Store Analytics Without Another
Monthly Dashboard Subscription
Lifetimely and Polar Analytics are dashboards you rent by the month, good at watching KPIs, thin on answers. When you need a deep answer, LTV cohorts, RFM segments, churn drivers, seasonality, commission it once: upload your Shopify or order CSV, describe the question in plain English, and get a statistical report with the R code included, re-runnable on fresh exports whenever you want.
The right way to answer each question, not one way for all of them.
Most tools have one way to answer. We have several, and the question decides which: a quick read when that is enough, a tested method when it is not, and an analysis built for your data only when nothing else fits. The method is named in every report.
A Dashboard Shows You the Number. It Doesn't Explain It.
Subscription dashboards are built to monitor: the same predefined KPIs, refreshed daily, for a monthly fee. The questions that actually change what you do next need analysis, not another chart.
Statistical Analysis, Not KPI Tiles
A dashboard tells you repeat rate dropped. A commissioned analysis tells you which cohorts drove the drop, whether the change is statistically meaningful, and what the seasonality-adjusted trend actually looks like, with the methodology and executed R code included in the report.
Pay Per Analysis, Not Per Month
No standing subscription quietly renewing whether you looked at the dashboard or not. Start with a free Answer, no card required, and spend credits only when a question needs deeper work. If a build fails, you're never billed.
An Analysis You Keep
Cancel a dashboard subscription and the dashboards are gone. A commissioned analysis is a re-runnable asset in your library: export next quarter's orders, run the identical methodology again, and compare like with like.
Export. Ask. Get an Answer You Can Act On.
No integrations, no onboarding call, no data warehouse
Upload Your Order Export
Export orders from Shopify, WooCommerce, or Amazon as a CSV and drop it in: from your browser, or directly inside Claude, ChatGPT, or any MCP client. Your data is encrypted and stays yours.
Ask in Plain English
Describe the question the way you'd brief an analyst: "Which cohorts have the best LTV?" "What's driving churn?" You don't pick a model, a method or a depth: the question decides, and you get an Answer, Analytics, or an analysis built for your data.
Get a Re-runnable Report
A pipeline of AI agents selects a statistical method, writes and executes real R code, and delivers an interactive report: methodology and code appendix included. Re-run it on a fresh export anytime.
What You Get With MCP Analytics
Built for the store questions a KPI tile can't answer
Answer
A quick question about your data, answered in seconds from all of it, with how it was worked out. An Answer is free.
Analytics
A tested statistical method from our library, run on your data and written up: the chart, the numbers and the method that produced them.
Custom Analytics
A full analysis built for your question when nothing in the library fits: re-runnable, yours to run on fresh data.
LTV & Cohort Analysis
Lifetime value by acquisition cohort, with real statistical treatment, not a single blended average on a tile. See which months, channels, or campaigns produce customers worth keeping.
RFM Segmentation
Recency, frequency, and monetary segmentation of your customer base, who's a champion, who's drifting, who's worth a win-back campaign, from a plain order export.
Churn & Seasonality
Repeat-purchase and churn driver analysis, plus seasonality decomposition and forecasting, so you can tell a real trend from a December.
Executed R Code Included
The exact script that produced your results ships inside the report. Not pseudo-code, not a summary: the code that ran, so the numbers are verifiable.
Delivered Where You Work
MCP-native: run and read your analyses inside Claude, ChatGPT, or any MCP client, or in the browser. We come to your chat; we don't lock results in our app.
Nothing to Set Up
No model to pick, no method to choose, no dashboard to build. A question and your data, by the site, by email or from your own AI.
MCP Analytics vs Lifetimely & Polar Analytics
Based on each product's own public positioning: different tools built for different jobs
When a dashboard subscription is the better choice
Honestly: Lifetimely and Polar Analytics are good at the job they were built for. Pick a subscription dashboard over MCP Analytics if what you want is:
- Daily KPI monitoring: live numbers you glance at every morning without asking a new question
- An always-on dashboard the whole team keeps open, updating automatically as orders come in
- Centralizing data from many channels (ads platforms, email, storefronts) into one continuously synced view
Pick MCP Analytics when you need the answer, not the feed: a deep analysis with its method named you commission once, own outright, and re-run on fresh data whenever you want. Plenty of stores use both.
Lifetimely & Polar Analytics Alternative FAQ
Can MCP Analytics replace Lifetimely or Polar Analytics?
It replaces the deep-analysis part of the job, not the always-on dashboard part. Lifetimely and Polar Analytics are subscription dashboards built for continuous KPI monitoring, if you want live numbers on a screen every morning, that's what they're for. MCP Analytics is for the questions a dashboard summarizes but never really answers: which cohorts actually drive lifetime value, what's behind your churn, how seasonality decomposes. You commission that analysis once, get a statistical report with the R code included, and re-run it on fresh exports whenever you want, with no standing subscription. Many stores use both; some find commissioned analyses on a fresh export cover everything they actually looked at the dashboard for.
What data do I need to get started?
An order export as a CSV: Shopify, WooCommerce, Amazon Seller Central, and most other platforms can export orders in a couple of clicks. A typical export with order IDs, customer IDs, dates, and order values is enough for LTV cohorts, RFM segmentation, and churn analysis. No integrations to configure, no pixel to install, no data warehouse required: upload the file, ask your question, and re-upload a fresh export whenever you want the analysis re-run on current data.
What kinds of analyses fit an e-commerce store?
The classic deep questions a store's order history can answer: customer lifetime value by acquisition cohort, RFM segmentation to find your best and most at-risk customers, churn and repeat-purchase driver analysis, seasonality decomposition, and revenue forecasting. Because you describe the question in plain English rather than picking from predefined metrics, you're not limited to a fixed menu: if your order data can answer it, a commissioned analysis can be built for it.
Do I need a data team or know statistics to use this?
No. You describe the question the way you'd brief an analyst: "which customer segments are worth a win-back campaign?" or "how much of my revenue swing is seasonal?", and a pipeline of AI agents selects a statistical method, writes and executes real R code, and delivers a report that explains the findings in plain language. The methodology and the exact code that produced your numbers are included, so anyone technical can verify the work, but nobody on your team has to write it.
What does MCP Analytics cost compared to a dashboard subscription?
There's no subscription. You can start free: an Answer costs nothing and requires no credit card. Deeper work, Analytics and Custom Analytics, uses credits, you pay per analysis, and a failed build is never billed. Dashboard tools charge a monthly subscription for continuous access; here you pay when you commission an analysis, and what you bought stays yours to re-run afterward. Full details are on the pricing page.
Keep Exploring
The store analyses behind this page, and how we compare
Comparisons
Get the Answer, Skip the Subscription
Bring a dataset and a real question. Get back an analysis with the method written down, the code that ran, and a report you can hand to anyone.
Or email ds@mcpanalytics.ai a file and a question. No account, no sign-up first.
Not sure yet
Send the question first. See what comes back, then decide.
MCP Analytics
Data science by email
Send us your data and your question, and we will reply.