Software Surety Bonds

I read your other post about programming, but i am curious,
does Anyone know where i can get a bond on programming?
i just need a contract completion bond. i already have the contract, but my customer wants me to get a bond for it.
if you have any information i would appreciate it, thanks.
-=levi=-

[Admin]

Finding a bond to guarantee the software code you develop will be quite a challange. I'm not saying no one will do it, but certainly the 16 sureties my agency represents will not. I have bonding companies willing to guarantee the installation of mass produced software, but none of them want to take the risk on the possible bugs of the software. To be honest with you, I think it is a good idea for the sureties to stay away from guaranteeing software development. I was a C.S. major and got to see first hand how troublesome software development can be! I would refer you to another agency, but I don't know of any that can write the development. Sorry I couldn't help more, good luck!

[tachyon]

I, too, am looking for bonding services for software development. It seems to me that if it is in demand, underwriters should consider it despite the high risk. Besides, can't they simply charge a high premium for 'high risk' software projects? I agree that it is time to rethink surety bond underwriting.

I hate dealing with insurance companies because they all are stupid when it comes to technology. When I applied for E&O insurance the agent had me fill out a form that used industry terminology incorrectly and further demonstrated that insurance companies haven't got a clue and need to hire a few more geeks. If they can't even ask the right questions how can they accurately access risk and give an educated premium quote? Also, many of these types of policies are geared towards giant software companies and are ignorant to the needs of small firms.

In a world where just about everyone and their dog is a self-proclaimed webmaster it would be nice to distinguish my company as bonded.

Also, when it comes to suretyship, underwriters tend to be very cautious when it comes to any software performance guarantees. To date the only bonds I have seen approved for software is to guarantee a service contract on the installation of software, never the programming of it.

Anyone want to help me change the industry?

Ben Damman
ben@tachyoninter.net

Tachyon Technologies, Inc.
PO Box 564
Spooner, WI 54801

(715) 635-4344

[Admin]

Great post tachyon! It appears you have done a good amount of research.

It is true that surety underwriters are very conservative in nature. It is also true that most do not have the slightest idea about technology. Which is precisely why most bonding companies do not want to touch it. They are content with the lines of business they write and are hesitant to write contracts they are not familiar with.

The other problem for for software development guarantees is that they are can not be underwritten using traditional methods. The bonding companies do not know how to go about assessing the risk of a programmer. Keeping in mind you can not look at their code, how would you differentiate one programmer from another? Being a programmer yourself, would you feel comfortable guaranteeing (using your money to back it) another programmer's work if they were a complete stranger?

I must say I am not suprised that most bonding companies are not willing to write these types of risks. Per the article you read, they are very stuck in their ways.


PS - I myself, am yet another self proclaimed webmaster/programmer. To be perfectly honest, I would find it difficult to guarantee software projects if I were in their shoes.

[rmontoya]

I am a homeowner in Southern California going through a very tough kitchen remodel. The contractor has a $10,000 construction bond for the project. It's been seven months and the job is still not complete. The problems have been with several cabinet doors, and a granite countertop that was cracked on installation. I'm trying to research and get educated on construction bonds. What does a Construction Bond cover? How and when do I enforce the bond? Any help would be appreciated! Thanks!!!

[tachyon]

Originally Posted by Admin

The other problem for for software development guarantees is that they are can not be underwritten using traditional methods. The bonding companies do not know how to go about assessing the risk of a programmer. Keeping in mind you can not look at their code, how would you differentiate one programmer from another? Being a programmer yourself, would you feel comfortable guaranteeing (using your money to back it) another programmer's work if they were a complete stranger?

Thanks for the kudos, but I'm just a really good "gopher." Here's what I got out of your reply:

1. Bonding companies do not have the know-how to assess a programmer's risk.
2. They can't inspect the source code and therefore cannot determine it's efficacy or integrity.
3. They are afraid of trying new things.
4. Keeping in mind you can not look at their code, how would you differentiate one programmer from another?
5. Being a programmer yourself, would you feel comfortable guaranteeing (using your money to back it) another programmer's work if they were a complete stranger?

