Analytics
Data, attribution, unit economy. UTM, end-to-end analytics, DRR, bounce rate, CJM, NPS - terms for those who make decisions based on numbers.
Cohort Analysis - dividing clients into groups by date of attraction and analyzing their behavior over time.
An A/B test is an experiment where two versions (A and B) are shown to different audience groups to compare metrics.
Attribution is the distribution of credit for a conversion between the channels that affected it. It depends on the model which channel the lead will “receive” in the report.
UTM tags are parameters in the link (source, medium, campaign), by which analytics understand where the referral came from. The basis of fair attribution.
DRR (share of advertising expenses) - advertising budget ÷ advertising revenue × 100%. Russian analogue of ACoS, the inverse of ROAS.
End-to-end analytics - a combination of advertising, website and CRM in one chain: you can see not only clicks and leads, but also revenue attributable to a specific ad.
Unit economics - calculating profit from one client or order: will the business remain profitable when scaled? Calculate it before increasing the advertising budget.
Bounce rate is the percentage of visits without action on the site. Yandex.Metrica and Google Analytics count differently - this is often confused.
CJM (Customer Journey Map) is a map of the customer’s journey from the first touch to the purchase: stages, emotions, touch points and places where it falls off.
NPS (Net Promoter Score) is a loyalty index based on one question “would you recommend us from 0 to 10.” Promoters minus critics.
Server-side Tracking - transfer of conversion events directly from the server to the advertising platform, bypassing browser blockers.
Cross-device - tracking the path of one user on different devices: phone, tablet, computer for correct attribution.
MMP (Mobile Measurement Partner) is an independent attribution platform for mobile applications: AppsFlyer, Adjust, AppMetrica.
DAU and MAU - the number of unique users who opened the product per day and per month. The DAU/MAU ratio indicates the “stickiness” of the product.
ARPU - revenue ÷ all users. ARPPU - revenue ÷ only those who pay. The difference between them shows the share of audience monetization.
LTV/CAC is a key indicator of business health: the ratio of customer value to the cost of attracting him. Normal >3, catastrophe <1.
Payback Period - the number of months for which the client returns the CAC spent on attracting him. The norm for SaaS is up to 18 months.
Heatmap - visualization of user behavior on the page: where they click, where they look, where they scroll. Tool for CRO.
Webvisor is a Yandex Metrica tool for recording user sessions: literally a video of how a person walked around the site.
GTM (Google Tag Manager) is a container for managing all counters and pixels on a website without editing the code every time it changes.
GA4 - Google Analytics 4, event-based analytics, replacing Universal Analytics in 2023. It is built around events, not sessions and page views.
Statistical significance is the confidence that the difference between the A/B test options is real and not random. Standard: p < 0.05 (95% confidence).
MTA (Multi-Touch Attribution) - conversion attribution to several channels involved in the customer’s journey, and not just the first or last touch.
MMM (Marketing Mix Modeling) is an econometric model: it evaluates the contribution of each channel to sales based on historical data, without tracking users.
First-party Data - data collected directly from your audience: email databases, CRM, purchase history, behavior on the site. Do not depend on third parties.
Cookieless - marketing and analytics without relying on third-party cookies. Relevant after blocking third-party cookies in Chrome and iOS restrictions.
Last Click is an attribution model in which all the credit for a conversion is given to the last channel before a purchase.
CM1, CM2, CM3 - contribution margin levels: sequential subtraction of direct, variable and marketing costs from revenue.
NRR is net revenue retention: how much money the same cohort of clients brings in a year later, taking into account churn, upgrades and downgrades.