- Press
- Pitch
- Blog
Nationwide survey · N = 2,601 · Fieldwork 5–8 Aug 2026 · Ages 18–48
The economics say move in. The fear says don't.
2,601 people across Greece on flat sharing: who has done it, what stops everyone else, and why the second bedroom stays empty.
One market, two forces
Demand and resistance are each driven by a single, dominant factor. Nothing else comes close.
of those open to a stranger roommate cite the cost of living. Coliving motivation overall: 70% economic, 22% social, 8% location.
of those who refuse cite safety and trust — and 70% of everyone ranks safety as the most important platform feature.
The strategic reading: demand does not need to be created — expensive rent is doing that job (73% call rents the #1 search difficulty, rising to 83% among 18–26s). The product's entire job is to lower one well-identified barrier at the front door.
The fear is of the unknown — literally
Ask everyone how comfortable they'd be moving in with a stranger tomorrow, and 55% say "not at all." Ask only the people who have actually done it, and the distribution inverts: comfort ("quite" + "very") jumps from 17% to 53%.
All respondents vs. those who have lived with a stranger before
- All respondents
- Lived with a stranger before
Data table
| Answer | All respondents | Lived with a stranger before |
|---|---|---|
| Not at all | 55% | 15% |
| A little | 32% | 32% |
| Quite | 12% | 38% |
| Very | 5% | 15% |
Combined with 77% of cohabitants reporting a positive social impact, this says the fear is anticipatory, not experiential — people fear the unknown, not a remembered bad outcome. The implication: the first match is the entire funnel. Anything that de-risks a first cohabitation (trial periods, structured onboarding, mediation, exit guarantees) manufactures the strongest persuasion available — the experience itself.
Verification alone is not the product
Third-party verification gets only a 35% firm "yes" (45% "maybe", 20% "no"). Checks are table stakes. Trust here is built by process, not by a badge.
A large latent market, a thin active slice
Share of all respondents at each stage
- 80%Know what flat sharing isawareness· −20 pts
- 60%Would consider a strangerintent· −43 pts — the trust gap
- 17%Comfortable doing it tomorrowreadiness· −5 pts
- 12%Actively searching right nowtiming
Data table
| Stage | Share | Loss to next stage |
|---|---|---|
| Know what flat sharing is (awareness) | 80% | −20 pts |
| Would consider a stranger (intent) | 60% | −43 pts — the trust gap |
| Comfortable doing it tomorrow (readiness) | 17% | −5 pts |
| Actively searching right now (timing) | 12% | — |
Two products hide in this funnel. The trust gap (60% → 17%) is the conversion problem the safety features must solve. The timing gap (only 12% searching at any moment) argues for capturing users before active search — compatibility profiles built early, waitlists, alerts — rather than competing purely on listing volume.
Match on chores, not demographics
Household chores are the #1 compatibility priority (63%; women 64% / men 56%), the #2 anxiety in the open-ended responses, and the #1 reported real-world conflict. The pattern sharpens with experience: people who've lived with strangers care more about chores and habits, and dramatically less about social proof.
All cohabitants vs. stranger-experienced (multi-select — shares do not sum to 100%)
- All cohabitants
- Stranger-experienced
Data table
| Factor | All cohabitants | Stranger-experienced |
|---|---|---|
| Household chores approach | 63% | 70% |
| Habits (smoking, alcohol…) | 38% | 56% |
| Financial / professional situation | 17% | 20% |
| Spoken language | 16% | 12% |
| Relationship status | 15% | 5% |
| Mutual acquaintances | 11% | 3% |
| Dietary habits | 9% | 5% |
Product translation: matching should weight cleanliness standards, chore-sharing expectations, and habits (smoking, alcohol, schedules) far above age, occupation, or social-graph overlap. The people who know best have already said what matters.
Stranger coliving is a different product
Ask cohabitants what the best part was, and the answer depends on who they lived with. Partners and friends deliver companionship. Strangers deliver something else: financial relief and new people.
