This past week, Google pre-announced a new live Internet TV service called YouTube TV.  YouTube TV is a “skinny bundle”: a pay-TV service with a subset of channels commonly available on American cable and satellite TV at a price that’s lower than what pay-TV subscribers pay for TV programming.

I used this announcement in my latest column on Forbes.com as a “news peg” to enable to me to talk about some fascinating research on pay-TV bundling that I saw recently, from the Boston-based consulting firm Altman Vilandrie & Co (AV&Co).  AV&Co does an annual consumer survey on attitudes towards pay-TV services, in which it gathers lots of data about (among other things) programming preferences, consumers’ willingness to pay for certain channels, and consumers’ interest in paying proportionately to the number of channels in a package.

This data drives a sophisticated algorithm that produces predictions of take rates — percentages of people who will subscribe to a given offer — for certain bundles of channels.  They can take a bundle of channels, such as YouTube TV’s  bundle of 44 channels, and predict how it will fare at YouTube TV’s $35/month price point as well as any other price point.  The fees that Internet TV providers pay to programming networks are confidential, but another research firm, SNL Kagan, publishes estimates.  AV&Co can use those numbers to predict how profitable an Internet TV service will be.

Until now, Internet TV service providers have grappled with issues other than divining who wants to watch which channels.  Their ability to construct bundles — skinny or otherwise — has depended on programming networks’ willingness to make deals at all (given that there may be competitive or other strategic concerns), and even if so, whether networks are willing to unbundle their channel lineups for licensing.

Skinny bundles are particularly problematic: most Americans get their TV from the same company that provides their basic Internet service, and they typically pay $40-50 per month for TV programming on top of that.  That means that an Internet TV service must charge $40/month or less to attract cord-cutters or cord-nevers.  This doesn’t leave much room for error in picking the right set of channels to attract a targeted audience; and on the other hand, many people are satisfied with Netflix or Hulu Plus at $10/month or less (not to mention broadcast television and various free video content).

As a result, the market for skinny bundles is small — there are only three contenders in the U.S. market (from Dish Network, AT&T, and Sony) — as are audiences, so far.

In the Forbes piece, I take AV&Co survey data on the channels that respondents said they “must have” in a pay-TV service and use it to show the relative values of the existing skinny bundles — Dish Network Sling TV Orange and Blue, DirecTV Now “Live a Little,” and Sony PlayStation Vue Access — as well as YouTube TV.  Spoiler alert: it doesn’t look good for the newcomers.  (I’ll report back once Hulu announces its channel lineup.)

In any case, research like AV&Co’s could help take the guesswork out of skinny bundling and lead to more willingness to license and experiment.  These are necessary steps toward the long-promised world where consumers will actually get to choose the TV they pay for.