Apply HN: Jury Board – helping attorneys win cases through jury selection
Problem: Scientific jury selection is limited to the largest firms usually making over hundreds of millions of dollars a year. When selecting juries, attorneys of other, smaller, firms have only the information they get through asking questions to make their selection. This can be greatly improved by providing attorneys with a tool that allows them to do data mining from social media and any other sources and aggregate it to a probability of how that juror is likely to vote.
A tool that gives an attorney more insight into jurors would be incredibly valuable to them. If you have millions on the line, paying even a few hundred to get information about a potential jury is a bargain.
While giving one attorney a strong advantage, thereby effecting the outcome of a case, can seem questionable, the justice system designed jury selection to be in the lose/lose quadrant of possible scenarios. Ideally a more equitable and fair jury is chosen by both attorneys eliminating those against them as best they can. This also means though should one attorney have an effective tool selecting jurors others are highly incentivized to use that same tool. This should produce a strong network effect for my product.
31 comments
[ 4.9 ms ] story [ 29.0 ms ] threadThis would require knowledge of the case and case law. Which would then require a full AI.
A probability based off a set of given facts and not every fact is still useful and doesn't require AI.
To fully replace a jury consultant, yes that would require AI, but at their going rate of >$250/hr I think an app could find a price point reflecting the diminished result that is still very attractive in comparison.
1. If you're already going to trial, especially with "millions on the line", paying out a few thousand / tens of thousands to a consultant doesn't necessarily seem to bad.
2. Given how much jury tuning is going on right now, I would have worries about the quality of the "actual data" and how much insight you could get from such a contaminated corpus.
3. Prosecutors already have an incredible amount of power in jury selection, why do they need more?
2. The biogdaphic/demographic data is all from forms and attorneys would have no reason to lie to their own device they are using to reference juror information from. I think most of the information would be reliable.
3. The defense cab use it too ;). But seriously, the juror selection process is designed for each attorney to be able to get rid of anyone they think is biased against then with the hope that by both attorneys doing this those left create a fair hury. Is this optimal? No. Would my tool make a suboptimal solution better? Yes I believe it will.
I think it is the programmer optimization drive in us that makes this seem crappy that it isn't being completely overhauled since it isn't optimal, but my solution does help.
Let's say the juror is "Joe Smith from Alabama". What is the probability that:
1. he has a twitter account
2. he provided his full name in the account
3. Which of the following accounts belong to that particular Joe Smith: https://twitter.com/search?f=users&q=joe%20smith ?
What other "social media" data can you provide? Facebook? (it's mostly private and not available for low-grade stalking). Tinder?
What is your addressable market? How many jury selections there are in US where "millions are on the line"? And if there really are "millions on the line", can't they afford $250/hr jury consultant?
How did you validate your idea? How many attorneys at small companies did you talk to? What is the "incredible value" of attorney knowing twitter posts of a potential juror?
A sample of information, just what I remember, I'm not looking at the list.
Name
age
education
limited previous legal experiences
Address
Employer and length of time with employer
members of household with relation and age
Criminal history, down to parking records, doesn't come with the juror info sheet, but is usually easily accessible by attorneys.
Even simple things that can be grabbed from facebook/twitter/anything would be useful. Often the halo effect or a person trying to get out of jury duty cause some degree of deception. Anything that can verify what they said in court is valuable. A jury could consist of as few as 6 people, finding that one of them is biased against you and you can strike them is great. The smallest thing can mean a great deal. Say under their favorite books they list civil disobedience, you could easily strike them from a case where one side may be some sort of conscientious objector.
True, if there are millions on the line the attorney would probably hire a consultant, but in that case I think a consultants job would still be easier with my tool than without it and should result in a sale for me, just from the consultant instead of the attorney.
As far as market, not much like this exists, but it is being asked for. There are roughly 80,000 trials per year in the US, a prosecution and defense for each, so 160,000 opportunities. This also makes it such that there are 800,000 potential jurors per year. Some pricing structures I think are reasonable could result in a revenue of under one billion per year.
I'll be honest I've not talked with many people, > 10, but the reactions I received from those I've talked to has convinced me. I've talked to a few civil and criminal lawyers, both public servants and private firms and the feedback is usually ecstatic. "incredible value" might be my words, but "invaluable" was what one lawyer kept saying. Further validation comes from a rudimentary and necessary building block of this, note taking electronically during jury selection, is immature to the point it almost doesn't exist was something I was approached by attorneys to create. The data mining and giving suggestions just seemed like a natural evolution, which so far has been received well.
