End-to-end analytics
for an online school
We collected all data on advertising, sales, and clients into one dashboard. Now the owner can see which advertising is generating revenue and which is wasting the budget — and makes decisions based on numbers, not intuition
dashboard screens
data rows
data sources
attribution models
What the client came with
An online school with growing revenue, but no understanding of what exactly is working
Data is scattered across 7 systems
Yandex Direct, Google Ads, Metrica, GA4, Bitrix24, payment systems, website — each has its own statistics, and nowhere is there a complete picture
It's unclear which advertising is paying off
Money is spent on 5 advertising channels, but no one knows which one actually generates sales and which one is just wasting the budget
Half of the clients are without a source
Only 49% of applications could be linked to a specific advertising channel. The rest are 'unclear where they came from'
Manual reports in Excel
The marketer spent hours every week to compile data from different systems into a table. The data became outdated faster than it could be collected
Decisions based on intuition
Budgets were allocated based on feelings: 'it seems like Direct is working better'. Without numbers — without confidence
Unable to plan growth
There is no tool for forecasting: how much to invest in advertising to get the desired revenue next month
What we did
We collected all the data into one dashboard with 12 screens. Each screen answers a specific business question
How much was earned
Main screen: revenue, number of sales, average check, returns, cost of customer acquisition — all in real time. Comparison with the previous period in one click
Where people come from
How many people visited the site, where they came from (search, advertising, social networks), how long they stayed on the site, from what devices
How SEO works
Search engine positions, number of clicks from Google and Yandex, which pages generate the most traffic from search
Which advertising pays off
Expenses, clicks, cost per click and leads for each advertising channel and each campaign. You can see where every ruble goes
Who actually brought the client
6 models for determining the source of a sale. A client could see an ad on Yandex, then come from Google, and make a purchase through social networks — the system takes into account the entire journey
Where we lose customers
Funnel from the first visit to purchase. You can see at which stage people leave: didn't reach the catalog? Abandoned the cart? Didn't complete the payment?
What is being bought
Which courses sell best, average check, coupons, cashback, referral program — a complete picture of sales
How much a client costs and how much they will bring
Cohort analysis: how much money a client will bring over time, how often they make repeat purchases, which channels bring the most valuable clients
How the sales department sells
Data from CRM: how many leads are in progress, at which stage deals get stuck, processing speed, manager effectiveness
How much to invest to earn X
You set the desired revenue — the system calculates how much to spend on advertising, how many leads and sales there will be. Three scenarios: optimistic, basic, pessimistic
What to do next — written by AI
Artificial intelligence analyzes all metrics and writes a marketing report with conclusions and specific recommendations. Automatically every week
Settings and connections
Managing data sources, manually entering expenses for channels without API, setting up users and access.
Why businesses need this
A dashboard is not just pretty graphs. It's a tool that saves money and helps earn more
You see where every ruble goes
All advertising expenses are tied to actual sales. Not 'we spent 500 thousand on marketing', but 'Yandex Direct brought 47 sales worth 1.2 million, and VK brought 3 sales worth 80 thousand'.
Find and disable unprofitable channels
In this project, we found 'dead' expenses of 150,000 rubles per month — advertising that didn't bring a single sale. This money was redirected to working channels.
Scale what works
When you see that one channel brings customers for 300 rubles and another for 3,000, the decision is obvious. Data eliminates guesswork and provides a clear action plan.
Plan growth based on numbers
Media planning module: enter 'I want 5 million in revenue in April' — the system calculates how much to invest in each channel, how many leads and sales there will be. Taking into account seasonality and trends.
Save 10+ hours per week
No manual reports in Excel. Data is collected automatically every day. AI writes a marketing report with conclusions — open it in the morning and immediately see what has changed.
Data on your server
This is not a SaaS you pay for every month. The dashboard runs on your server, the data belongs to you. No one will raise the price or cut off access.
How quickly will analytics pay for itself?
Simple math based on a real client example
that's how many times analytics pays for itself in the first year
real ROAS for a client after connecting
How it works: you don't just save — you redirect money from ineffective channels to those that actually drive sales. Same budget, more customers.
Results in practice
Here are real dashboard screens.
Each figure is updated automatically every day.
Skills Up School
Online school • EdTech • Data for 2 months
What we found thanks to the dashboard
We discovered that VK Advertising provides the best return on investment (5.2x) with the smallest budget — we redistributed 20% of the budget from Yandex
Reduced the time for preparing a weekly marketing report from several hours to 8 seconds
Attribution showed that 41% of paid deals are not tied to an advertising source — we connected a backup binding method and returned 3.6M ₽ to analytics
Real indicators are on demo dashboard
How it works under the hood
Data is collected automatically every day — without human involvement
Data collection
11 sources are connected via API: advertising cabinets, analytics systems, CRM, payment systems. Data is downloaded automatically according to a schedule.
Processing
Raw data is cleaned, combined and stored in the fast ClickHouse database. 10 million rows — response in fractions of a second.
Customer matching
The system links an advertising click to a specific sale: UTM tag → Metric identifier → GA identifier → contact in CRM. This is end-to-end analytics.
Visualization
12 dashboard screens, each updated automatically. Plus an AI report with conclusions and recommendations every week.
Do you want to see where every ruble of advertising goes?
Tell us about your business — we'll show you what your dashboard will look like. The first consultation is free.