Avoiding Point Solution Proliferation in AI and Legal Tech

The False Dichotomy of Buy vs. Build

Legal technology buyers are often confronted with a dilemma of choice:

Buy it all: On the one hand, they can purchase many point solutions that solve specific problems. These point solutions are often offered as Software-as-a-Service (SaaS) products, making them easy to buy and start with.

Build it all: On the other hand, they can build their own solutions in-house. This approach allows them to tailor solutions to their specific needs and wants, but it can be costly and time-consuming to build and maintain.

The problem with the first approach is that it can lead to a proliferation of point solutions, each with its own cost of developing and maintaining. This can lead to a significant amount of technical debt, a high total cost of ownership, and a number of compliance and security headaches.

On the other hand, the problem with the second approach is that the task seems insurmountable - can you really build everything you need in-house?

What if there were a third option?

Buy 80% and build 20%: The third option is to buy a platform that solves 80% of your problems, reducing your focus and effort to the remaining 20% that is unique to your organization and each application. This hybrid approach allows you to focus on what matters most while still getting the benefits of buying an enterprise solution that solves most of your problems. Furthermore, it unlocks the ability to manage your data and applications in a single place, reducing the counterparty and compliance headache that inevitably emerges when you have dozens of vendors and point solutions.

The same 80%: In reality, many vendors are solving the same 80% of problems in their products, like building contract data models or parsing LEDES files. Each charges you for their own cost to build this functionality though. Do you really need to pay for the same thing over and over again?

In the world of point solutions, the answer is yes.

In the world of platforms, the answer is no.

Doing the Math

Let’s say you have 10 problems that need solving.

Under the buy it all approach, you would buy 10 point solutions, each solving one of your problems. Let’s say the cost for each vendor to develop these solutions is $10; therefore, each vendor will likely charge you at least $10 to acquire their product. This means you’ll need to budget at least $100 for these solutions.

Under the build it all approach, you would build 10 solutions in-house. Let’s assume that your cost to build is higher than the vendor - $15 each; therefore, you might think you need to budget $150 to build these solutions.

However, under the buy 80% and build 20% approach, you would buy a platform that solves 80% of each problem, reducing your focus and effort to the remaining 20% that is unique to each problem. Under this model, your marginal cost to build each solution is only 20% of $15, or $3. This means that your internal cost to build 10 solutions is only $30; as long as the platform costs less than $70 - or seven times each application - you’re better off buying the platform and building the remaining 20%. In reality, most platforms are available for less than 2-3x the cost of a single application, making this approach even more attractive.

Let’s summarize the costs of acquisition for each approach:

The Total Cost of Owning (Someone Else’s Product)

The above analysis strongly suggests the hybrid approach is the best option, but it doesn’t tell the whole story. In reality, the cost of buying a product is often even higher when one considers that the cost of ownership is not just the cost of the product itself, but also the cost of configuring, integrating, and customizing the product to your needs. This is especially true for point solutions in the legal industry, where many key systems of record are heavily customized or hosted on-premises or in locked-down cloud environments.

As a result, the total cost of ownership for a multitude of point solutions can be much higher than the cost of the products themselves. This is especially true when one considers the cost of integrating these systems together, which can be significant. In fact, many organizations spend more on implementation and integration than they do on the products themselves.

In contrast to buying a product designed to optimize vendor margins, buying a platform designed to fit your requirements can significantly reduce the cost of ownership. This is because the platform is designed to be configured, integrated, and customized to your needs from the start, reducing the cost of implementation and integration. Furthermore, the platform is designed to be flexible and extensible, allowing you to customize it to your needs as time goes on even if your needs differ from the priorities on the vendor’s roadmap.

Building your own product on top of an enterprise platform makes a lot more sense than buying a product designed for someone else’s needs.

The Total Benefit of Owning (Your Own Product)

The total benefit of owning your own product is not just the benefit of the product fit or the cost savings, but also the legal and business autonomy that comes with it. This is especially true for legal technology, where compliance and risk management considerations are paramount to the safe, secure, and efficient operation of the business.

Counterparty Risk

When you buy a product, you are taking on the counterparty risk of the vendor. This risk is especially high for Software-as-a-Service (SaaS) or cloud products in the legal industry, where the data entrusted to vendors is often business-critical, highly-sensitive, and subject to retention regulations. When there is a risk that a vendor may go out of business, be acquired, or change their product in a way that impacts these strict requirements, the risk of using a SaaS product can be material.

On the other hand, when you build your own product on a platform, you are in control of the data storage and processing and can ensure that your requirements are met. Organizations can still deploy their applications in their own public cloud environments, but they can also deploy them on-premises or in hybrid cloud environments, which are increasingly common in the legal industry. Some platforms like our Kelvin Legal Data OS even include traditional source code escrow rights, providing you with the ultimate insurance in the event that counterparty risk materializes.

Compliance Risk

When you buy a product, you are taking on the compliance risk of the vendor. This risk is especially high for global organizations, where the data entrusted to vendors is often subject to a wide variety of regulations and ethical obligations that vary by jurisdiction. Most vendors, however, deploy and operate their products in a single jurisdiction, which may not be the jurisdiction where your data should be stored or processed, resulting in a compliance risk.

For AI products in particular, the risk is even higher, as the AI model may be trained on data that is subject to strict regulations or ethical obligations. In the extreme case, the use of specific models or training data may be illegal in certain jurisdictions, which can prevent the use of the product altogether in some jurisdictions.

In contrast, when you build your own product on a platform, you are in control of the data storage and processing and can ensure that your requirements are met. Organizations can deploy their products where they need to and choose the AI models and training data that are appropriate for their needs and compliant with their requirements.

(AI) Technology Risk

When you buy a product, you are taking on the technology risk of the vendor. This risk is especially high for AI products in the legal industry, where technology is moving at a rapid pace and the state-of-the-art AI model may be obsolete in a matter of months. For vendors who often lock themselves into AI partnerships with specific vendors, this can result in a product that is obsolete within a matter of months.

By comparison, when you build your own product on a platform, you are in control of the AI models and training data. When new techniques or models are developed, you can easily swap them out to replace the old ones. When new data is available, you can easily retrain the models to ensure they are representative of your data and compliant with your requirements. While rules related to AI will likely continue to diverge across jurisdictions, a platform-based approach means that you can swap models in different jurisdictions yet still retain the majority of your functionality and user experience.

Embracing the Ensemble

Build it all and Buy it all are just two extremes of a continuous spectrum. The reality is that most mature organizations will end up somewhere in the middle. The key is to understand the ensemble benefits that platforms can provide by reducing the cost of ownership and time to develop for organizations that invest in their own systems.

In the legal industry, the ensemble benefits of platforms are especially important as they related to data protection and compliance. As a result, we believe that the future of legal technology will be built on platforms like our Kelvin Legal Data OS, which allow organizations to achieve quality and efficiency at scale while retaining the legal and business autonomy critical to their success.

Jillian Bommarito, CPA, CIPP/US/E

Jillian is a Co-Founding Partner at 273 Ventures, where she helps ensure that Kelvin is developed and implemented in a way that is secure and compliant.

Jillian is a Certified Public Accountant and a Certified Information Privacy Professional with specializations in the United States and Europe. She has over 15 years of experience in the legal and accounting industries.

Would you like to learn more about risk management for AI-enabled legal tools? Send your questions to Jillian by email or LinkedIn.