Real-Time Company Survey Method
Build a current economic picture by surveying operators across industries
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 98%
The method constructs a near-real-time economic picture from recurring surveys of people who directly run businesses. Researchers divide the economy into relevant industries, recruit operators with detailed knowledge of current orders, sales, staffing, pricing, and activity, and ask comparable questions on a regular schedule. Aggregating those responses reveals whether weakness or strength is isolated, widespread, accelerating, or fading. Participants receive the resulting research, giving them an incentive to contribute and a benchmark against competitors. The framework does not treat anecdotes as a substitute for statistics; it organizes many informed observations into repeatable sector indicators that can be compared with government releases, market prices, and prior cycles.
Origin
Ed Hyman adapted the idea from Johns in red book's retailer survey, then expanded it across roughly thirty industries at ISI Evercore. Extracted from Masters in Business.
Core principles
- 01Operators know their own businesses in granular detail
- 02Primary observations can complement delayed official statistics
- 03A diverse panel reveals both sector trends and economic breadth
- 04Participants contribute more reliably when they receive useful results
- 05Repeated surveys become more valuable through consistent comparison
How to run it
- 1
Map the economy
Select industries that collectively provide a useful view of economic activity. Include narrow subsectors when they offer distinctive leading information.
Pro tip Mix consumer-facing, industrial, transportation, and geographically exposed businesses.
Watch out A panel concentrated in one sector can create a misleading macro signal.
- 2
Recruit informed operators
Find executives and operators who live their business conditions every day and can report concrete changes. Build enough participation to prevent one company from dominating the result.
Pro tip Offer aggregated research so participants can compare their experience with peers.
Watch out Do not disclose identifiable competitor responses.
- 3
Standardize recurring questions
Ask a stable set of questions about activity, direction, expectations, and relevant operating variables. Repeat the survey on a consistent weekly or monthly cadence.
Pro tip Preserve question wording so changes remain comparable over time.
Watch out Constantly changing the instrument destroys the historical baseline.
- 4
Aggregate sector signals
Convert responses into consistent industry indicators and compare them with prior readings. Look for changes in direction, magnitude, and breadth.
Pro tip Investigate divergences between a company and its industry panel.
Watch out Do not overreact to a single volatile reading.
- 5
Cross-check the mosaic
Compare company surveys with markets, official statistics, and other primary observations. Use disagreements as prompts for investigation rather than automatically choosing one source.
Pro tip Give more weight to signals confirmed across independent industries.
Watch out Primary research can still contain sampling and sentiment biases.
- 6
Maintain signal quality
Review whether each survey continues to explain real conditions and remove instruments that stop working. Preserve useful panels long enough to establish cycle-to-cycle context.
Pro tip Document structural changes that may break comparability.
Watch out Do not retain a survey merely because it once produced a strong call.
In the wild
Hyman's team surveys the people who grow Christmas trees, the companies that transport them, and the sellers operating in cities. Combining observations from each stage provides a current view of demand and supply conditions rather than waiting for a broad retail release.
→ Researchers obtain a granular, cross-checked snapshot of a seasonal market.
A recurring panel of trucking companies receives the aggregated survey in exchange for participating. If one carrier reports strong business while the industry measure drops, or its own activity drops while peers remain stable, management can investigate whether the difference is company-specific.
→ The same survey informs macroeconomic research and gives participants a competitive diagnostic.
Common mistakes
Treating one anecdote as the economy
A vivid operator account is useful but cannot represent an entire sector. Build a sufficiently diverse panel and examine the breadth of responses.
Changing the survey repeatedly
Inconsistent questions and timing make successive readings difficult to compare. Preserve a stable core instrument while documenting necessary revisions.
Ignoring participant incentives
Operators have little reason to provide recurring, thoughtful responses without receiving value. Give them useful aggregated findings while protecting confidentiality.
Is it for you?
Best for
It is best for investors, economists, and strategists who need timely evidence from the real economy.
Not ideal for
It is not ideal for teams without sustained access to representative business operators or the capacity to maintain a consistent panel.
From the transcript
“If you meet somebody, say that runs a business, and you say house business, they'll always tell you with actually vivid detail, real granularity because…”
“So now we survey about thirty industries, maybe three hundred companies in each industry, three hundred companies overall, thirty industries like retail for example, or…”
“So if they participate with us, I send them our research.”
From the episode
Ed Hyman on How Investors Should Use Economic Data
Ed Hyman