- Uses explanatory metric feedback to help reflective optimizers learn patterns and improve prompts.
- How deep it goes
- Defines feedback, contrasts it with scores, and illustrates its use in an optimizer metric.
- Examples
- Illustrates feedback with uppercase, 10-word output rules and a DSPy metric returning overlap scores and explanations.
- What you can do after
- Apply it via DSPy metric functions to return explanatory feedback alongside a score for SIMBA.
Read how Building LLM Applications with DSPy explains feedback →
Machines that Think
Feedback is information about the effects or outputs of a system that can influence subsequent behavior or processing.
- How it is used
- Frames feedback as a constraint on learning and as part of human–machine interaction and AI-supported education.
- How deep it goes
- Explains sparse, delayed, and noisy feedback and names resulting learning approaches, but stops short of implementation detail.
- Examples
- Contrasts sparse, delayed, noisy biological feedback with AI assistants’ educational feedback, dialogue, and practice.
- What you can do after
- Evaluate how feedback can shape learning and interaction without treating responsive output as proof of understanding.
Also covered in
- Writing for Developers: Blogs that get read
- Bug BashIn early access (MEAP)
- Understanding AIIn early access (MEAP)
- Bootstrapping Microservices, Second Edition: With Docker, Kubernetes, GitHub Actions, and Terraform
- Tech Startup Toolkit
- Deep Learning with R, Third Edition
- How Large Language Models Work
- The Architecture of ReasoningIn early access (MEAP)
- Grokking AI Algorithms, Second Edition
- Starting Data Analytics with Generative AI and Python
- Knowledge Graphs and LLMs in Action
- Collaborative Software Design: How to facilitate domain modeling decisions
What to know first
- language modelA language model is a computational model that represents patterns in linguistic sequences to assign probabilities, predict, generate, or evaluate language.
- software testingSoftware testing is the systematic evaluation of software under specified inputs or conditions to assess its behavior against expected results or requirements and identify defects.
- promptA prompt is text or other input supplied to an AI model to provide instructions, context, or other information that guides its response or generated output.
- accuracyAccuracy is the degree to which a result agrees with a specified truth, reference, or criterion, commonly measured as the proportion of evaluated predictions that are correct.
Related concepts
- criticismincludes · Criticism is an evaluative judgment or communication that identifies perceived shortcomings, faults, or undesirable qualities in a subject, work, behavior, or result.
- AI feedbackincludes · AI feedback is evaluative or supervisory information generated by an artificial-intelligence system about another system’s outputs, actions, or behavior.
Questions about feedback
What is feedback?
Feedback is information or evaluation about an action, system, or outcome that can inform assessment or subsequent decisions and behavior.
Which Manning books explain feedback?
Lead Developer Career Guide, Think Like a Software Engineering Manager, "Looks Good to Me", The Art of AI Product Development, Software Architecture Decision-Making, Design for Developers, Enterprise RAG, Learn Go with Pocket-Sized Projects and 17 more.
What should I know before learning feedback?
Manning books that teach feedback usually introduce language model, software testing, prompt and accuracy first.
What kinds of feedback are there?
Kinds of feedback include criticism, AI feedback, periodic feedback, rapid feedback, session-based feedback and singleton feedback.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.
























