The leasing evidence described demand from years earlier
A public equity team held shares in a listed landlord that was repositioning older offices, upgrading them and putting them back in front of occupiers. The value of that plan rested on demand that had not arrived yet: enough companies choosing that kind of space, on terms the landlord could live with. The team had to decide whether to keep the holding at its current size.
The usual evidence is signed leases. A lease tells you what an occupier wanted at the moment it signed, often years earlier and under different assumptions about how its people would work. Research built on past take up reads the past confidently. Reading forward meant asking the occupiers themselves, and asking enough of them to tell real intent from general enthusiasm.
We asked the occupiers what they would commit to
We built the survey with the analyst team and put it to people who make property decisions inside occupier organizations. Respondents were screened on the size of the organization, the sector it works in, and whether they genuinely held authority over a property decision rather than an opinion about one. Former corporate real estate leads and former letting agents were included as well.
The questions were written around commitment rather than preference. Occupiers will say they want amenities they would never agree to pay for, so we asked what they would sign for and on what terms. Screening by size and sector meant the answers could be cut by the segments the building was actually aimed at. Every participant cleared our compliance framework before the survey went out.
The demand was narrower than the plan assumed
The survey showed which occupier segments would genuinely commit to the repositioned space, and on what terms. The demand was there, but it sat in fewer segments than the landlord’s plan assumed.
The team reduced its position while the evidence was still ahead of the market, and could show the committee exactly where the conclusion came from. The survey was kept so the same questions could be put again later, turning one reading into a signal the team could track on the name.
