Housing Ecosystem Triangulation
Combine proprietary channel data to form an independent market view
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 98%
Housing Ecosystem Triangulation builds an independent market view from proprietary information gathered across interconnected industry silos. The analyst first maps the ecosystem, then develops relationships with private builders, brokers, lenders, manufacturers, landlords, and other operators. Responses are aggregated within each group to identify its internal direction, then compared across groups for confirmation or contradiction. Mortgage conditions might be checked against builder orders, broker activity, land lending, and building-product demand. The mechanism is stronger than a single survey because each silo provides a different view of the same underlying cycle. Persistent relationships and repeated questions turn scattered channel checks into longitudinal evidence that can support forecasts and investment recommendations.
Origin
Extracted from Masters in Business, where Ivy Zellman described the private-company network and multi-silo research system developed throughout her housing-analysis career.
Core principles
- 01Study the operating channel rather than relying solely on public management
- 02Collect comparable proprietary data from distinct industry silos
- 03Use adjacent sectors to confirm or challenge each signal
- 04Build relationships that improve data quality over time
- 05Form conclusions independently of consensus forecasts
How to run it
- 1
Map the ecosystem
List the participant groups that influence supply, demand, financing, pricing, and transaction volume. Include private operators whose activity may appear before public-company results.
Pro tip Separate groups by economic function rather than company label.
Watch out Leaving out financing or distribution can hide the mechanism driving reported sales.
- 2
Build channel relationships
Recruit knowledgeable operators from each silo and establish a recurring exchange of information. Offer useful aggregated findings without exposing individual responses.
Pro tip Start with phone conversations to build trust before automating surveys.
Watch out A large but unrepresentative network can produce false confidence.
- 3
Collect structured evidence
Ask recurring questions that measure current conditions, changes from the prior period, expectations, and constraints. Preserve consistency so trends can be compared over time.
Pro tip Combine quantitative questions with limited qualitative prompts explaining unusual changes.
Watch out Constantly changing questions destroys longitudinal comparability.
- 4
Aggregate each silo
Summarize the direction, breadth, and magnitude of responses within every participant group. Identify outliers rather than allowing them to dominate the conclusion.
Pro tip Track both averages and the percentage of respondents reporting improvement or deterioration.
Watch out Do not treat anecdotes as representative data.
- 5
Triangulate across silos
Compare related signals and trace disagreements to timing, geography, customer segment, or methodology. Raise confidence only when the proposed mechanism fits several independent observations.
Pro tip Use contradictions as investigation prompts rather than discarding them.
Watch out Correlation across silos does not prove causation when all respondents share the same external narrative.
- 6
Form an independent thesis
Translate the reconciled evidence into a market view, forecast, or investment recommendation. State what future observations would confirm or invalidate it.
Pro tip Document the thesis before reviewing consensus forecasts.
Watch out Avoid adjusting the conclusion merely to align with management guidance or market sentiment.
In the wild
A housing analyst sees weaker private-builder orders, tighter land-development lending, slower broker transactions, and falling product shipments while public builders still issue optimistic guidance. Because independent silos point in the same direction, the analyst lowers forecasts before quarterly earnings reveal the slowdown.
→ The cross-silo evidence produces an earlier and more defensible forecast than management guidance alone.
National broker data looks healthy, but builders report weakness. The analyst segments both datasets geographically and discovers that broker strength comes from Florida and Arizona while builder weakness is concentrated in California luxury markets. The apparent contradiction becomes a regional allocation insight.
→ The analyst avoids an inaccurate national conclusion and identifies the affected markets and price points.
Common mistakes
Listening only to public management
Management teams may share the same incentives or lagging information. Independent private channels provide a necessary external check.
Pooling incompatible respondents
Combining different regions, price points, or business models without segmentation can conceal the mechanism behind changing conditions.
Forcing agreement across silos
Conflicting evidence may reveal timing differences or a divided market. Investigate it instead of selecting only the data that supports the preferred thesis.
Is it for you?
Best for
It is best for investors and strategists studying fragmented industries with many private operators and interconnected supply chains.
Not ideal for
It is not ideal for quick decisions when there is no time or access to build a credible participant network.
From the transcript
“We're taking all of these silos, and we're aggregating data that is proprietary data within each silo, and we're then triangulating it amongst this ecosystem…”
“you need to go dig in the channel. You need to find private companies.”
“we will give you the research for free, for free if you fill out the survey.”
From the episode
Ivy Zelman Discusses Real Estate