Law Enforcement Transcription Rollup
Decided not to pursue this because the total market turned out to be too small ~150M.
Context:
Public interactions with law enforcement officers are recorded, including 911 calls, body cam recordings, and court recordings.
Conversion of these audio files into text is a legal requirement.
Small mom-and-pop agency services exist to handle this task.
Employees of these agencies listen to the recordings and manually transcribe them into text.
The industry is regulated, and compliance with CJIS (Criminal Justice Information Services) and FedRAMP (Federal Risk and Authorization Management Program) is necessary to operate.
Some agencies claim that human transcriptions are more accurate due to the transcribers' understanding of legal terminology and ability to comprehend accents.
The business model is straightforward: Transcribers are paid by the hour, and clients are charged per minute of audio transcribed.
What is the idea?
Acquire multiple transcription services in the industry.
Consolidate these services into a single large entity.
Implement automation technology to streamline and optimize the transcription process.
Reduce costs by laying off a significant portion (50-70%) of the workforce.
Maintain the same revenue while cutting down on labor expenses.
Acknowledge that complete automation may not be feasible, necessitating the retention of a workforce for reviewing and quality assurance purposes.
Aim to increase profitability through efficiency gains and cost reduction.
Why now?
Transcription software, such as OpenAI's Whisper, has become highly affordable, user-friendly, and accurate.
OpenAI Whisper, released approximately 9 months ago, offers advanced transcription capabilities.
Automation of the transcription process can be achieved through the utilization of Whisper and similar software.
Reviewing and correcting the transcripts may also be automated using ChatGPT, an AI language model.
Leveraging these technologies can potentially reduce the need for manual workforce involvement in the transcription and review processes.
The combination of Whisper and ChatGPT can contribute to significant cost savings and improved efficiency in the transcription business.
Why is it a good idea?
The transcription industry consists of numerous small players that can be targeted for acquisition.
The industry is regulated, which poses barriers to entry for new players, making acquisitions a logical approach for growth.
Acquiring smaller transcription services allows for consolidation and centralized operations.
Implementing technological transformations can lead to economies of scale and increased operational efficiencies.
Centralizing operations enables better resource allocation and optimization of processes.
The combination of acquisitions and technology adoption can enhance competitiveness and profitability in the industry.
Why might it be a bad idea?
It's a poor choice if the industry is dying, or if the entire business will be automated away.
I believe the need for transcription won't disappear.
I think businesses won't vanish either; they will instead look a lot more like "robot-assisted assembly lines" rather than everything being done manually. By this, I mean that businesses and people will persist to ensure accuracy, reliability, and liability. The bulk of the work will be carried out by machines, with humans at the forefront to ensure everything keeps running smoothly and there is someone to hold accountable if anything goes wrong.
Parallels in other industries?
Thrasio in e-commerce. They acquired a bunch of Amazon FBA brands to consolidate and gain economies of scale.
How can we invalidate it? What are the next steps? Answers to figure out?
I need data on, number of such businesses, their annual revenue, costs of FTE, licensing requirements, variance from state/county level, resistances to automation.
Build a financial model to prove the model makes sense. This will also tell us what are the revenue multiple in which it makes sense to acquire the business, also include people cost, negotiation cost, legal cost etc.
[Note: I got to read an ibis world report about transcription businesses, which lead to the ~150M market size number given above]
Appendix:
Examples of law enforcement transcription services, How to find these transcription services?
https://www.dittotranscripts.com/
http://www.ckreporting.com/
https://www.bwtsi.com/
What is their revenue like?
One I looked at had ~$700K annual revenue. Need more.
How many of such businesses are there?
To find out