Weighted Behavioral Demand Measurement
Measure audience demand by weighting observable actions by their intensity
- Difficulty
- Advanced
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 95%
Weighted Behavioral Demand Measurement replaces a single undifferentiated popularity count with a portfolio of observable audience actions. Begin by collecting signals such as social likes, searches, discussions, and downloads. Then weight each signal according to the commitment it implies: spending hours downloading a program expresses more demand than clicking a lightweight reaction. Aggregate those weighted actions to estimate demand for each title rather than extrapolating solely from a small survey panel. The same data can also reveal audience demographics and affinities with other products, making the result useful for content valuation and advertising decisions. Tuckman presents this emerging method as a complement to established ratings systems while behavioral data becomes more mature and commercially accepted.
Origin
Tuckman explains the method through Parrot Analytics, where he was a board member, contrasting internet demand signals with Nielsen's panel-based television ratings.
Core principles
- 01Observable behavior reveals demand more directly than a small survey panel
- 02Different audience actions express different levels of commitment
- 03High-effort actions deserve more weight than low-effort reactions
- 04Demographic and product affinities can make aggregate demand commercially useful
How to run it
- 1
Define demand signals
Identify observable actions that people take when they care about a show, product, or topic.
Pro tip Include signals from several types of platform rather than treating one network as the whole audience.
Watch out Do not assume every visible action reflects equal interest.
- 2
Collect the behavior
Gather the relevant likes, searches, discussions, downloads, and other actions at sufficient scale.
Pro tip Preserve the market and audience attributes attached to each signal where possible.
Watch out A narrow source can systematically omit parts of the audience.
- 3
Weight signal intensity
Assign greater importance to actions requiring more time, effort, or commitment.
Pro tip Treat the weighting model as an explicit assumption that can be tested and revised.
Watch out Adding raw counts without weighting can equate a casual like with hours of committed behavior.
- 4
Aggregate demand
Combine the weighted signals into a comparable demand measure for each subject and market.
Pro tip Retain the underlying components so the total can be explained.
- 5
Profile the audience
Use demographic and affinity data to understand who is expressing the measured demand and what else they value.
Pro tip Translate audience affinities into testable advertising or distribution hypotheses.
Watch out An aggregate demand score alone does not explain who produced it.
- 6
Triangulate decisions
Use behavioral demand alongside established ratings and transaction data while the method matures.
Pro tip Investigate disagreement between measures instead of automatically choosing one.
Watch out Do not present an emerging demand measure as a direct count of viewers.
In the wild
A viewer might like Game of Thrones on Facebook, read about Jon Snow on Wikipedia, discuss the show on Reddit, or download four hours of episodes. The model collects those actions but gives greater weight to the download because it expresses substantially more commitment.
→ Different online behaviors become one intensity-weighted estimate of audience demand.
After measuring demand for Game of Thrones, an analyst examines the people behind those signals and finds affinities with other products, including Lincoln Motors and Dannon yogurt.
→ Content demand becomes a commercially useful audience profile rather than a popularity total alone.
Common mistakes
Counting every action equally
A lightweight reaction and a high-effort action do not express the same intensity of demand.
Confusing demand with viewership
The method measures observable expressions of demand, not a literal census of who watched a program.
Discarding established measures
Tuckman describes behavioral-demand data as a complement to Nielsen while the new industry develops.
Is it for you?
Best for
It is best for media owners, marketers, and analysts comparing demand for content across audiences and markets.
Not ideal for
It is not ideal when relevant audience behavior leaves little observable digital trace.
From the transcript
“Parrot is listening and scraping to all of these messages in the Internet, listening to everything, but more importantly weighing it all.”
“If you download four hours and pirated of Game of Thrones, you're pretty passionate about a show. If you're liking it on Facebook, you like…”
“if you weigh all that together, you could come up with a real, um, actual, genuine amount of data that people are expressing demand for…”
From the episode
Interview With Bruce Tuchman: Masters in Business (Audio)