Skip to main content

Technical PM Skills

This section covers the technical knowledge product managers need to work effectively with engineering teams. The goal is productive collaboration and informed decision-making, not deep technical expertise.

Required Technical Knowledge

Web Application Architecture

Web applications follow a standard request flow:

User → Client (browser/app) → API → Server → Database

ComponentDescriptionPM Relevance
ClientBrowser or mobile appChanges require app updates and store review
APIInterface between client and serverDetermines what data is available
ServerProcesses requestsChanges can be deployed immediately
DatabaseStores dataStructure affects query performance

Feature scoping considerations:

  • Client-only changes: Require app store approval (1-7 days)
  • Server-only changes: Can deploy instantly to all users
  • Full-stack changes: Require coordination across teams

API Fundamentals

API (Application Programming Interface) defines how software components communicate.

HTTP MethodPurposeExample
GETRead dataRetrieve user profile
POSTCreate dataSubmit new order
PUT/PATCHUpdate dataModify settings
DELETERemove dataDelete account
Status CodeMeaningTypical Cause
200SuccessRequest completed
400Client errorInvalid request format
401UnauthorizedAuthentication required
404Not foundResource does not exist
500Server errorInternal system failure

Database Types

TypeExamplesUse CaseTrade-offs
SQL (Relational)PostgreSQL, MySQLComplex queries, data integrityStrict schema, slower writes
NoSQL (Document)MongoDB, DynamoDBFlexible schema, high scaleLimited query complexity
NoSQL (Key-Value)Redis, MemcachedCaching, session storageSimple data structures only

Caching

Caching stores frequently-accessed data closer to the user to reduce latency.

Cache LocationLatencyUse Case
Browser cacheUnder 1msStatic assets (images, CSS)
CDN10-50msContent delivery
Application cache1-10msDatabase query results
Database cache5-20msQuery optimization

Cache invalidation consideration: Cached data can become stale. Profile updates may not appear immediately due to caching.

Synchronous vs Asynchronous Processing

TypeBehaviorExampleUser Experience
SynchronousWait for completionLogin authenticationLoading spinner
AsynchronousReturn immediatelyVideo upload"Processing" status

Asynchronous operations require:

  • Status tracking
  • Failure handling
  • User notification system

Working with Engineering Teams

Effective Communication Patterns

Instead ofUse
"How long will this take?""Can you help me understand the complexity for prioritization?"
"The competitor has this feature""Users need X. Competitor Y solved it this way. What approach works best for us?"
"We need this by Friday""Here's why Friday matters. Is that achievable? What would we need to cut?"

What Engineers Value from PMs

BehaviorDescription
Clear problem statementsDefine the problem, not the solution
ContextExplain why this matters, who is affected, what happens if delayed
Honesty about uncertaintyAcknowledge unknowns rather than false confidence
Respect for expertiseAllow engineers to propose technical solutions
Commitment follow-throughDeliver on promises, communicate changes immediately

Behaviors That Damage Relationships

BehaviorImpact
Mid-sprint requirement changesDisrupts planning, erodes trust
Ignoring technical concernsCreates technical debt, reduces credibility
Treating engineers as implementersReduces engagement and solution quality
Making assumptions about effortLeads to unrealistic commitments

Platform PM Responsibilities

Platform PMs build products for developers. The user base has different expectations and behaviors.

Developer Experience (DX) Metrics

MetricTargetWhy It Matters
Time to first API callUnder 30 minutesMeasures onboarding quality
Documentation completeness100% endpoint coverageReduces support burden
Error message clarityActionable in 95% of casesEnables self-service debugging
SDK qualityUnder 5 open critical bugsAffects developer productivity

API Design Considerations

FactorDescriptionImpact
Backward compatibilityChanges must not break existing integrationsRequires versioning strategy
VersioningSupport multiple API versions simultaneouslyMaintenance overhead
DeprecationAdvance notice before removing functionalityTypically 6-12 months
Migration supportTools and documentation for version upgradesReduces customer friction

Platform PM shipping velocity is typically slower than consumer PM due to backward compatibility requirements.

Technical Debt

Technical debt is accumulated shortcuts that increase future development cost.

Indicators of Technical Debt Problems

SignalDescription
Slow feature deliverySimple features take weeks instead of days
Recurring bugsSame issues appear repeatedly
Engineer frustrationTeam avoids certain codebase areas
System fragilityChanges in one area break others

PM Role in Technical Debt Management

ResponsibilityAction
Advocate for allocationReserve 15-20% of sprint capacity
Understand trade-offsKnow which shortcuts are appropriate
Balance prioritiesWeigh new features against stability
Communicate impactTranslate technical debt to business terms

Interview Questions and Responses

"How do you work with engineers?"

Response structure:

  1. Problem framing: Provide context and goals, not implementation details
  2. Collaboration: Let engineers propose approaches
  3. Constraint awareness: Understand technical limitations before committing
  4. Knowledge gaps: Ask questions rather than make assumptions

"How technical should a PM be?"

Response framework:

LevelDescription
RequiredUnderstand system architecture, technical vocabulary, read documentation
Not requiredWrite production code, deep language/framework expertise, make low-level technical decisions

"Describe a technical trade-off decision"

STAR format example:

  • Situation: Need for notifications system
  • Task: Choose build vs. buy approach
  • Action: Analyzed current volume (low), projected growth (18+ months to justify build), integration effort (2 weeks vs 3 months)
  • Result: Used third-party service, shipped faster, learned requirements, built custom system later with better specifications

Developing Technical Knowledge

MethodDescriptionBenefit
Engineer conversationsAsk about system design decisionsContext and relationships
Architecture documentationRead existing system docsFoundation knowledge
On-call shadowingObserve incident responseUnderstand failure modes
Personal projectsBuild simple applicationsExperience development constraints

Company-Specific Technical Expectations

CompanyPM Technical BarNotes
StripeHighRead code, review API designs, opinions on DX
GoogleSplit rolePM focuses on what/why; TPM handles execution
AmazonModerateStrategic tech understanding; SDMs handle implementation
MetaModerate-HighRead internal wikis, understand performance implications
StartupsVariableSmaller teams require broader knowledge

Platform PM Considerations

FactorConsumer PMPlatform PM
User toleranceModerateLow (developers switch quickly)
Documentation prioritySecondaryCore product component
Breaking changesFix and iterateMonths of deprecation notice
Success metricsStandard analyticsInferred from API patterns, surveys
Customer segmentsRelatively uniformIndividual developers vs. enterprises