My thoughts on each point, respectively:

1. They can easily obtain the resources and know-how to evaluate a programmer's risk, so they're not off the hook just yet.
2. It is not unreasonable for a bonding company to require the programmer to share source code while it's being developed. It's also possible to request snippets of software code from the programmers past works done for previous clients. Of course, this would require bonding companies to have knowledgeable programmers on staff to review these types of code submissions. Oh, and I forgot to mention that a significant portion of a web developer's source code is available for anyone to see. Just go to the View -> Source menu in your browser and you have it.
3. I think bonding companies should move back in with their mommy if they are afraid to acknowledge the real world. (hehe) Although it's nice of them to leave an opportunity for a more forward-thinking bond underwriter.
4. If we assume that it is impossible for the bonding company to inspect the software code, then a number of criteria can be used to judge a programmer or small software firm:

Work history, level of education, verified letters of recommendation, credit score, give you a programming test to see if you fail, any pending lawsuits?, etcetera

5. Scenario time... Let's be realistic. Instead of pretending to be me, let's assume I'm a bonding company that likes bonding software programmers. First of all, I'm the surety for many programmers. That means that I have a good idea who my most dependable, problem-free programmers are. Now let's say you contact me to "become bonded" because a biotech client has asked for that credential. But you're a young guy fresh out of college and very risky. We probably shouldn't bond you, but I accidentally stamp "APPROVED" on your application and ran out of white-out so I just leave it as-is. With surety bond in hand you get to work on the biotech software project. A few weeks go by and we receive a notice from the biotech company letting us know that you are in default of your obligation. Uh oh!
Since I am a bonding company with many successful, dependable programmers (unlike you) I can easily find a qualified programmer to fix your screw-up. Maybe the biotech company pays you using an escrow. That means you weren't paid all the money for the project yet, so we can easily shift those funds to your replacement. Heck, we might even outsource the job to India where programmers are way cheaper and, arguably, just as good as US-based programmers. If I were you, the programmer in default, I would be happy that my client is being "taken care of" and that they aren't going to sue me.

I guess my point here is that it could be done. If bond underwriters would simply take a deep look at some niche markets (like software development) and innovate a little, they could create new demand for surety bonds.

[Admin]

Originally Posted by rmontoya

I am a homeowner in Southern California going through a very tough kitchen remodel. The contractor has a $10,000 construction bond for the project. It's been seven months and the job is still not complete. The problems have been with several cabinet doors, and a granite countertop that was cracked on installation. I'm trying to research and get educated on construction bonds. What does a Construction Bond cover? How and when do I enforce the bond? Any help would be appreciated! Thanks!!!

Please create a "New Thread" when posting new topics.

You will want to read through our bond information section and the Surety Bond Blog. Between the two you will find everything you will ever want to know about surety bonds and more.

[Admin]

Originally Posted by tachyon

Thanks for the kudos, but I'm just a really good "gopher." Here's what I got out of your reply:

1. Bonding companies do not have the know-how to assess a programmer's risk.
2. They can't inspect the source code and therefore cannot determine it's efficacy or integrity.
3. They are afraid of trying new things.
4. Keeping in mind you can not look at their code, how would you differentiate one programmer from another?
5. Being a programmer yourself, would you feel comfortable guaranteeing (using your money to back it) another programmer's work if they were a complete stranger?

My thoughts on each point, respectively:

1. They can easily obtain the resources and know-how to evaluate a programmer's risk, so they're not off the hook just yet.
2. It is not unreasonable for a bonding company to require the programmer to share source code while it's being developed. It's also possible to request snippets of software code from the programmers past works done for previous clients. Of course, this would require bonding companies to have knowledgeable programmers on staff to review these types of code submissions. Oh, and I forgot to mention that a significant portion of a web developer's source code is available for anyone to see. Just go to the View -> Source menu in your browser and you have it.
3. I think bonding companies should move back in with their mommy if they are afraid to acknowledge the real world. (hehe) Although it's nice of them to leave an opportunity for a more forward-thinking bond underwriter.
4. If we assume that it is impossible for the bonding company to inspect the software code, then a number of criteria can be used to judge a programmer or small software firm:

Work history, level of education, verified letters of recommendation, credit score, give you a programming test to see if you fail, any pending lawsuits?, etcetera

5. Scenario time... Let's be realistic. Instead of pretending to be me, let's assume I'm a bonding company that likes bonding software programmers. First of all, I'm the surety for many programmers. That means that I have a good idea who my most dependable, problem-free programmers are. Now let's say you contact me to "become bonded" because a biotech client has asked for that credential. But you're a young guy fresh out of college and very risky. We probably shouldn't bond you, but I accidentally stamp "APPROVED" on your application and ran out of white-out so I just leave it as-is. With surety bond in hand you get to work on the biotech software project. A few weeks go by and we receive a notice from the biotech company letting us know that you are in default of your obligation. Uh oh!
Since I am a bonding company with many successful, dependable programmers (unlike you) I can easily find a qualified programmer to fix your screw-up. Maybe the biotech company pays you using an escrow. That means you weren't paid all the money for the project yet, so we can easily shift those funds to your replacement. Heck, we might even outsource the job to India where programmers are way cheaper and, arguably, just as good as US-based programmers. If I were you, the programmer in default, I would be happy that my client is being "taken care of" and that they aren't going to sue me.

