GTM Engineering glossary and flashcards

Drill the terms

Read the term. Say the definition out loud before you flip, then be honest. A card you know moves up a box and comes back less often. A card you miss goes back to the first box. Ten minutes a day beats an hour on Sunday. Your boxes stay in this browser only.

Every term, A to Z

Search terms

Action title

A slide title written as the finding, such as "Tasks grew 34%", not a topic label. Stage 11

Activation

The first use of the core feature that shows an install is real. Stage 9

Agglomerative clustering

Grouping by repeatedly merging the two closest groups. Stage 7

Aha moment

The first time a user feels the product's value, measured through a proxy action. Stage 4

Alert fatigue

People ignoring alerts because there are too many of them. Stage 2

Anchoring

The pull that a first-seen number exerts on later price judgments. Stage 12

ASC 606

The US accounting standard that says when a company may count customer contract money as revenue. Stage 5

Authorization code flow

The OAuth flow where the app gets a one-time code through the browser and swaps it for tokens on the server. Stage 9

Battle card

A one-page guide for handling one objection or competitor. Stage 7

Benchmark

A median or percentile from a sample, only meaningful with its sample and definition. Stage 12

Bootstrap interval

A range found by redrawing your own data with replacement many times, which shows how noisy a number is. Stage 10

CDX server

The Wayback Machine index that lists every capture of a URL with timestamps. Stage 12

Censoring

A customer still active today whose final lifetime is not yet known. Stage 8

Charm pricing

Ending prices in 9, such as $19.99, to raise demand, supported by field tests on retail items. Stage 12

Chunking

Cutting a transcript into overlapping pieces for search. Stage 7

Claygent

Clay's AI agent that takes inputs, follows your instructions and writes a structured output. Stage 3

Cohen's kappa

A score of how much two coders agree after removing chance. Stage 10

Confounder

A third factor, such as company size, that drives both cause and outcome and fakes a link. Stage 10

Contract liability

The accounting record for cash received before the service is delivered, also called deferred revenue. Stage 5

Conversation intelligence

Software that records, transcribes and analyzes sales calls. Stage 7

Cooldown

A rule that blocks repeat alerts for the same account for a set time. Stage 2

Cosine similarity

A score for how closely two vectors point the same way. Stage 7

Credit burn-down

Spending a prepaid balance as usage happens, in a fixed order, until it reaches zero. Stage 5

Custom object

A table you define inside the CRM to store your own concept, such as a billing account. Stage 1

Dark social

Sharing through private channels that carries no referrer, so analytics show it as direct. Stage 10

Data leakage

Using information in training or features that would not exist at prediction time. Stage 8

dbt model

A SQL select statement that dbt materializes as a table or view. Stage 1

Deanonymization

Working out which company or person is behind anonymous website traffic. Stage 2

Debouncing

Making sure many triggering events cause only one alert. Stage 4

Decoy effect

A lab result in which adding a clearly worse option shifts choices toward the option it resembles. Stage 12

DKIM

A cryptographic signature on each message, checked against a public key in DNS. Stage 3

DMARC

A DNS policy that says what to do when SPF and DKIM fail to align, and where to send reports. Stage 3

DRQL

DevRel Qualified Lead: anyone from the community who can add value to the company. Stage 10

EBR

Executive business review, a less frequent meeting for executives that focuses on business outcomes. Stage 11

ELT

Load raw data into the warehouse first, then transform it there with SQL. Stage 1

Embedding

A list of numbers that places text in space by meaning. Stage 7

Expansion

Extra revenue from an existing customer through seats, tiers, usage or new products. Stage 8

Extremeness aversion

The lab finding that the middle option in a set gains appeal, the basis of good-better-best ladders. Stage 12

First-party intent

Intent collected on your own website, app or email. Stage 2

First-touch and last-touch

Models that give all credit to the earliest or the latest recorded touch. Stage 6

Gross revenue retention (GRR)

The share of starting revenue kept after churn and contraction, capped at 100%. Stage 8

GRR

Gross revenue retention: the share of starting recurring revenue kept, ignoring any expansion. Stage 12

GTM (go-to-market)

Everything a company does to find, win and keep customers. Stage 1

HDYHAU

Short for "How did you hear about us?", the open question behind self-reported attribution. Stage 10

Health score

A number for how likely an account is to renew, from rules or a model. Stage 8

Holdout test

An experiment that withholds an activity from a random group to measure lift. Stage 6

Human in the loop

A person approves, edits or rejects each AI draft before anything is sent. Stage 3

Hybrid pricing

A recurring fee combined with usage, credits or overages. Stage 5

ICP

Ideal customer profile, the description of companies and people most likely to buy. Stage 3

ICP

Ideal customer profile: a description of the companies you most want as customers. Stage 4

Idempotency

The property that repeating an operation has the same effect as doing it once. Stage 1

Idempotency

A property of a process that gives the same result whether you run the same input once or many times. Stage 4

Idempotency key

A unique ID sent with an event so a repeat of the same event is ignored. Stage 5

Idempotent

Safe to run twice, because the second run changes nothing. Stage 2

Identity resolution

Deciding which records across systems are the same person or company. Stage 1

Incrementality

Conversions caused by an activity beyond those that would have happened anyway. Stage 6

Influenced pipeline

Pipeline credited to every activity that touched the account, so channel totals overlap. Stage 6

Integration-led growth

A go-to-market motion where connecting to the tools buyers already use drives discovery, trust and retention. Stage 9

Lift

The gap in outcome rate between two groups, such as word-of-mouth accounts versus other accounts. Stage 10

