How it works

From raw customer data to an explainable risk score — live in an afternoon

No data team, no months-long rollout, no consultants. Here's exactly what happens between installing Lonieta and seeing your first at-risk customer.

Lonieta gets you from raw customer data to an explainable churn risk score in an afternoon by running as a single self-hosted installation on your own server: you import your data, an opinionated default model scores every customer, and your team starts working the highest-risk accounts the same day — no data warehouse, no integrations project, no consultants.

Step 1 — Install on your own server

Lonieta is self-hosted: you run it on your own server or a VM you control, not in our cloud. There's an installer that gets a working instance up in under an hour on a standard Linux server. Nothing about your customer data ever needs to leave your own infrastructure — that's true from the first minute, not a configuration option you have to remember to enable.

Step 2 — Import your data (no warehouse needed)

Drag in a CSV or Excel export of your customers, contracts, payments and usage, or connect Lonieta read-only to your existing database (PostgreSQL, MySQL, MSSQL). There is no data warehouse to build first and no schema project — Lonieta maps your columns and recognises customers, contracts and activity out of the box.

Step 3 — An explainable risk score, not a black box

Every customer gets a churn risk score built from a weighted model: usage trend, payment behaviour, support pressure, contract renewal date and NPS. Unlike an opaque machine-learning model, every factor is visible and adjustable — you decide how heavily each one counts, and for every score you can see exactly which factors pushed it up. That matters practically (your team can act on it) and for compliance (a score you can explain is a very different thing, under GDPR, than an opaque model making decisions on its own).

Step 4 — Work the list, not the dashboard

Instead of a dashboard you have to interpret, Lonieta gives your team a daily worklist: highest-risk customers first, with a suggested playbook and a draft action already prepared. A human decides and sends — Lonieta prepares, it never acts autonomously.

Step 5 — Measure whether it actually worked

Every followed-up action is tracked through to outcome: did the customer stay, and how much revenue did that save. That gives you a real save rate and a churn trend versus baseline, instead of a warning light with no feedback loop.

Why self-hosted changes the timeline

Cloud-only customer success platforms typically require a data integration project, a security review of a third-party vendor handling your customer data, and a multi-week (sometimes multi-month) onboarding with consultants. Because Lonieta runs on infrastructure you already control, most of that goes away: there's no vendor security review to run because your data isn't going to a vendor's cloud in the first place.

Frequently asked questions

How long does it take to set up Lonieta?

Most teams are running with real data the same day: installation takes under an hour on a standard server, and importing a CSV or connecting a database takes minutes. There's no multi-week onboarding project.

Do I need a data team to use Lonieta?

No. Lonieta ships with an opinionated default risk model that works out of the box. You can adjust the weighting yourself through sliders in the app — there's no model to build or data science required.

What does "self-hosted" mean in practice?

You run Lonieta on your own server or VM. Your customer data is processed and stored on infrastructure you control and never leaves it — it is never sent to Lonieta's own servers or any third-party cloud.

Is the risk score a black box?

No. Every risk score is built from a weighted, adjustable model. You can see exactly which factors (usage, payment behaviour, support tickets, NPS, renewal date) contributed to a given score, and how much.

See it running on sample data

Play with the live demo, or talk to us about installing Lonieta on your own server.

To the live demo Read about security