Chief Marketing Officer / Japan
What accuracy cost, and what it returned
- Group revenue
- ¥14.6 bn
- Year ended March 2026
- Rental homes managed
- 6,800
- For about 1,150 private owners in Nagano Prefecture
- Marketing budget
- ¥620 m
- Year to March 2027; team of twenty-six
Decisions & consequences
01Withdrawing rental listings on the day each home was let
In October 2021, six months after I became CMO, the listing-standards team audited all 2,340 rental homes we were advertising on portals and our own website. We found 212 faults. In all, 131 homes had already been let but remained online, some for more than a month. Another 44 used photographs of a renovated unit elsewhere in the same building, and 37 described walking routes, bus services or shops that had changed. Complaints about listing accuracy had reached 58 in the previous year.
Branch managers were not careless. Their performance measures rewarded enquiries, and a stale listing for a popular home still produced calls that staff could steer to a similar vacancy. Two of our four regional managers argued that faster withdrawal would cut enquiries without improving lettings. Several owners had supplied the renovated photographs themselves and preferred them. I thought the redirected call was precisely the problem: a person had been drawn in by a home we could not offer.
With the president's support, we linked portal publication to the leasing system so that a listing came down on the day an application was accepted. We photographed each advertised unit and labelled any comparison images. Branch staff re-walked routes every quarter, and branch reviews began to include listing accuracy alongside lettings. The additional photography and checking cost about ¥38 million a year.
Enquiries fell 17% over the following six months, and the sceptical managers were right that the first quarter looked worse. Lettings recovered within the year. The viewing-to-application rate rose from 31% to 39%, average vacancy fell from 41 to 38 days, and accuracy complaints fell to nine. Some owners with dated interiors saw slower lettings and asked us for advice on modest refurbishment.
Next time I would involve owners before deciding. We told them about unit-specific photographs by letter after the event, and several felt their homes had been exposed rather than represented. A short meeting and an offer to share photography costs would have been fairer.
Publishing condition and repair estimates beside vacant-house prices
By 2022 our vacant-house brokerage handled about 150 sales a year, mostly older wooden houses in three towns where we worked closely with the municipal vacant-house registers. In the year to March 2023, nine of forty-one signed contracts were cancelled or renegotiated after buyers' own surveys found roof, damp or termite damage. Each failed sale cost a seller months, and some heirs gave up and left the house empty.
I proposed that we offer a building condition survey before listing, with the company paying half of the roughly ¥90,000 fee, and publish a summary with a repair-cost range from a partner builder. Our head of brokerage disagreed. He expected sellers to resist and buyers to be frightened by a list of defects, and he was partly right. Several heirs felt the summary made their parents' house look neglected. We were asking sellers to accept that visible problems would shape the price from the start instead of surfacing after a contract.
From April 2023 we offered the survey on every new vacant-house instruction. Eleven of sixty-four sellers declined and withdrew or moved to another broker. For the others, the listing showed the asking price, the survey summary, the repair range and the inspector's qualification. We trained staff to talk buyers through the findings, and the subsidy cost about ¥2.4 million in the first year.
Over the following twelve months, 38 of the 53 surveyed houses sold, against 29 of a similar number the year before. Post-contract cancellations fell from nine to two, and average time on the market fell from 214 to 151 days. Achieved prices averaged 6% below initial asking prices, a wider gap than before, because buyers negotiated on visible defects at the outset. Two of the towns later adopted a similar summary format for their own registers.
I would explain the approach to heirs in person rather than by letter. Most of the withdrawals came from sellers we had never met.
Marketing resort apartments abroad without a projected yield
In 2024 we converted an old ski lodge in the northern resort area into forty-eight apartments, priced from ¥38 million to ¥96 million. Overseas buyers were expected to take most of them, encouraged by the weak yen. The two overseas agencies we worked with asked for prices in Hong Kong, Singapore and Australian dollars and for a headline projected rental return of 6%, which they said competing developments offered.
I declined both. A projected return depended on occupancy, nightly rates and costs we could not control, and a converted price would be wrong by settlement. Our resort sales director disagreed strongly. He expected the building to lose its place in agents' presentations, and one agency did stop actively marketing it. We also gave up the faster sales our plan had assumed.
Instead, our materials in English, traditional Chinese and Japanese gave yen prices, three seasons of actual occupancy for comparable apartments we manage, and a full schedule of annual costs: management fee, repair reserve, fixed asset tax, utilities and the rental-management commission. A typical one-bedroom apartment carried about ¥1.1 million a year in holding costs before any rental income. We stated that the Japanese contract would govern, provided reference translations and advised buyers to take tax advice in both countries.
By the start of the 2024–25 season, thirty-one apartments had sold against a plan of forty. The remaining seventeen sold by September 2025, nine of them after a 4% price reduction that cost about ¥22 million, and the later sales moved revenue into the following financial year. In the first season we received no complaints about undisclosed costs from the new owners, compared with eleven from a 2019 project marketed with a projected return.
I would share the cost schedule with agents before prices were set, not at launch. Part of their resistance came from seeing our figures only after their own presentations were prepared.