I measure an insurer's advertising at the moment of the claim, because that is when the customer finally learns whether our promise and their policy were the same thing.
I am executive officer and chief marketing officer of Tomoshibi Life, a Tokyo direct insurer that sells medical, cancer and term life cover by telephone and online. We are wholly owned by Kawasemi Holdings, a listed television-shopping and catalogue group, and at March 2026 we had 1.32 million policies in force and ¥64.9 billion of annualised premiums. I am responsible for a ¥12.6 billion annual marketing budget, 64 colleagues in brand, creative, digital, partnerships and customer research, and every advertisement that carries our name. I report to our president, Masayo Tsuchiya, and present a marketing and conduct report to our board each quarter.
I started in 1997 writing late-night television commercials for a direct-response agency in Osaka. In 2003 I moved to the insurer's side of those calls, building a mail-order medical channel at a mutual life company. I joined Tomoshibi Life in 2008, before it had sold a policy, and left in 2013 to run the insurance business of a comparison site, where I built a paid ranking I later came to regret. I returned to Tomoshibi in 2018 and have led its marketing since 2021.
Most of my work now concerns the distance between what an advertisement leads someone to expect and what a policy pays. That distance is widest for older applicants, for conditions that are hard to say in fifteen seconds, and in channels paid by the application rather than by the customer who stays. Narrowing it has cost us response, some volume and a few arguments, and has made our brand easier to defend in an industry the public has had reason to distrust.
Advertising that says what the first year pays
From late 2021 our television advertising for a relaxed-underwriting medical plan stated aloud that first-year hospital benefits are halved, and applicants were checked against the cheaper standard plan first. Complaints about first-year payments fell from 214 to 61 a year; 9,800 applicants bought the standard plan instead.
Out of rankings we could not label
In October 2023 we left two comparison sites that would not mark paid positions individually and withdrew a satisfaction claim built on a survey of non-customers. Comparison sites fell from 29% to 18% of new policies, and thirteen-month persistency from that channel rose from 84% to 90%.
Partners paid for policies that stay
We stopped paying affiliates and point sites per application in 2024. Digital cost per application rose from ¥31,000 to ¥44,000, but the cost of each policy still in force at thirteen months fell from ¥96,000 to ¥71,000, and digital spending fell by ¥0.7 billion.
A claimant question that reaches the board
Since April 2024 every claimant is asked whether we paid more, about the same or less than expected. The share answering less fell from 14% to 9% by the first half of 2026, and the board sees it beside acquisition cost each quarter.
Current focus
I am preparing our 2027 plan, in which television carries less of our acquisition and more of our explanation, and proposing that the claimant measure join the figures behind our executive officers' bonuses. We are also testing a text-based two-call process for older applicants with hearing loss.
Conversations I welcome
Another executive appointment is not something I am pursuing. With the agreement of our president and parent company, I would accept one outside-director seat at a listed consumer company outside insurance and outside Kawasemi's retail markets. I give a few talks a year and compare notes with other insurers' marketing leaders on specific problems.
