# Closing a product that was already in the market

Simbiose Ventures · 2021
Author: Henrique Senra (https://henriquesenra.com)
Web version: https://henriquesenra.com/cases/simbiose-permissionlayer-moovme

> I took over an internal data-security product, made the case that it carried too much risk for clients and too little interest, and closed it. The team moved to projects that ran one to two years.

## Key outcomes

- Closed an internal product that had been taken to market and found no fit.
- The team of five developers, a process manager and a product marketing manager moved to projects that ran one to two years.
- Told a prospect that the product was not what they needed, which avoided another mismatched sale.
- Cost redirected to paying work, directional. No saving was measured.

Tags: Product discovery, Kill decision, Roadmap, Simbiose

## TL;DR

- **Problem.** PermissionLayer was an internal product with a full team and no usable, proven-valuable version. Clients had a pain they were not feeling.
- **My role.** In June 2021 I became its product owner and closed it myself. I also told one prospect, Moov.me, that they wanted something we were not selling.
- **Decision and cost.** I closed it for too much risk to clients and too little interest. The cost was the product itself and the work already spent.
- **Outcome.** The team went to projects that ran one to two years. Business end: cost redirected to paying work, directional. No saving was measured.

## Context

Simbiose Ventures built data products. PermissionLayer was conceived in March 2020, after the company's focus left SlicingDice, its data platform. It was a row-level permission layer for any SQL tool. My LinkedIn text calls it a "man in the middle" application that gives companies column and row permissions over any data source in their stack.

It was an internal product, as opposed to products built for outside clients. Its team was five developers, a process manager, a product marketing manager and me. In my notes on whether to stay at the company, I set the goal that it be the first product to turn a profit. I also wrote that we did not have a usable product, and that we were not sure we had a valuable one.

Between February and August 2021, several new clients arrived, Moov.me among them.

## What I owned

In June 2021 I went on vacation. When I came back, a new product owner had taken FastCRUD, the internal tool that every later Simbiose product was built on, and I became product owner of PermissionLayer. I closed it. Asked who did, I answered "That was me."

What discovery looked like, from my working notes.

- A criterion for product-market fit, in revenue. We knew neither the total nor the serviceable market, so any number was arbitrary. Ten to twenty "crazy people who buy" would indicate it.
- Two lines of work. One revisited the problem space with pain and job hypotheses, a market survey and a new Mom Test interview. The other pushed a hypothesis around Metabase, a dashboard tool, with emails to its users and other channels.
- A risk I wrote down: the product was the cheap version of an enterprise Metabase feature, and Metabase could shut us off by removing the user ID.
- After comments from an outside contact, the team saw data-security professionals, such as security, information and data officers, as the segment to attack first.

## Options

```mermaid
flowchart TD\n  A["PermissionLayer in the market, no usable proven version"] --> B["Keep pushing the Metabase route"]\n  A --> C["Go after the data-security segment"]\n  A --> D["Close the product"]\n  D --> E["Team moves to projects that ran 1 to 2 years"]
```

The record does not hold a formal list of alternatives. The choices visible in my notes were these.

| Option | Where it stood |
| --- | --- |
| Keep pushing the Metabase route | A dependency on a tool that could cut us off. |
| Go after the data-security segment | Interviews were planned. The record does not say what they found. |
| Close the product | What I chose. |

I closed it because the risk to clients was too large and the interest too low. The pain was real, but the clients had it without feeling it. In my vacation notes I wrote that I could not believe in the product, since it had been taken to market and had no fit. The cost I accepted was the product itself and the work already spent on it.

A second decision sat beside it. Moov.me wanted to build a minimum viable product and then decided to run the process by forms and spreadsheets and not by software. I made clear that what they wanted was not what we were selling, and I influenced them to leave. In my own comparison, this avoided another mismatched sale like the one on the Scoutz project, where the client wanted one thing and was convinced to buy another.

## Results

The closing freed the team. In my words, they were low salaries but valuable assets that were directed to projects that lasted one to two years within the business.

Nothing else is measured. I do not have a figure for the investment saved or the time saved. The older figure of hundreds of thousands in savings has no documented basis, so I do not use it. My career narrative also ties a strategic shift, growth from near zero to 20 thousand in monthly recurring revenue in nine months, and a team growing from 10 to 90 people, to this period. My own answers attribute them to many clients and to software-house work, not to this decision, so I do not claim them for it.

## Friction and what I would change

The part I noted as hard was engagement. I wrote that I felt the biggest difference was in engagement, and that wear and the absence of full-time people got in the way. I asked myself how to handle part-time people on a team. In my comparison with a client team that had more structure, PermissionLayer's team had one practice, putting items on the calendar and warning when the deadline blew.

## Limits

- No saving, revenue or margin is measured for this decision.
- Do not claim that the closing caused the later growth in revenue or headcount.
- The count of seven people is the team around me, and I know where they went only in broad terms.
- What the data-security interviews found is not recorded.
