The Austin Broker Barometer · Methodology

How the Barometer is made

For reporters, researchers and anyone citing the numbers: who is surveyed, how the index is calculated, and what we leave out. The short version for brokers is on the main Barometer page.

At a glance

Who counts

An answer counts toward the index only when all of these are true:

• The respondent works deals. The first question asks for the broker’s role. Producing brokers, and managing brokers or team leads who also produce, are counted. Anyone who isn’t working deals (managers, support staff, owners) stops at that question and isn’t counted.
• The work is in the Austin area. A broker whose only submarket for the quarter is outside the Austin area is not counted.
• It is complete. All four headline questions must be answered.
• It is the broker’s last answer. One answer per business email: if a broker submits twice, the latest complete answer before the close counts. Answers that arrive after the close are not counted.
• It is not ours. Answers from Austin Office Space email addresses are dropped before anything is calculated.

The headline number: Broker Sentiment Index

Four questions make the index. Each is a five-step scale, and each answer gets a fixed score: the most positive answer 1, the next 0.75, the middle 0.5, the next 0.25, the most negative 0. This is the diffusion-index weighting used by Japan’s Economy Watchers Survey.

Reading it. 50 is the midpoint: above 50 the weighted average leans positive, below 50 it leans negative. Each release labels the number with a fixed band on the index rounded to one decimal: under 2 points from 50 is “little changed”; 2 up to 6 is “modest”; 6 up to 11 is “moderate”; 11 or more is “strong”, with “expansion” above 50 and “contraction” below.

“N/A” answers. The three questions about the quarter just closed offer “N/A” for a broker with nothing to compare (new to the market, or no deals in either quarter). Those answers are reported as a share and left out of that component’s average; the broker still counts in the overall total. Every release prints the number of answers behind each component.

Reported beside the index, never blended into it

• Year-ago comparison: tours and inquiries vs. the same quarter a year earlier, scored the same way as the components. It puts seasonal swings in context.
• Pipeline index: active requirements or deals in negotiation vs. last quarter, scored the same way.
• Conditions index: “Austin leasing conditions overall, right now” (good / fair / poor), calculated the way the National Association of Home Builders reports its builder survey: (% good − % poor + 100) ÷ 2.
• Shares: where negotiating leverage sits, the direction of effective rents, one word for the market, and the biggest factor moving it. The factor list is shown in a different random order to each broker so no option gets a first-position advantage; the report ranks them by how many brokers picked each, with “No single factor” and “Other” always last.
• Cuts by product type, side of the table (tenant rep, landlord rep, both), submarket (a broker counts in each submarket they name, up to three) and firm type. National or global firms are single-owner corporate platforms, identified by business email domain; affiliate networks of locally owned offices, such as Lee & Associates and Cresa, count as local or regional. These describe; they never re-weight the index.

Publication rules

• Every quarter, whatever the number. A down quarter is the headline.
• Sample-size tiers. 40 or more counted answers: a full release. 30 to 39: published and labeled a small sample. Under 30: we publish participation and method, not the index.
• No margin of error. This is a panel of working brokers, not a random sample, so a margin of error would be false precision.
• Small groups stay hidden. No breakdown is printed for fewer than 10 brokers; the cell reads “insufficient”.
• Early vs. late check. We compare answers from the first and second halves of the survey window, so a news event mid-survey shows up rather than hides.
• From the second edition, the headline change is printed two ways: all respondents, and only brokers who answered both quarters.
• Corrections are dated in the change log below. Questions are not changed while a survey is open.

Limitations

• Respondents come from one brokerage’s email lists and the brokers they forward the link to. The panel is broad but not random.
• Answers are brokers’ own reports of their business and their outlook, not transaction records. The Barometer is an early read on direction, not a measure of absorption, vacancy or rent levels.
• Austin Office Space competes with the brokers it surveys. That is why the firm does not answer, publishes every quarter on a fixed method, and keeps its own view to a separate, signed note that never touches the number.

Comments and attribution

Each broker may leave one comment and chooses how it is used: quoted by name and firm, or anonymously. Only brokers who choose name and firm are named, and choosing it also means a reporter covering the results may contact them. We may quote part of a comment; we never change its meaning. Brokers can withdraw a comment by email.

The questionnaire

Exactly as brokers see it for Q3 2026. Every scale is shown most-positive first.

How to cite

Media contact

Chad Jewell, Founder & Principal Broker, Austin Office Space
chad@austinofficespace.com · 512-633-7703

Change log

First edition published Oct 29, 2026. No changes to the method yet; any change will be dated here.