You might have seen last Friday’s post about Avvo Legal Forms (here). It prompted enough comment that I thought some follow-up was in order.
In my post I was categorical that Avvo Legal Forms is, as I put it in the title, a real stinker. That prompted a response from Avvo’s CEO, Mark Britton, as reported by Bob Ambrogi here.
But Britton’s remarks to Bob address primarily what Avvo hopes to achieve with Avvo Legal Forms. It all sounds plausible enough, depending on how you feel about consumers creating contracts, but it has nothing to do with the quality of their forms.
Here’s what Bob’s piece says regarding quality:
Britton says that Avvo hired lawyers to create the base forms and to be sure that they are compliant in all states. It is unfair, he suggests, to compare these forms to ones created by a lawyer.
“You cannot compare a bespoke product from a lawyer that will cost you thousands of dollars to a product that is an entry-level product designed for people who are doing everything they can to avoid a lawyer,” he says. “Let’s get them that product and then start the conversation from there.”
The first paragraph is unintentionally amusing: We hired lawyers to create these forms—don’t compare them to ones created by lawyers!
That aside, Britton’s point—that it’s unfair to compare Avvo’s forms with bespoke work—was also made by Josh King, Avvo’s chief legal officer, in a comment to my post, here. I have two points to make in response.
A Failure of Quality Control
First, nothing in my post compares Avvo’s nondisclosure agreement with bespoke work. Instead, I consider their product on its own terms. The implications of allowing consumers to generate their own contracts are certainly worth discussing, but that’s not what my post was about.
Instead, I detailed how as a foundation for automated contract creation, Avvo’s nondisclosure agreement is of head-scratchingly poor quality. There’s the too-bare-bones technology that precludes meaningful customization, the glitchy questionnaire, the English-but-not-English wording of the questionnaire, the problematic substance, the bloated prose.
Among the tweets launched in response to my post, one of my favorites is this one by Bob Ambrogi:
@mikewhelanjr @ronfriedmann @KonciseD @Mark_Britton No. They'll add more forms & features. But they don't agree that the forms so far suck
— Bob Ambrogi (@bobambrogi) April 8, 2016
In alluding to a do-they-suck standard (originated by someone else), this tweet cheerfully surpasses the plain speaking on display in Friday’s post. I suggest that anyone who thinks Avvo’s nondisclosure agreement doesn’t meet that standard isn’t paying attention.
A Failure of Imagination
But what’s more perturbing about Avvo’s response is their aspirations for Avvo Legal Forms, or rather their lack of aspirations.
When used in connection with preparing contracts, the word “bespoke” is a euphemism for inefficiency. It’s standard practice for lawyers to prepare a new contract by retrieving a precedent contract or, less often, a template and making changes to reflect the new transaction. There are three drawbacks to that process.
First, thanks to generations of copying and pasting without the benefit of guidelines or real training, whatever is copied to create a new contract is itself likely a showcase for the dysfunction that is traditional contract drafting.
Second, revising a contract used in one transaction to fit another transaction is too often a haphazard process, as it’s difficult to climb back up the decision tree. You might end up incorporating deal points and concessions that have no bearing on the new transaction, and it’s all too easy not to address issues that aren’t reflected in whatever you’re copying.
And third, the process of revising a template or precedent contract to reflect a new transaction is relatively slow.
Automated contract creation allows you to escape these constraints. Because you’re creating templates to be used by many, you can take advantage of economies of scale: You can invest in technology that allows users to customize templates to reflect the key alternative scenarios. You can enlist subject-matter experts. And you can make sure that your templates say clearly and concisely whatever they need to say—an attribute that consumers in particular would presumably appreciate.
With Avvo Legal Forms, Avvo has made no attempt to do any of this. That’s what prompted my review and determined its tone. I’m inured to dysfunction in traditional contract drafting. I know what causes it, and I face it every day, so for the most part I no longer get worked up about it. But I find aggravating the idea of using automated contract creation not to effect change but instead to enshrine dysfunction.