Loaded hourly cost

A wage plus the employer's cost of benefits and taxes, used to price saved time. Stage 11

Logo churn

The share of customers, counted by number, that left in a period. Stage 8

Marketplace review

The vendor's check of security, scopes, listing and user experience before an app can be listed. Stage 9

Meter

A rule that turns a stream of events into one number per customer per period. Stage 5

Mirror sync

A sync that also removes destination records whose source rows disappeared. Stage 1

MQL

Marketing qualified lead: a lead qualified by marketing activity such as a form fill. Stage 4

MRR

Monthly recurring revenue, which only means something once its rules are written down. Stage 1

Net revenue retention (NRR)

Revenue now from last year's customers, including expansion, divided by their revenue a year ago. Stage 8

Non-response bias

The error from the people who answer a survey differing from those who skip it. Stage 10

NRR

Net revenue retention: the share of starting recurring revenue kept after churn and expansion. Stage 12

Outcome-based pricing

Charging when a defined result occurs, such as one resolved support conversation. Stage 5

Overage

Usage billed at a list rate after included allowance or prepaid credits run out. Stage 5

Payback period

The time it takes for monthly net benefit to repay the upfront cost. Stage 11

PKCE

An add-on where the app sends a hash first and proves it holds the original secret at token exchange. Stage 9

Placeholder token

A marker such as {hours_saved} that the model writes and code replaces with a real number. Stage 11

PLG

Product-led growth: a go-to-market motion where the product drives acquisition, conversion and expansion. Stage 4

PQA

Product qualified account: an account where enough users show PQL behavior. Stage 4

PQL

Product qualified lead: a user or account qualified by in-product behavior plus fit and intent. Stage 4

Precision at k

The share of the top k flagged accounts that truly churn. Stage 8

Prompt injection

Instructions hidden in scraped content that try to change what the model does. Stage 3

QBR

Quarterly business review, a recurring meeting where a vendor reviews progress against goals with a customer. Stage 11

Recall bias

The error from people forgetting or misremembering how they found you. Stage 10

Refresh token

A longer-lived credential that gets a new access token when the old one expires. Stage 9

Removal effect

In a Markov model, the drop in conversion probability when one channel is removed from all journeys. Stage 6

Retrieval-augmented generation

Fetching trusted text first so the model answers from it. Stage 7

Reverse ETL

Copying modeled warehouse data back into operational tools such as the CRM. Stage 1

Reverse IP lookup

Matching a visitor's IP address to the organization that owns that address range. Stage 2

Right-censoring

Missing outcomes for recent cohorts that have not had time to convert. Stage 6

Risk adjustment

Shrinking estimated benefits and raising estimated costs to allow for uncertainty. Stage 11

Rolling window

A time window that moves with each event, such as the last 7 days up to now. Stage 2

Scope

A named permission, such as commands, that limits what an access token can do. Stage 9

SDR

Sales development representative, the person who finds prospects, sends the first message and books meetings. Stage 3

Second-party intent

Another company's first-party data, such as review-site activity, shared or sold to you. Stage 2

Self-reported attribution

Asking the buyer how they found you and using the answer as an attribution signal. Stage 10

Shapley value

A channel's average marginal contribution across every order channels could join. Stage 6

Signal

A recorded event that suggests an account may be ready to buy. Stage 2

Snapshot date

The date when a prediction is made, so only earlier data may be used. Stage 8

Sourced pipeline

Pipeline credited to the activity that created the opportunity. Stage 6

Speaker diarization

Labeling which speaker said each part of a recording. Stage 7

Spend cap

A limit, enforced before the work runs, on how much a customer can consume in a period. Stage 5

SPF

A DNS record listing which servers may send mail for your domain. Stage 3

Structured output

A model reply forced to match a JSON Schema so code can use it without parsing free text. Stage 3

Structured output

A model reply forced to match a JSON schema you define. Stage 7

Structured output

A model reply forced to match a JSON schema. Stage 11

Surge score

A vendor score comparing an account's recent topic reading with its own earlier baseline. Stage 2

Swap test

Replace the prospect's name with another name; if the email still reads fine, it is not personalized. Stage 3

Teardown

A written analysis of how one company sells, built from dated, checkable evidence. Stage 12

Template merge

Copying a slide template and replacing tags like {{name}} with real values. Stage 11

Third-party intent

Topic-reading data pooled across many publisher sites by a vendor. Stage 2

Tracking plan

The written list of events, names, properties and owners a team agrees to send. Stage 4

TTFHW

Time to first hello world: how long a new developer takes to reach first successful use. Stage 10

Usage-based pricing (UBP)

A model where the customer pays for units consumed, such as tokens, runs or API calls. Stage 5

UTM parameters

Link tags such as utm_source and utm_medium that tell analytics where a visit came from. Stage 6

Value metric

The unit you bill, chosen so the bill grows when the customer gets more value. Stage 5

Value metric

The unit a vendor bills, such as seats, credits or tokens, chosen so the bill grows with customer value. Stage 12

Value realization

Showing that the customer reached the result they bought the product for, against an agreed baseline. Stage 11

Vector database

A store that quickly finds the vectors nearest a query. Stage 7

Warehouse

A database built for large analytical queries, such as Snowflake, BigQuery or DuckDB. Stage 1

Waterfall enrichment

Asking data providers one after another and stopping at the first that returns a result. Stage 3

Webhook

An HTTP request a vendor sends to your server when an event happens. Stage 1

Zap template

A pre-built Zapier workflow that users switch on in a few clicks, available only for public integrations. Stage 9