Project Description

authorAPPROVE and CheckThisContract are part of the author product suite by software developer McCarthyFinch/Onit. It’s an advanced legal AI email and web application that reviews, amends and approves contracts for business users.

Industry: Legal Software Development
Product team:  Myself on UX/UI + Graphic Design, 1 x Legal Tech, 1 x Product Owner, 1x Product Designer

What was developed

authorAPPROVE enables business users to self-service contract reviews in minutes. Users send a contract to approve via email or a web interface and their companies corporate standards are automatically applied to the document, helping manage legal risk quickly.

Over the course of the app’s development, I was responsible in tandem with our product team for identifying key user concerns, workflows, and personas. I was also responsible for wireframing, prototyping, user testing, and taking the app to the final product. Inclusive of this was the development of a full support portal with documentation and user tutorials.

The problem being solved

Most businesses are reliant on their in-house legal teams, or external legal counsel when it comes to contract review and approvals. This can result in lost revenue due to slow turnaround times caused by already overtasked legal teams.  This backlog also leads to a large volume of low-end contracts being signed without ever being viewed by legal counsel, increasing risk for businesses.

Our challenge was to create a low-touch product that would allow contracts to be quickly and accurately reviewed, redlined (marked up with advice), and edited. Effectively allowing business users to confidently turn around contracts within minutes.

UX research + artifacts

UX was an extremely important part of the development process in order to ensure adoption of the product.

To aid this the product team and I:

  • Interviewed business users and lawyers across firms and in-house to understand common practices and workflows.
  • Developed user personas and use cases
  • Performed competitive analysis of other products in the market.
  • User tested and iterated with clickable prototypes developed in InVision.

As the product went live we performed:

  • User testing
  • User interviews
  • Studies to gauge the effectiveness of time saved and productivity increases.
  • Implemented iterative improvements and updates to the product over time.

Designing for trust with Business Users

The user profile for authorAPPROVE is quite different from our product authorDOCS. AuthorDOCS required an ‘open-box that allowed lawyers to pop the hood and understand how advice was being provided. authorAPPROVE users have the exact opposite profile. Business users don’t want to be sucked into the minutia of understanding legal context. They simply want to know if they can or cannot sign a contract.

Initially, we designed authorAPPROVE with a companion web app that allowed users to jump through their document review and edited changes. This app provided context and advice around the changes that had been made.

Initially, the idea of the companion app tested well with users. But once implemented we found that most users were overwhelmed by the need to ‘understand’ contract changes. In reality, they simply wanted to know ‘do I sign or don’t I?‘.

This lead to the retirement of the companion app. Which we replaced with a more comprehensive email that business users received along with their reviewed contract. The emails provide a summarised overview of changes along with the thumbs-up approval to sign or advice to seek further counsel from the legal team.

Designing for trust with Lawyers

While authorAPPROVE’s primary users are Business Users the app exists in service of an in-house legal team or external law firm. To this end legal users needed to feel comfortable the AI was providing correct legal advice.

To aid this concern we developed a review builder that allowed legal users to select the contract elements they wanted the AI to look for then provide rules and customized advice around each element that was found. Thus alleviating the concern that the incorrect advice would be provided.

Follow Up Study

We engaged several clients as beta users over a trial period to gauge how much time was saved on low-risk contract approvals. The outcome of the beta trials showed that contract approvals were sped up within the organizations by between 60% and 70%.

Conclusion

In conclusion, we identified that authorAPPROVE business users are far happier to trust the AI outcomes than our more legally accomplished authorDOCS users. In the case of this app we found that less information and less user input resulted in greater trust and adoption.