Using AI to breathe new life into the $120 billion deathcare industry

If we told you that AI could help you pre-plan your funeral and would write your letter to your loved ones when you pass, would you be upset, confused, put-off? Or would you be relieved? 3 years ago, you might not have believed it at all. But today, with everyone from middle schoolers to grandparents trying out ChatGPT and every software launching new AI-powered features, it’s certainly, at the very least, feasible. 

At Gitwit, almost any business that we’ve spoken with over the last 6 months has mentioned AI and is trying to “find applications for AI in their businesses”, because there is an increasing feeling that you have to — not only to get ahead, but to stay relevant. What we’ve found is that the truly powerful AI tools are not bolt-on features to an existing product or workflow, they are applications that have fundamentally changed, or even replaced, the tools that were once used or the way teams once worked. 

In other words, companies are often first looking to quickly apply a general large language model to an existing aspect of their business to create efficiencies either for themselves and/or their customers. What usually happens is that the hyped up potential of AI meets the brick wall of specificity. How do we apply it to our specific situation? This is the next question companies are asking.

The common mistake we have found is that companies start first with the technology and search for an opportunity, rather than starting with the nitty gritty details of a user experience or customer journey and identify what could and should be better – then use the lens of AI potential to see if something much better is now possible.  

Some of the greatest opportunities that we’ve found for AI are in the most difficult, human moments, where some form of automation seemed impossible — or even preposterous to involve a machine dealing with sensitive issues — just over a year ago. Let us tell you how our team found an opportunity in the deathcare industry and created an AI-driven product that is changing how funeral directors, customers, and families are talking about end-of-life wishes.  

‍‍

The Insight 

The Gitwit team built Prelude within our venture studio, 19days. As we were researching the funeral home space and the deathcare industry at large, we noticed an overall trend. Cremations are on the rise, leading to lower average costs of funerals.  As a result, funeral homes need to be more proactive in sales — relying on pre-planned funerals for a greater portion of their revenue. 

However, a pre-planning customer is much different than an at-need customer. This customer must be motivated now to create a plan and pay for something that they do not currently need and won’t even get to experience. This customer is pre-planning so that their family does not have to worry about arrangements when the time comes, and to ensure they get the funeral they prefer. It’s easy to put off when it doesn’t feel tangible or urgent.

In our research process, we found a subject matter expert (SME) (more on finding a SME here) who had an uncanny ability to sell pre-planned funerals — her close rate on pre-planning policies far exceeded the average. We brought her on to the Prelude team and dug into what made her approach different. One of the innovative things she did was create a letter to the family of the person who had pre-planned. 

After sitting down with someone and taking the time to walk through an incredibly thoughtful pre-planning experience, she would manually type up a letter for the person to provide to their family — who could deliver it either before their passing or after their death. This letter included both the tactical elements of the plan as well as the personal wishes and reasoning of the person that she’d pre-planned with. She would spend hours writing this up, but always found that the time investment was worth it.  After seeing this letter, people would almost always take steps to finalize their funeral policy — even after they had told her that they didn’t intend to do so.

The Business Problem 

Without the letter, the pre-planning process feels cold and transactional (our team pre-planned our own funerals with a handful of funeral homes to audit their processes). Without a strong why — without feeling the strong sense of empathy for close family members who will be left with the burden of planning in a moment of grief — customers are faced with something that they likely don’t want to think about and is probably more expensive than they considered. This meant that people tended to walk away without closing, put less money into their policy, and lapse in payment on pre-planned policies as time went on. 

Plus, with nothing more than a bill of sale for the transaction, there was often confusion amongst families when the death occurred for what had been purchased and paid for, and, most importantly, why or why not. Lack of clarity can create frustration that ends up creating a perception that the family is being taken advantage of by the funeral home.

The Consumer Challenge

In a typical pre-planning scenario the person leaves the meeting with a new insurance policy and a bill of sale. It’s on them to figure out what they’re going to communicate to whom and how. They may not realize the value of documenting the details of what they’ve planned and why. Or they may feel like a letter is a good idea but then they have to sit down with a blank page and figure out what to write. It’s an incredibly heavy topic emotionally, with a ton of details that they need to document accurately, and it’s something they’ve never had to do before. Plus, most people don’t love writing to begin with. 

The Technical Opportunity

In the letter, we found an extremely effective sales tactic that is also likely the most valuable part of the entire experience for consumers. Yet, the tactic would be incredibly difficult to replicate at scale across funeral homes (our SME spent hours writing each letter).

Without AI, the best solution would have been a mad-libs style plug & play template, either generated by hand by the funeral director or generated digitally. The end product would have been obviously generic or linguistically awkward. 

However, with modern AI tools like a large language model (LLM), we have the opportunity to essentially make the AI model an expert in what a great letter looks like and a person’s individual motivations, wishes, and emotions. It could be just as good as, and much faster than, our SME. 

Plus, our LLM is able to cross-reference a funeral home’s specific list of services and prices and automatically create a bill of sale. Again, this would have previously been a manual, painstaking process or something that needed to be coded and updated each time prices change. Now, it’s as simple as uploading a new price document to the LLM. 

How did we make AI feel human? We had to prompt engineer the humans first.

The Prelude guided pre-planning process begins with  a series of very thoughtful, empathetic questions designed to build an understanding of both the person and their wishes. As the funeral director or staff member takes the pre-planner through this journey, they select from specific options and also jot down free form notes to capture personalized details, additional wishes, and the pre-planner’s “why.” 

We had to think through the human journey that was needed in order to prompt people within the conversation in a natural way that also fuels the AI. 

Behind the scenes, AI utilizes these inputs along with the funeral home’s list of services and prices to create:

• An editable list of merchandise and services with pricing that aligns with the unique offerings of that funeral home. The selections from the list of merchandise and services then inform the content of the letter.
• A personalized letter that captures critical information in very human language.

Both are created right there, during the session. 

The Outcome 

By creating this sophisticated AI driven experience, we’ve made it more human for both the funeral director and the pre-planner. The funeral director doesn’t have to worry about finding pricing and calculating financials during the session. They can just be present, guiding the pre-planner through the questions, listening to their answers, and asking follow ups. The burden of ‘selling’ is removed from their role, and they are able to simply advise. 

The pre-planner is taken through a thoughtful conversation about their end-of-life wishes. It’s reflective and personal. Prelude guarantees that they are getting a best-in-class experience, no matter who their funeral planner is. Then, they leave not just with a bill of sale, but an emotional, personal letter to their family member. It feels like they, personally, have accomplished something meaningful during that time. 

At no point does the director or pre-planner feel like they are interacting with AI — which is even more important considering the average age of our end user. The technology is behind the scenes, yet carefully conducting the whole experience.

Your next AI opportunity

Your next AI opportunity doesn’t start with technology, it starts with an insight about a very human problem in your operations, process, or end user experience. The best way to start developing an AI tool, is not to look for a feature, it’s to find the people most familiar with these components (ops, processes, user experience) or work with a team that is expert in finding insights in these areas. Then, look for areas where there are opportunities where improving or scaling steps would create outsized value. Finally, think about whether AI can power these innovations, and, if possible, do so without the humans on either side knowing that they are working with AI. 

‍

‍

‍

‍

‍

‍