Evicta

Everyone in the SaaS world is obsessed with Monthly Recurring Revenue (MRR). But when I started validating my latest project, I realized forcing MRR onto a "One-and-Done" problem is a great way to kill your conversion rate.

The Problem I Found
I noticed a massive pain point for technical teams: migrating off Zendesk. If a company wants to move to a new helpdesk, or an ML team wants to use their historical support tickets to train an AI bot, getting the data out is a nightmare. The native exports are flat CSVs that completely miss deeply nested JSON "Custom Objects" and conversation threads.

The existing ETL tools (like Fivetran) charge by Monthly Active Rows. But if you just want to extract your data once and cancel Zendesk, a recurring pipeline makes no sense. You just want the data.

The Product (Evicta)
I spent the last month building Evicta (evicta.dev). It’s a stateless extraction engine. It uses a dynamic rate-limit governor to handle the Zendesk API, an LLM to surgically map nested custom objects, and streams the data directly to a clean SQL or JSONL archive.

I built it entirely stateless on purpose. 0 bytes of the customer's PII ever touch my disks; it streams directly into an ephemeral, encrypted R2 vault.

Why I Chose a $500 Flat Fee:

1. Alignment: The customer's goal is finality. They want their data so they can close a chapter. A flat fee aligns perfectly with their psychological state.

2. The "Discretionary" Threshold: In most mid-market companies, $500 is under the CTO or VP of Engineering's discretionary limit. They don't need to ask Procurement or Finance for approval. They can just put it on a corporate card and save their engineers 2 weeks of writing custom Python scripts.

3. The Upsell: I do have a $1,499 tier for Enterprise teams that require a Zero-Trust Local CLI to run the extraction inside their own VPC.

Most advice says if you do B2B, you must charge a subscription.

I'm curious to hear from other founders—have any of you successfully built a high-ticket, flat-fee SaaS? Do you think I'm leaving money on the table by not forcing them into a monthly pipeline model?