All experienced respondents → stranger-experienced subgroup
- Rises among the stranger-experienced
- Falls among the stranger-experienced
Data table
| Benefit | All cohabitants | Stranger-experienced |
|---|---|---|
| Companionship | 41% | 15% |
| Financial support | 20% | 27% |
| New acquaintances | 5% | 22% |
| Chores help | 14% | 18% |
| Safety | 12% | 12% |
| Personal development | 6% | 6% |
Positioning implication: don't sell stranger coliving as ready-made friendship — the people who've done it don't describe it that way. Sell it as financial relief plus an expanding social circle, with friendship as a possible upside rather than the promise.
The hardest segment to convert is the one that needs it most
All respondents
Data table
| Answer | Share |
|---|---|
| No | 42% |
| Yes | 38% |
| Maybe | 20% |
Share answering not at all
Data table
| Answer | Share |
|---|---|
| Women | 58% |
| Men | 42% |
This overlap is the report's most actionable segment finding: the largest underserved need sits behind the highest conversion barrier. It is a direct argument for safety-forward features built for women — verified women-only matching, facilitated in-person meetings, visible reporting and mediation — as growth strategy, not compliance.
What people actually worry about, in their own words
Open-ended responses, ranked by prominence. No percentages exist here — this is qualitative texture, and it maps almost perfectly onto the quantitative findings.
- fear of the unknown
- cleanliness & chores
- delinquent behaviour
- character matching
- habits
- privacy
- communication
- trust
- safety
- schedule
- finances
- shared spaces
- sexuality
- arguments
- values
The two biggest entries are the whole thesis: the dominant anxiety (the unknown) is the one that experience dissolves (FIG. 1), and the second (chores) is the one an algorithm can screen for (FIG. 3).
Monetization signal — promising, held loosely
of tenants say they would pay for facilitation services
of property owners say the same
willingness on both sides of the match
These are unusually high figures, and they deserve a caveat in any deck: stated willingness to pay is not revealed willingness to pay. Present them as a directional signal, not a revenue forecast. The questionnaire's guarantee-specific question appears to have been merged into the general facilitation figures; recovering or re-fielding that answer matters, because a vetting guarantee is the most naturally monetizable safety product this data points to.
Three ways to tell this story
Same dataset, three different leads.
The societal angle
“Nearly half of Greeks can't afford to rent alone — but fear of strangers keeps flats half-empty”
Lead with the national numbers: 42% cannot afford solo rent, 73% call rents the top search problem, 60% open to sharing, 88% blocked by trust. Human-interest close: 77% of those who tried it say it improved their social life.
- → FIG. 0 (floor plan)
- → Tension stats (79 / 88)
- → Affordability + gender
The investor angle
“Demand exists. One barrier blocks it. Experience removes the barrier. We manufacture the experience.”
Four-slide spine: the tension → the funnel (trust gap + timing gap) → the inversion chart as proof the barrier is solvable → two-sided willingness to pay, honestly caveated.
- → FIG. 1 (inversion)
- → FIG. 2 (funnel)
- → WTP stats
The product angle
“We asked 2,601 people about roommates. The dealbreaker isn't who you are — it's who does the dishes.”
Lead with the chores finding, use the slope chart as the surprise ("strangers are not friends — and that is fine"), close with the anxiety cloud and what it taught the matching algorithm.
- → FIG. 3 (compatibility)
- → FIG. 4 (slope)
- → Anxiety cloud
Methodology & honest footnotes
Nationwide survey, N = 2,601 (55% women / 45% men, ages 18–48), fieldwork 5–8 August 2026, online and face-to-face in major urban centres. Figures marked with ~ in the source are approximate.
- Multi-select questions (compatibility factors, cohabitation benefits) do not sum to 100% — label them as multi-select wherever republished.
- The stranger-experienced subgroup is ≈ 250 respondents: directionally solid, but quote margins of error if publishing subgroup comparisons.
- The guarantee-specific willingness-to-pay answer is missing or merged in the source data — recover it before building pricing claims on it.
- 60% "would consider" vs. 17% "ready tomorrow" is an attitude–behaviour gap, not a contradiction; treat them as separate funnel stages.