Free Can Cost You Trust: An AI Companion for Seniors Is Caring, but It Makes Money Selling Supplements
The Case: An AI companion app for seniors. It's completely free, so where does the money come from?
It was an accelerator selection review, and an AI startup took the stage. The founder, Simon (a pseudonym), has spent years in AI, and his partners are all skilled with language models. They built an AI companion app for seniors: seniors just speak, and the AI chats with them, reminds them to take their medicine, and observes changes in memory and mood from their conversations, turning them into an easy-to-read report for adult children working in other cities.
The technology really was impressive. The all-voice interface works even for seniors who can't type, and the AI doesn't just chat; it can explain why it thinks Mom has been more forgetful lately, so it isn't a black box. The judges kept nodding.
Then Simon turned to the business model slide: "Our app is completely free. We'll build a large base of active families first, then make money through targeted ads, affiliate commissions and brand sponsorships."
He showed a huge SOM chart to illustrate how big the silver economy is.
When it was my turn, I asked: "Suppose one day an ad for a health supplement pops up on your mother's phone because the app 'heard' that her memory has been slipping. Would you feel cared for, or betrayed?"
The room went quiet for a few seconds.
📘 Startup Glossary | TAM / SAM / SOM TAM is the size of the whole market, SAM is the market your product can actually serve, and SOM is the share you can realistically capture within a few years. A big TAM doesn't earn points with judges; they want to know how you'll win the SOM.
The triangle of conflict between users, payers and advertisers
That day I wrote in my review: "The completely free strategy relies entirely on monetizing back-end behavioral data, making the business model highly sensitive to regulation; the conflict between advertising and the nature of caregiving also warrants further consideration."
There are actually three roles in this business model, and what they want conflicts:
Seniors are the users: they want companionship and peace of mind, and don't know their words become ad targeting tags
Adult children are the decision-makers: they install the app for their parents because they trust it will look after their family
Advertisers are the ones paying: they want precise lists, and the more they know about who "needs" supplements, the better
When the advertiser is the one paying, the product quietly tilts toward the advertiser. And in a field like caregiving, where the bar for trust is extremely high, one uncomfortable ad is enough for the children to delete the app, and tell their relatives and friends too.
📘 Startup Glossary | User vs. Customer The User is the person who uses the product every day, the Customer is the person who pays, and they are often not the same person. One of the first questions a proposal must answer is: who pays, and why are they willing to?
Add two more realities:
First, data related to memory and health is extremely sensitive. Inferring a senior's physical and mental state from conversations may fall under special rules for health-related data in personal data protection law. Using that data for advertising carries high legal and ethical risk.
Second, "free" is not an advantage. Governments and nonprofits already offer free outreach and long-term care services. When everyone is free, free is no reason to choose you.
📘 Startup Glossary | Freemium A model where basic features are free and advanced features are paid. It's different from "completely free and ad-funded": in Freemium, the money comes from users deciding it's worth it themselves, not from selling users to a third party.
If I were them: from "selling data" to "selling trust"
If I were Simon, I wouldn't give up on the technology, but I would rethink the business model from scratch.
Step 1: Find out who is truly willing to pay.
Leave the lab and interview three kinds of stakeholders: adult children working in other cities, managers of community care centers, and care supervisors at long-term care facilities. Ask how they care for seniors now, what worries them most, and what they would pay for.
📘 Startup Glossary | Stakeholder Everyone who affects your product or is affected by it: users, payers, decision-makers, partners, even regulators. The most common startup mistake is interviewing only users and forgetting the people who actually pay and make the call.
📘 Startup Glossary | Customer Development Steve Blank's method: before building a lot of product, get out of the building and use interviews to verify that customers really have the problem and are willing to pay to solve it.
📘 Startup Glossary | Willingness to Pay The highest price a customer is willing to pay for a given value. In interviews, don't ask "do you think this is good?" Ask "how much do you spend solving this problem today?"
Step 2: Change who pays.
Instead of charging advertisers, sell the product to care facilities or community centers: seniors use it for free, and facilities pay for care reports that help them spot people who need attention early. That way, the people paying stand on the same side as the seniors.
📘 Startup Glossary | B2B2C Delivering a product to end users (C) through a business or institution (B) that pays for it. In fields like education, caregiving and healthcare, B2B2C is often more workable than charging consumers directly.
What if facilities still won't pay? Step back: run a willingness-to-pay test first.
You don't need to wait until the product is finished. Build a simple plan page that clearly states "advanced reports for adult children, monthly subscription price," and see how many people leave their details or book a trial. Even if few people click, you get a real market signal, which is far more solid than assuming ad revenue in your proposal.
📘 Startup Glossary | Smoke Test Testing market response with a landing page, pre-orders or a booking form before the product is ready. It validates not the technology, but whether people really want it and will pay.
Step 3: Turn privacy into a selling point.
In caregiving, "we don't sell your data" is itself a competitive advantage. Data minimization, clear consent explanations, and letting children delete data at any time can all be built into the product design and the marketing message.
📘 Startup Glossary | Privacy by Design Building privacy protection into the product from the very start of design, rather than patching it in after launch. For startups handling sensitive data, this is the foundation of earning trust.
📘 Startup Glossary | Moat An advantage competitors find hard to copy, such as patents, unique data, brand trust or network effects. In caregiving, being trusted is itself a deep moat.
Jesse's Take: Free is not a business model. Trust is.
I admire the technology of teams like this, and I believe AI really can help seniors get better care. But I have to say it plainly: "Free" is only a growth tactic (Scale & Grow), not a business model.
When your users are vulnerable groups like seniors, children or patients, your most valuable asset isn't data, it's the family's trust. Trading it for ad revenue means swapping your most valuable asset for your cheapest income.
Two reminders for founders writing proposals:
First, write in one sentence who pays and why. If you can't write that sentence, even the biggest market size chart is just an imaginary picture.
Second, the compliance cost of sensitive data belongs in the budget, not in a footnote. Legal advice, security and consent mechanisms all cost money. When judges see you've budgeted for them, they know you're serious about building this business.
Technology can make judges' eyes light up. A business model is what makes them comfortable recommending you.
Startup Glossary Summary
Term | In one sentence | What you've proven |
TAM / SAM / SOM | Total market, serviceable market, obtainable market | You know which slice you can win |
User vs. Customer | The person who uses it and the person who pays | You know who will pay |
Stakeholder | People who affect or are affected by the product | You've heard from every key role |
Freemium | Free basics, paid upgrades | Users themselves think it's worth it |
Customer Development | Validate customer problems before building | The problem really exists |
Willingness to Pay | The price a customer is willing to pay | Someone will pay for it |
B2B2C | Serving end users through institutions | Payers and users are on the same side |
Smoke Test | Testing market response before the product is done | The market really wants it |
Privacy by Design | Privacy built in from the start of design | You deserve to be trusted |
Moat | An advantage that's hard to copy | Others can't easily replace you |
This case is adapted from my experience across many reviews and mentoring sessions. Industries, people and details have been rewritten and do not refer to any specific team.
Further Reading
Related posts on JCXO
Comments