Evaluating the ROI of policy review automation (Part 2)
In this second installment, Cortical.io’s CMO does a deep dive into the policy review automation ROI equation.
Editor’s Note: Part one of this series can be found here.
One model for evaluating the ROI of a business process improvement is to determine the cost of the current process and calculate the potential direct savings and attempt to quantify, if possible, the indirect benefits. The ROI of policy review automation on an annual basis can then be calculated with the following formula:
Direct savings
To calculate the costs of your current policy review process, you need first to answer these three questions:
- How many documents are reviewed on an annual basis?
- On average, how much time does your quote specialist spend reviewing a document?
- What is the average hourly cost of a quote specialist?
The average time for processing a document depends on the size of the document, but also on the number and difficulty of extractions. You may want to create several categories of policies, depending on the difficulty of reviewing them and calculate the ROI for the different policy categories.
Take an insurance company that processes 50,000 Tier 1 policies per year — Tier 1 policies being the easiest documents to review. Each Tier 1 policy requires on average 15 minutes of manual work by a quote specialist whose fully loaded hourly rate is $50. That brings the yearly costs of manually processing these documents to:
Depending on the use case, customers typically save between 30% and 60% of the time required to review and analyze documents. In the above example, that means that the direct savings range between:
In this example, it means that the implementation of an automation solution will enable the company to pull the equivalent of two to four full-time quote specialists out of the policy review process and reassign them to more sophisticated tasks, making better use of their skills.
Indirect benefits
Indirect benefits are much harder to quantify but should definitely be considered when evaluating the ROI of policy review automation. Here is a shor tlist of the indirect benefits customers usually see:
- Reduced company risk: Because it is more accurate than manual labor, policy review automation greatly reduces the risk of missing important clauses or misinterpreting provisions. Quote specialists may overlook two to three provisions for a given number of plans. Each overlooked provision has a cost associated with it (higher risk for the company, negative impact on profitability, etc.).
- Shorter sales cycle: Because automation accelerates the policy review process, quote specialists can process the quotes quicker. An insurance company will be able to submit quotes much faster and consequently close more deals. Note that the faster response time enabled by automation also positively impacts customer satisfaction.
- Improved pricing strategy: Accrued accuracy indirectly affects the bottom line because it enables a better pricing strategy. Human accuracy is generally estimated at around 80%, while a good review automation tool reaches about 90% accuracy. For an insurance company, this translates in more accurate quotes and therefore either better margins or more competitive quotes, resulting in a higher probability to close a deal.
- Easier knowledge transfer: When employees move up the ranks, leave or retire, a document review solution with trained models helps transfer knowledge about extractions and how to interpret them to new employees.
How can these indirect benefits be quantified?
According to Nucleus Research, quantifying indirect returns requires a structured approach and educated assumptions, testing the validity of estimates against a sample population and projecting a best-case and a worst-case scenario. Based on the case studies the research organization has conducted, Nucleus estimates that, on average, indirect benefits account for half of technology ROI. To be conservative, we will assume that indirect benefits are only 1/3 of total benefits or 50% of direct benefits in our ROI calculation.
Let’s return to our example company that processes 50,000 policies every year. In our lower-end scenario (30%), the direct savings were estimated at $187.500, while the upper-end scenario brought the estimated direct savings to $375,000. Applying Nucleus’ assumption to these figures gives us the following estimated indirect benefits on an annual basis:
Cost of policy review automation
There are different pricing models for document review solutions, either document-or user-based. Infrastructure cost, while minimal, should also be considered. You will also need to take the model training time and cost, which may vary with each solution, into consideration. This is a one-time expense and should be amortized over the life of the project (e.g., three years).
Let’s assume the annual subscription for an automated review solution is $230,000 for 50,000 documents per year and requires $20,000 for initial training and consulting (to be amortized over 3 years).
Total annual costs of policy review automation:
Justifying the implementation of new solution requires an evaluation of the ROI. With the direct savings, indirect benefits and cost of automation, the ROI of the solution can be estimated based on business process improvement.
Let’s go back now to our original ROI formula and integrate all the variables we obtained:
For our company processing 50,000 policies per year, this means a yearly ROI ranging between:
When calculating the ROI of a document automation solution, it is essential to consider both the direct savings induced by automation and the multiple indirect benefits such as increased efficiency and shorter sales cycle. In fact, experts acknowledge that software that improves productivity generates a higher proportion of indirect benefits than supply chain software, for example.
It is also important to be aware that the ROI of a policy review solution greatly depends on the use case and company environment. The type and number of policies that must be analyzed on an annual basis, their level of complexity, but also the labor costs associated with the manual review, the degree of experience of your quote specialist, will all affect the results.
Steve Levine is CMO of Cortical.io. He brings extensive marketing and sales experience to his role. Most recently, he led marketing for Civic Connect, a GovTech startup. He has consulted for a number of cybersecurity companies including Flashpoint, RiskSense, Qualys & Panda Security. Previously, Steve was CMO at publicly traded Edgar-Online and financial services startup UB matrix. Steve has held vice president of marketing positions at Oracle, Cassatt, Ketera and Arcot. While at Oracle, he led the company’s first global e-commerce marketing campaign. Steve also brings a sales perspective having held business development and sales roles at Tektronix and ParcPlace Systems. Steve has a B.S. in computer science from Southern Methodist University. He can be reached at s.levine@cortical.io and on LinkedIn.
Opinions expressed here are the author’s own.
Related: