Independent CRM product intelligenceGlobal edition · Updated 06 Aug 2026
CRMatlas
Index / Sales skills
Prospecting · Research · Deals · Sales · CRM

Run the complete sales motion.

Agent skills for finding, researching, and winning customers—plus operational CRM playbooks for the system behind the work.

01
GTM · Sales

Sales skills library

38 practical agent skills for finding the right accounts, researching buyers, progressing deals, and building a repeatable sales motion.

10 skills

Prospecting

Build clean target lists, define the ICP, enrich contacts, and map the buying committee.

View source collection ↗
04

ICP ↗

Open-source procedure · GitHub

12 skills

Research

Turn public and internal evidence into account briefs, battlecards, market intelligence, and confident next actions.

View source collection ↗
10 skills

Deals

Find deal risk, strengthen stakeholder coverage, run pipeline reviews, and identify expansion opportunities.

View source collection ↗
6 skills

Sales execution

Improve discovery, qualification, coaching, objection handling, and the operating sales playbook.

View source collection ↗
Adapted and independently edited by CRMatlas from the open-source GTM Skills catalogue (MIT). Every skill links directly to its source directory on GitHub; descriptions and information architecture are CRMatlas editorial work.
02
CRM · Operations

CRM playbooks that end in execution.

Practical procedures for teams selecting, implementing, and improving a CRM. Every playbook moves from a real problem to a concrete deliverable and a clear definition of done.

01
Starter · 2–3 hours + testing

Build a CRM selection scorecard

Outcome: A shortlist based on your process, not the demo

When to use it

When several systems are being compared and each stakeholder uses different criteria.

Steps

  1. Describe the 5–7 critical moments in your sales process: qualification, lead handoff, proposal, follow-up, forecast, and reporting.
  2. Weight the criteria. Process fit and adoption should matter more than the number of features.
  3. Define disqualifiers such as no full export, SSO, local support, or a required integration.
  4. Run the same scenario and dataset in every system.
  5. Record the score, strength of evidence, and cost of each compromise.
02
Starter · 1 workshop + 2-week pilot

Design a pipeline sellers will keep clean

Outcome: A readable pipeline and a forecast people can trust

When to use it

When stages describe seller activity, records age quickly, and managers correct the forecast manually.

Steps

  1. Give every stage one unambiguous entry and exit condition based on customer behavior.
  2. Limit required fields to information needed for the next decision.
  3. Set a maximum idle time and a rule for moving back or closing an opportunity.
  4. Make next step, date, and owner the minimum standard for an active opportunity.
  5. Pilot with one team and adjust stages using real exceptions.
03
Advanced · 2–6 weeks

Migrate CRM without losing history

Outcome: Clean data and a controlled cutover

When to use it

When replacing a CRM or consolidating databases after a reorganisation, acquisition, or branch migration.

Steps

  1. Inventory objects, fields, owners, automations, attachments, and integrations.
  2. Classify data as migrate, archive, or delete—do not copy unused disorder.
  3. Agree deduplication keys and value mappings before the first import.
  4. Run a trial migration, reconcile a control report, and obtain process-owner approval.
  5. Freeze changes, make a recovery export, run the final migration, and reconcile record counts.
04
Intermediate · 30 days

Launch a CRM in 30 days without redesigning the company

Outcome: A working MVP scope

When to use it

When implementation is stuck in a wish list and the team still lacks a usable core system.

Steps

  1. Choose one process, one team, and three measurable implementation outcomes.
  2. Build the minimum data model, roles, pipeline, and essential reports.
  3. Import a representative data sample and test real scenarios.
  4. Train users on their actual opportunities, not feature slides.
  5. Run a daily office hour, issue list, and configuration decision log for two weeks.
05
Intermediate · 2–4 weeks

Fix CRM adoption without adding fields and policing

Outcome: A system updated as work happens

When to use it

When sellers update CRM before a manager meeting instead of while selling.

Steps

  1. Measure entry delay, opportunities without a next step, and activity happening outside CRM.
  2. Remove fields and tasks that do not support a decision or automation.
  3. Connect email, calendar, and communication where it removes duplicate entry.
  4. Redesign management rituals so pipeline reviews use CRM only.
  5. Publish a weekly team-level data-quality report, not a public shame ranking.
06
Advanced · 2 workshops + 1 quarter of calibration

Build an evidence-based forecast

Outcome: Visible risk and a repeatable forecast

When to use it

When probabilities are arbitrary and commit changes without new customer evidence.

Steps

  1. Define pipeline, best-case, and commit categories with required evidence.
  2. Separate stage probability from the assessment of a specific opportunity.
  3. Track risks: decision process, sponsor, competition, date, and next step.
  4. Compare forecast with outcome weekly and record the cause of every variance.
  5. Calibrate quarterly using cohorts, not memorable wins.
07
Intermediate · 1 week design + ongoing cadence

Create a data hygiene system that needs no rescue projects

Outcome: Data ready for reporting and automation

When to use it

When duplicates, missing owners, and inconsistent values block reporting or automation.

Steps

  1. Select critical fields and assign a quality owner to every object.
  2. Define rules for creating, updating, merging, and deleting records.
  3. Validate at entry instead of periodically cleaning the whole database.
  4. Create an exception queue with an owner and due date.
  5. Monitor completeness, uniqueness, freshness, and validity separately.
08
Advanced · 1–3 weeks

Automate CRM without creating invisible debt

Outcome: Less manual work without losing control

When to use it

When automations duplicate actions, lack owners, or break after process changes.

Steps

  1. Map event → conditions → action → owner → log.
  2. Estimate frequency, manual effort, and error cost; automate frequent, reversible tasks first.
  3. Add an error path, alert, and retry mechanism.
  4. Test edge cases: missing owner, duplicate, reversed stage, and delayed integration.
  5. Quarterly remove automations without an owner, use, or measurable outcome.