Door #1, Door #2, or Door #3? How do P&C Insurers Decide Where to Place their Bets on AI?
Last week, I wrapped up my 100th business day as President of Appulate and thought I would share some observations and insights I have garnered over the past few months. During this time, I have spoken with executives from roughly 50 of our current and prospective customers, including carriers, MGAs, wholesale brokerages, and retail agencies. I’ve noted several common themes throughout these conversations:
- Insurers are aggressively working to identify AI projects to improve their operations and send a message to their respective Board of Directors, investors, reinsurers, distribution partners, and policyholders that they are embracing AI and its potential benefits.
- There is no shortage of opportunities to leverage AI to attempt to make an impact on business operations. Whether streamlining new business submissions, enhancing underwriting success, or improving the efficiency and effectiveness of claims, policy servicing, billing, premium audit, or loss control teams, there are dozens of proven use cases where AI can be deployed to deliver a positive impact for insurers.
- Insurers’ boards, investors, and business partners want to see, not just an investment in AI, but actual results, delivered promptly. Given the plethora of AI opportunities facing P&C executive teams, one of the biggest challenges is selecting the right AI initiatives to prioritize.
- Minimize Risk / Maximize Return. Okay, I get it. Minimizing risk and maximizing return on any IT project is easier said than done. However, with careful research and a methodical ROI approach, insurers can increase the likelihood of selecting AI initiatives that deliver solid impact with a reasonable investment while minimizing the risk of runaway projects.
- Are you beginning by identifying a business problem, or have you simply identified a technology that your competitors have deployed? — You see where I am going here. Don’t jump on a technology bandwagon just because you have seen press releases from your competitors that they have deployed some new tech. Make sure that you actually have a business issue to solve, and then begin your evaluation of options.
- Are there metrics that can be used to evaluate ROI on the aforementioned business problem(s)? If so, then great! Now you can study the likely outcome(s) of your AI initiatives to assess what your expected business impact should be, the probability of success, and how long it will take to begin seeing results. Whether partnering with an AI solution provider or developing something in-house, make sure you feel truly comfortable with the impact metrics, or it could be difficult to evaluate success.
- Have you selected AI initiatives that can be time and cost-boxed? Do your internal or external AI partners have legitimate proof points that can be evaluated to understand how long the solution will take to deploy? Can they truly give accurate estimates on project costs? If not, tread very carefully here, as what initially looks like a 2–3-month project can easily morph into 12-18 months. It’s no fun going to your board/investors and explaining that the project will have cost overruns and/or timeline overages of 2x – 3x!!
- Is there Executive Buy-in on Adoption & Change Management? This happens all too often with insurers. Someone sees a “shiny new penny” of a project and sells it to the executive team, only to watch the project wither on the vine from poor business user adoption and/or ineffective executive management oversight and review. Each year, companies spend millions of dollars on IT projects that either (a) are implemented without ensuring user adoption, or (b) never go live because users are too busy to effectively evaluate and test the solutions before moving to production. For an AI project to succeed, you need clear, ongoing executive support - leaders who actively reinforce priorities, push adoption, and are engaged throughout the process from implementation and into production.
The AI Project Holy Grail – Producing and Proving Positive ROI in Months, not Years.
So, you’ve been inundated with sales calls, online articles, webinars, and case studies describing the “next best thing” in AI investment. Where do you begin when narrowing down dozens of potential AI investment options to just a few projects to focus on this fiscal year?
Let’s start with a basic set of questions designed to evaluate the right AI projects for your team. This approach attempts to consider a holistic view of AI ROI estimation — from project selection through deployment and production:
Multiple studies show that the success rate of AI projects hovers between 30% to 40%, which is a scary statistic. The reality is that most unsuccessful AI projects fail quietly, not with a loud explosion, but with a slow drift towards irrelevance. There is no silver bullet for picking the right AI initiatives or for ensuring their success. However, Business, IT, and Innovation Leaders can improve their odds of a positive return on these investments by keeping a handful of guiding principles in mind:
- Begin with a business problem, rather than a new tech trend
- Define ROI metrics before you start
- Find ways to put a box around scope, time, and cost
- Ensure executive oversight and effective change management to avoid project “decay”
My advice? Start your AI project wish list with proven solution partners who have been doing similar projects for years and can point to a track record of effective deployments that finish on time and on budget, delivering strong impact. Yes, my bias for Appulate’s 20 years of experience is showing, but insurers can reduce risk for their AI initiatives by following the steps above and selecting business partners with the right track record.
Wishing you all the best in ’26 and beyond with your AI optimization projects!
READ NEXT:
.png)
