Microspeak sighting: over-index

The other day, I heard the Microspeak term over-index used in the wild. It's like birdwatching, but for words.

In a brief interview with Redfin CEO Glenn Kelman, they talked about the significance of having Amazon headquartered in your city.

And Amazon can take some of the wealth of Seattle or Silicon Valley and bring it to the center of the country, and that can be really good. But I think people are overindexed on Amazon. There's going to be Apple, and then there's going to be Dropbox, and there's going to be Intel.

The term here is being used in the Microspeak way: Given too much prominence in a discussion. He's saying that people are focusing too much on Amazon and missing the other companies.

Comments (6)
  1. Peter Doubleday says:

    But why an “index?”

    There’s a perfectly cromulent CS term that applies here: it’s called a “reference.”

    You can have strong references, weak references, phantom references, and I forget the other one but it’s a reference anyhow. That’s, like, four uses of references in a Powerpoint presentation.

    Beats indexing by a factor of two (if you include direct indexing and, say, Trie referencing).

    1. voo says:

      What meaning of the word “reference” do you mean? I can’t think of any that makes sense in this context in any way (for me from a more theoretical CS pov a reference is a piece of data that allows the retrieval of some other data – but we aren’t referencing anything here)

      The word index on the other hand has a rather obvious connection. To quote the oxford dictionary:

      2. A sign or measure of something.
      ‘exam results may serve as an index of the teacher’s effectiveness’

      I can see how you arrive at “overindex” from that.

    2. Brian says:

      In a database, if you over-index, you end up a) wasting space, b) wasting machine cycles (inserts take longer), c) getting very little benefit (by definition, if something is over-indexed, you don’t need the index you just created) and finally d) make something seem more important than it really is.
      In the context of the Amazon HQ2 quote, that actually pretty closely matches the point of what Mr. Kelman is trying to say.

    3. ErikF says:

      It’s probably from the economic term index (it could also be from statistics, but is more likely the economic version.) Overindexing would be favouring one group of data over other groups.

      Just in case the previous comment was in jest: :-)

  2. Scarlet Manuka says:

    I’m glad you added the clarification of the meaning. I initially read it the other way around, as saying that people were being given too much prominence by Amazon, which given the stories I’ve heard about the working conditions in the warehouses seems to indicate a race to the bottom in terms of treating people with a bit of basic respect.

  3. Spire says:

    Based on your definition, it sounds like Amazon is the one that’s overindexed — not people. But that’s not what Kelman is saying.

Comments are closed.

Skip to main content