Postgres Performance Kit

Find and fix your slow Postgres queries yourself — one payment, not $149/month monitoring

When Postgres gets slow, the market sells subscriptions: typical monitoring starts around $149 per server, per month at the time of writing, plan-review tools bill annual-only, and managed platforms point you up a compute ladder that costs more every year. This kit is the doing layer: a per-platform instrumentation runbook, an EXPLAIN ANALYZE playbook with a 12 bad-plan cheat sheet, index and autovacuum audit worksheets with read-only-first SQL, and a 100-point audit rubric with a client-ready report. One payment of $56 — no subscription.

Get the kit — $56
One-time payment · ≈ ฿1,990 · free updates for v1.x · instant download

What's inside the kit

One zip of plain Markdown, XLSX and SQL: instrumentation runbooks per platform, an EXPLAIN reading playbook, audit worksheets with ready-to-run read-only queries, decision trees and a scoreable 100-point rubric — everything re-runs on every database and every client. No software to install, nothing to subscribe to.

Instrument

Per-platform instrumentation runbook

Turn on pg_stat_statements, auto_explain and slow-query logging with exact steps for self-hosted, RDS parameter groups, Supabase and Neon — including what needs superuser and what a read-only role can do — plus read-only SQL that ranks your top-N slowest queries.

Triage

Decision tree + latency budget

Is it latency, throughput or connection errors? A symptom decision tree, a latency-budget worksheet, and the matrix that answers the hard question — fix the query, fix the schema, or rent bigger hardware — with the annual cost of each option in dollars.

Explain

EXPLAIN ANALYZE playbook

Read any plan from the bottom up: costs, rows, loops, buffers. The cheat sheet covers the 12 bad plans you'll actually meet — oversized sequential scans, nested-loop multiplication, disk sorts, hash spills — each with how to read it and the fix that usually applies.

Index

The index audit pack

Read-only SQL finds unused, duplicate and redundant indexes; a decision guide covers partial, covering and expression indexes; and a write-cost tradeoff worksheet prices every new index against the writes it will tax — so each one earns its keep.

Config

Autovacuum, memory & pooling

Autovacuum thresholds sized to your table churn, memory settings sized to your RAM, and a PgBouncer pooling SOP (3 modes) matched to the real connection limits of managed plans — plus a right-size sheet that compares upgrade rent against one-time fixes.

Report

100-point rubric + client report

A scoreable 100-point audit rubric across 7 categories, tied to a white-label report template you can hand to a client or your own team — agencies also get a per-engagement pricing guide. Single-buyer commercial license.

What this kit will never tell you

No speed guarantees. Nobody can promise "10x faster" — the right fix depends on your workload, and some slow queries are application problems, not database problems. What the kit does is give you the measurement process and the decision rules consultants use: baseline first, change what the evidence points to, re-measure after.

Read-only first. Staging always. Every diagnostic query in the kit reads catalog views — pg_stat_statements, pg_stat_user_indexes, pg_stat_activity — that are safe on production. Anything that changes your schema carries a written warning: test on staging first, run off-peak. No shortcuts that can destroy data.

A process, not a monitor. The kit doesn't watch your database around the clock — pair it with your platform's free dashboards or any monitor you already run. What it adds is the part none of them sell: the audit process, the decision rules and the report you can hand over.

How you'll use it

A one-week performance sprint — then the same loop on every database and every quarter.

Days 1–2 — baseline

Instrument and score

Follow the runbook for your platform, capture your baseline numbers, and score your database on the 100-point rubric. You now know where you stand — and what to chase first.

Days 3–5 — fix the top 3

Read plans, change little

Work your top-N query list through the EXPLAIN playbook, use the decision tree to choose index vs query vs config, and re-measure against the baseline. The report tells the story.

Every quarter — re-run

The rubric shows the delta

Re-score after each release cycle and after growth. Agencies: repeat the loop per client and deliver the white-label report as a paid audit.

One price. You own the process.

No monthly plan, no per-server fee, no annual-only billing — typical Postgres monitoring already charges enough of those.

$56 one-time payment

≈ ฿1,990

  • Per-platform instrumentation runbook + top-N slow-query SQL (read-only)
  • EXPLAIN ANALYZE playbook + 12 bad-plan cheat sheet, each with its fix
  • Index, autovacuum, memory and pooling audit worksheets + decision tree
  • 100-point audit rubric + white-label client report · free updates v1.x · 14-day refund

Secure checkout via Stripe · card payments · taxes calculated at checkout

Frequently asked questions

Is this a monitoring tool?

No — and on purpose. Typical Postgres monitoring starts around $149 per server per month at the time of writing, and the price multiplies with every database you add. This kit is the process layer they don't sell: how to instrument, read plans, audit indexes and config, and write up the result. It works alongside any monitor — or with your platform's free dashboards.

What exactly do I get?

A zip with plain files: 00-START-HERE.md (a one-week performance sprint), the triage pack (intake form, symptom decision tree, latency-budget worksheet, query-vs-schema-vs-hardware decision matrix), instrumentation runbooks per platform with top-N query SQL, the EXPLAIN ANALYZE playbook with the 12 bad-plan cheat sheet, the index audit pack, the config pack (autovacuum, memory, PgBouncer pooling), the capacity right-size sheet, the 100-point audit rubric with a white-label report, a recommended-alerts table, and a buyer README + license. Markdown, XLSX and SQL — no software to install.

Will my queries actually get faster?

We can't promise a number, and you should be skeptical of anyone who does. What the kit does: measure before and after, point the evidence at the most likely fix (an index, a rewritten query, a config change), and give you the decision rules that stop you from paying for hardware when a one-line fix exists. Slow queries that are application bugs are out of scope — the triage tree tells you so early.

Which platforms does it cover?

Self-hosted Postgres, Amazon RDS/Aurora (parameter groups), Supabase and Neon — each with its own runbook, because permissions and defaults genuinely differ: what needs superuser on self-hosted may be impossible with a managed read-only role, and connection limits differ per plan. The pooling SOP is written against those real limits.

Is the SQL safe to run on production?

The diagnostic queries read catalog views (pg_stat_statements, pg_stat_user_indexes, pg_stat_activity) and are designed read-only-first — that is the whole point of being able to run them in production. Statements that change your database (CREATE INDEX, config changes) appear only inside clearly marked warning boxes: test on staging, schedule off-peak.

Can I sell database audits with it?

Yes. The audit rubric ties into a white-label report template — remove our name, add yours — and an included pricing guide shows how agencies typically price a per-engagement audit. The single-buyer commercial license covers your own projects and your clients' work; reselling or sharing the kit files themselves isn't included.

What is the refund policy?

If the kit doesn't fit how you work, write to us within 14 days of purchase and we refund in full — the files are yours to keep or delete. One refund per purchase; we only ask for optional feedback.

Your database already knows what's slow. Go read it.

$56 one-time · ≈ ฿1,990 · free updates v1.x · instant download

Get the Postgres Performance Kit — $56