I know I won't be able to provide everything on every potential juror, but if I'm consistently striking one I'm showing value. And with a pool of > 200 people I'm sure there's someone that I can dig up dirt on.
Also, I've not really explored this avenue much, but any information brought up in this investigation or given to attorneys becomes a matter of public record. It seems like there would be a great deal you could do with that, but I'm not really into the personal data selling thing.
Realistically, you might need to create a MVP with lots of manual processing to test the market. Or, find a VC who will believe your projections.
And this is just the States, there are other countries that use similar systems for jury selection.
I agree I won't make that until there is a mature product, but the question was addressable market, not expected year 1 revenue.
The mvp I'm working on now allows the sharing of information about jurors. This should allow me to gather data on how jurors vote tat can be marketed while still making profit off the sharing app alone.
Since it is such a huge potential market, it might be a good idea to work on getting VC funding. The only caveat being, avoid "free" anything. Very hard to convert to paying down the track. You could give early users the option of investing and getting a chunk of discounted juror research for their money as well as equity. Since you are already in contact with lawyers you should be able to find some professional advice, etc.
On the flip side, where I'm located the 120k from yc seed money would be a year's runway for me and a cofounder or two. That and there is more talent here than people think, or at least more good devs looking for work than hiring.
If anyone has some contacts or VCs want to get in contact with me take my username at Gmail to talk.
Request the Startup Ecosystem Report from here:
http://startup-ecosystem.compass.co/ser2015/
Great research. Interesting notes about Chicago in there.
It's hard to believe this wouldn't end up mostly used by prosecutors.
Like I said though, with courtrooms being adversarial one side is incentivized to use any useful tool against the other. If the defense can't afford it and therefore doesn't use it, it becomes less likely the prosecution will use it then too. Maybe to counteract this I have to subsidize the cost of those using it for defense.
It is far to early to say for certain, but in order to generate better juror recommendations data is needed, I want as many people using it as possible to get data.
That's exceptionally unlikely, I think. The prosecution's job is to win, whatever the cost--that appears to be the current mode of operation in the US, anyways.
Having people pay what they will or can isn't a new idea. If the defense can't pay an individual license may be cheap, while an office license could be exorbitant if the DA will pay through the nose.
The main point of the last post is that I gain nothing if they don't use it, but if anyone is paying for it I still gain from other free users.
As I mentioned the I think system is suboptimal, I think this can improve it though. I'm not a jerk who would screw the Justice system for a buck, I really do think this can improve things. I don't think the possibility that it might hurt is reason enough not to try.
Technically, you would be doing a lot of site scraping and that would take a sizeable investment in development. Of course, assuming that you don't first need to enter the details from hardcopy. Even if you get the information in electronic form, it will probably be different for different courts, states, etc.
It is a remnant of the medieval legal distinction between nobles and commoners. It simply means that nobles can only be tried by a jury of other nobles, and commoners can only be tried by a jury of other commoners.
Attorneys/secretaries/paralegals already do this by hand. More than scraping I'm hoping to do something useful with what they already scrape manually. By giving them a platform that make it easy to share their notes and info about jurors I should be able to start using this information.
Why should a lawyer trust the outcome of an important case with "millions on the line" to an unknown startup rather than do what they do now to select juries (go with their gut, whatever)?
My MVP and what would likely become a free version is an app that allows for the sharing of information about jurors and taking notes about jurors during selection. The level of trust required for this is low, just don't delete their info and your good. But while using the free version I imagine that if you had a juror the attorney rated favorably and you had evidence they were an undeniable liability and you showed this to the attorney it would prove the paid tools worth and hopefully encourage them to pay to research the other potential jurors after seeing my tool prevent such a blunder.
Even if it had a low conversion rate the free version still makes sense, more users means more data to draw conclusions to give to those that are paying.
"I have a complex case, I need smart jurors" or "my case is based off emotions, not facts, so I need emotional jurors" are common thoughts. Because of this any insight into the potential juror is useful.
There is a great deal of rules of thumb. Strategies passed from attorney to attorney, but little verified knowledge. That is what I'm hoping I can provide.
There are sites you can download this information from now. It is freely available, but not freely distributed by the government. Somehow a file server costs ten cents a page served, but you can access it all.