critique my first sales letter

Hi There,
Long time lurker, just registered to post my first sales letter.
I'm a developer focusing on math and machine learning - not a copywriter.

This sales letter is to get interest in a service for debt collections agencies to improve their outgoing collections attempts.

-- About the industry --
The collections (or accounts receivable management if they're trying to sound sophisticated) industry is about as sleazy as you can get. People fail out of used car sales into it. They do not attract the best and brightest,
and their technology/processes are firmly rooted in the 70/80's

My service provides a solution to targeting who to call after the first letter (known as a Dunning letter) goes unanswered. This is where agencies spend the majority of their money. I can see a 30-325% improvement in recoveries, depending on the debt type. The 325% lift came from high volume, low value medical debt.

-- Letter follows --


A letter from the creator of Derisk


My name is Steven Shack, and I created Derisk after
industry veterans told me the solutions available to them
weren't effective. They had plenty of accounts, but even using
the tools currently available they couldn't effectively target the
ones that paid. This might be you too.


But perhaps I should step back for a moment and explain <i>why</i>
I was being told about this. I have spent the last 14 years of my career
building high performance data driven applications in racing,
military security and quantitative finance. Before marketing
terms like "big data" and "analytics" became en vogue.


The future is already here -- it's just not very evenly distributed.
-- William Gibson


During that time I built systems on the cutting edge of research.
Ones that learnt as conditions changed or new data came in, systems
that add real value to business. The sort of tools and techniques
you'll start hearing about in the press in a few years time.


From that background, I built Derisk to solve a single, highly
focused problem: Identify who is most likely to pay.
I based my work on modern statistical and computer science based
techniques. Techniques that have only been commercially viable in
the last few years.


Derisk takes anonymized data about an account like age, sex, balance,
location, and combines it with additional open and proprietary datasets
to make up an enriched account. This is quite normal, many collections
agencies do this in house using skip tracing and credit bureau scores.


Here we deviate from the exiting tools. Instead of relying on
statistical techniques that were cutting edge int he 80's
and 90's, Derisk takes approaches from the cutting edge of research.
Derisk uses techniques from Machine Learning which let Derisk
learn not simply what variables affect payment, but how those variables
affect each other, in nuanced ways to affect payment.


Derisk doesn't simply learn people in a certain age group pay X% of
the time, it learns mid-20's females, with urban lifestyles, pay less
frequently than the equivalent men. This is not something a
person can do reliably and traditional statistical tools and services
miss the nuance entirely.


How do I know it works? The same way we know Asprin works - science.
With each customer I start with a free mini-clinical trial, where
we test and measure exactly how Derisk improves their business. So I've done exactly that, Depending on the type of debt the results
are a lift from 30-325%.



Tools can actually impact the way we think. Derisk lets you separate
your flow of accounts so you can focus your planing at a higher level.
If you'd like to see a lift in your recoveries and start working at
higher level contact me and we'll get started right away.

Sincerely
Steven Shack
Creator of Derisk