I guess my point here is that it could be done. If bond underwriters would simply take a deep look at some niche markets (like software development) and innovate a little, they could create new demand for surety bonds.

You have many good arguments. Some of which I agree with others I do not.

1. Easily in your eyes, but not theirs. As previously stated, most bonding companies are comfortable with the business they write and are not looking to create a new surety department, especially when they are not familiar with the risk involved.

2. It is true the surety could inspect the code in an attempt to lower their risk. However, this is not a good solution to the problem. If the surety plans on charging reasonable rates and making a profit then they will not have much time or resources to read through code. If a software giant like Microsoft can't afford to produce 100% error free code then how would a surety have the resources to check for bugs?

It is a good idea if it could be made practicle. However, this is more of a problem of the software industry, as there aren't set programming standards, making the profession more of an art than a science at times.

3. I partially agree, as I feel bonding companies could do beter at year end if they thought a little more outside of the box. Unfortunately, guaranteeing software is not something that I think they should venture to just yet.

4. Precisely. You are now listing many of the traditional underwriting methods used for construction contracts. However, I personally believe this would not be enough to use for software underwriting. There are simply too many variables involved in software development.

5. You listed a scenario that often happens in traditional performance bond claims. Often if a contractor triggers a claim they will either rebid the contract or offer it to another contractor that the surety writes for. However, bond underwriting is not done with claims built in. In other words, they expect a 0% loss. Obviously this is impossible, but I wanted to make it clear that claims are a very bad thing, even if another company takes over the work.


It is true it could be done. I personally do not think it should be done until programming develops some more standards. If I were president of a surety, I would look at other untapped markets that can be underwritten more easily and more efficiently.

[BondLaw]

Very lively and entertaining discussion. My experience comes from inside the underwriting cube, and it has been my experience that the reasons why underwriters are generally reluctant to write bonds for technology contracts are pretty much as stated. But there have been a few who have taken a close look at the market potential for this type of business and have come to the conclusion that there isn’t much of a market. They find that when a technology company comes looking for a bond, it is usually required because the techies are working on orange crates in their garage and the buyer is more interested in testing their creditworthiness that their savvy.

Which brings me around to a couple of questions for Tachyon and, perhaps, the unregistered guest who started this whole thing. First, what are your actual bond needs? And who requires the bonds (government agencies private sector buyers)? Perhaps things have changed—it has been quite a few years since I was exposed to such a bond requirement—but my recollection was that technology providers were rarely asked to furnish bonds. Second, what do you believe the market potential is? Are there hundreds, perhaps thousands of well capitalized (read: qualified) companies in need of bonds? Third, what type of bonds are called for? A completion bond was mentioned, which sounds like the right instrument, but a couple I dealt with were looking for performance bonds, written on an AIA-312 no less.

[tachyon]

I don't even know what an AIA-312 is, so I'm hardly qualified to conduct a feasibility study. Software development projects are complex beasts and I believe the world would be a better place if there was a third party to these kinds of contracts that ensured that the other parties played by the rules and complete obligations to one another. However, I'm not certain that surety bonds are the only possible solution.

For example: Professional associations can hold programmers to higher quality standards. E&O insurance is helpful, too. Certifications and degrees are useful credentials. Having payments put in escrow during development removes other kinds of pitfalls.

Nobody has required any bond from me other than my word. But the question is: should customers than don't know me from Adam be this trusting? I suppose that's their only option because, according the the industry, bonding just isn't currently done for software projects.
Which came first: the chicken or the egg?

I'm thinking more along the lines of completion rather than performance bonds. (A performance bond is guaranteeing and extra level of quality assurance? Bug-free-ness?)

What exactly do you mean by "well capitalized" companies? I looked up the number of "Business Sales Leads" at infousa.com matching the following criteria:

SIC Code(s)
7371 - COMPUTER PROGRAMMING SERVICES
737102 - COMPUTER PROGRAMMING SERVICES

Location Employee Size
1 TO 499
Location Sales Volume
$500,000 TO $2.5 MILLION
Years In Business
5 TO 15+ YEARS
Credit Rating
A+ - 95 AND ABOVE
A - 90 TO 94
Type of Business-Additional Omit(s)
EXCLUDE FRANCHISE RECORDS

The result was only 1,334 companies. If I then include firms with "B" credit scores the number jumps to 6,054. Include "C" scored firms and it goes up to 10,654.

So, let's say that bonding companies offer software bonds at around 2% of contract value (on average) and if we also assume that an average bonded firm does $1M in business per year....

That's...WHOA...$213M in revenue (split by several bonding companies unless a monopoly exists.)

Of course, I know that some of these projects with fail miserably and eat up would-be profits. What do you think the default rate would be for computer programming contracts? What would the ultimate cost per-default be for the bonding companies? If half the obligations are met is the bonding company simply making up the differerence?