In their response to my post, Avvo in effect threw up their hands and pleaded that they couldn’t possibly compete with bespoke law-firm work. So Avvo considers the traditional process for creating contracts to be an unattainable ideal rather than the source of dysfunction that it actually is. It’s surprising that a company that presumably sees itself as part of the new legal economy should set the bar so low.
Getting serious about automated contract creation doesn’t have to be expensive. But that might not be feasible if, like Avvo, you plan on allowing people to use it for free or if, like Avvo, you plan on putting 200 templates—yes, 200—on your system in less than a year. I suggest that when it comes to automated contract creation, you do it properly or you don’t do it at all.
I leave the final word to Marc Lauritsen (@marclauritsen), who has way more experience than me in automated contract creation. Here’s something Mark tweeted last week regarding Avvo Legal Forms:
Sad that a major player with plenty of resources would enter the field with such a weak offering https://t.co/Gh9bjR2NqR
— Marc Lauritsen (@marclauritsen) April 8, 2016
(If you want someone else’s take on all this, I suggest you check out this post by Casey Flaherty.)
5 thoughts on “Revisiting Avvo Legal Forms, Part 1: Failure of Quality Control and Failure of Imagination”
For the record, it wasn’t me who raised the “do they suck” standard. I was responding to the tweet by @mikewhelanjr in which he said that Avvo was using a “we know it sucks, but…” argument.
Yes, I know. I tried to signal that with my verb choice, but that was probably too subtle. If you’d like me to change it, let me know.
My organization is the dev for A2J Author which has been used to automate hundreds of legal forms for SRLs by legal aid attorneys.
(1) The interview process/questions can serve as a “legal education moment” for SRLs. Without a lawyer, they ARE the lawyer and should want to learn (Why do you want this information? What if I answer “No”?, etc.).
(2) Court forms are temporary holding pens for data that eventually ends up in a court management system. We used paper forms and kept them short and condensed to make the transfer of that information as efficient as possible. That compactness led to brevity and complexity for the uninitiated – everything looks like its same level of importance on a single sheet of paper, but a single checkbox can change your life. Online form filler-outers can improve on that (hopefully), by using the software to say “are you sure?” and “have you considered?”.
(3) Legalese can be automated out (a little) using code that tests for readability. WriteClearly.org is an attempt at that in my space and we include Flesch–Kincaid scores for all text lawyers use in their A2J Author dialogues. It isn’t perfect, but it’s a start.
Finally, take a look at the NYCourts User Testimonials from people who used the A2J Author forms to deal with their legal issue. https://www.nycourts.gov/ip/nya2j/diytestimonials.shtml
In my opinion, it’s particularly silly to claim that an automated confidentiality agreement can’t compete with a “bespoke” one, since the small number of variables that affect most confidentiality agreements dictates that a well-thought-out template should be better than a single lawyer creating something specific. Having a lawyer customize a confidentiality agreement should be valuable ony if there’s some unusual aspect to what the parties intend to do that a template can’t capture; in reality, what a lawyer drafting a confidentiality agreement usually does is mostly word processing, likely introducing and perpetuating errors. Indeed, since many lawyers spend relatively little time thinking about confidentiality agreements, the quality of even purportedly “bespoke” ones is generally more dependent on how good the base document was than the skill of the drafter. If you don’t take the opportunity to clean up the language and eliminate fundamental misapprehensions like “in the public domain” from a template (which could be done for a minimal investment) you’re surrendering at least half of the potential benefit of automation.
Thank you for the sentiment that baking disfunction into new services is not the direction in which “legal tech” should move. So much being passed off as ‘innovation’ is nothing more than backwards iteration, leaving lawyers running to stand still. I remain confident, nonetheless, that more lawyers will come around to the idea that the law itself is due an upgrade. That realization will stimulate us all to take more creative advantage of what code-based approaches have to offer. Maybe we have no choice: the advent of blockchain-based distributed agreements not only clarifies the concept that law is social media, but demonstrates why the days of static, paper-based contractual models are numbered.