Slashdot Mirror

← Back to Stories (view on slashdot.org)

Facebook Discloses New Measurement Errors, Continues To Hone Its Math (marketingland.com)

Posted by msmash on from the math-is-hard dept.

An anonymous reader shares an article on MarketingLand: For the third time since September, Facebook is disclosing new measurement errors. The two new errors affected the reaction counts Facebook reports on Pages' Live videos, as well as the engagement figures Facebook reports for off-Facebook links; the latter link engagement metrics were recently used in investigations by BuzzFeed and The New York Times into fake news articles' performance on Facebook. In addition to acknowledging the two new errors -- of which one has been corrected and one is still being inspected -- Facebook has refined a measurement marketers may reference when buying ads through the social network. None of the aforementioned metrics had any impact on how much money Facebook charges advertisers for their campaigns. But they may have informed brands' Facebook ad-buying strategies as well as brands', publishers' and others' Facebook-related content-publishing strategies.

19 of 36 comments (clear)

  1. Imaginary benefits of social media advertising by sinij · · Score: 2

    It doesn't really matter how Facebook measures engagement and so on, much bigger problem is complete lack of results of all social media advertising. This is not 1-2% response ineffective, this is 10^-20 ineffective. It is quite likely that the entire Facebook advertising haven't sold a single thing today.

    1. Re:Imaginary benefits of social media advertising by fpophoto · · Score: 2

      The local garage sale groups in my area do pretty well though. It's probably the only regular use I get out of Facebook these days.

    2. Re:Imaginary benefits of social media advertising by rikkards · · Score: 1

      My wife got better results with her business using Facebook advertising than Google's adwords.
      That said my biggest annoyance with FB is the sale groups. Once I see them, if I don't add them and X them out, I don't want to see them again, same with suggested pages. The first I am pretty confident there is no monetary gain for FB other than potentially more use of their app (IOW no direct money for access), latter not so sure.

    3. Re:Imaginary benefits of social media advertising by Anonymous Coward · · Score: 1

      I worked on a retail eCommerce site for a while (five below) and they made bank on Facebook. It helps that their target market is teen girls and they sell shit that's only a few dollars, but it was super effective for them.

      Also there was this one girl from finance that thought we couldn't hear when she farted, but we could hear. we could hear.

    4. Re:Imaginary benefits of social media advertising by Anonymous Coward · · Score: 1

      I think big company advertisers are losing out massively but some smaller guys hocking cheap crap are making out like bandits.

      I suspect the opposite. Big companies reap the only generally agreed-upon benefit of online advertising - building brand awareness. The conversion rate on their ads can suck all it wants, they cash in when someone is deciding which cola to buy and seeing the Coke logo 500 times a day works its magic on their decision making.

      I suppose it is possible that some very savvy businesses are cashing in by pushing low cost impulse purchases.

      Honestly, I think most of FB's money comes from a huge and constantly revolving cast of failed business ideas that blew their wad on advertising.

    5. Re:Imaginary benefits of social media advertising by Hognoxious · · Score: 1

      Obviously a tranny. Everyone knows girls don't fart.

      --
      Confucius say, "Find worm in apple - bad. Find half a worm - worse."
    6. Re:Imaginary benefits of social media advertising by garcia · · Score: 1

      I work in the marketing analytics and attribution space and can confidently speak to this topic. While Social isn't the BEST performer, it doesn't carry with it the dire statement of a "complete lack of results" as you state.

      With dependencies on vertical and how the advertising is used in known conjunction with other channels, Social definitely does have an assister effect on those other channels. The problem you may be encountering is relying solely on outdated analysis methods which do not appropriately track credit for known users across the entirety of their path to purchase or you're simply just looking at in effective ad buying behavior resulting in poor ROAS.

      Done right, Social is definitely valuable for relatively low cost when compared to the much larger channels (based on investment) and can absolutely jack up your return on those other channels as an assister but definitely is not going to be a 1:1 return as the only advertising channel you should leverage if you are hoping for conversion.

    7. Re:Imaginary benefits of social media advertising by TechyImmigrant · · Score: 1

      It doesn't really matter how Facebook measures engagement and so on, much bigger problem is complete lack of results of all social media advertising. This is not 1-2% response ineffective, this is 10^-20 ineffective. It is quite likely that the entire Facebook advertising haven't sold a single thing today.

      Bullshit.
      My wife's store get the best result by far through Facebook advertising. Better than Google. Better than the local papers. It's measurable.

      --
      I should use this sig to advertise my book ISBN-13 : 978-1501515132.
    8. Re:Imaginary benefits of social media advertising by TechyImmigrant · · Score: 1

      It doesn't really matter how Facebook measures engagement and so on, much bigger problem is complete lack of results of all social media advertising. This is not 1-2% response ineffective, this is 10^-20 ineffective. It is quite likely that the entire Facebook advertising haven't sold a single thing today.

      Bullshit.
      My wife's store get the best result by far through Facebook advertising. Better than Google. Better than the local papers. It's measurable.

      Measurable by Facebook?

      Certainly not. Facebook don't have our sales data. It's measurable with a Chi-Square test of sales against the periods of active advertising.

       

      --
      I should use this sig to advertise my book ISBN-13 : 978-1501515132.
    9. Re:Imaginary benefits of social media advertising by sinij · · Score: 1

      Yes, that what all snake oil salespeople tell me - you are not measuring results right. Until you can measure "assisster effect" it is just as imaginary as Santa Claus.

      Yes, my original statement was generalization. There are few industries, like luxury accessories, that could benefit from social media. On other hand, if you are, for example, selling refrigerators or accounting software you need not bother.

  2. Re:Really by Pseudonymous+Powers · · Score: 2

    Last sentence makes no sense. They have informed of what?

    When they say "they informed... ad-buying strategies", they're saying that the erroneous metrics were used by companies to make strategic decisions, which may now turn out to be invalid.

  3. They didn't pay attention in math class by Idisagree · · Score: 1

    too busy on Facebook, I guess.

    (oh dear... looks like its going to be cheese and corn on the cob for dinner tonight!)

  4. Not God? by U8MyData · · Score: 1

    Oh, really? Please, the day will come when tech companies, although full of great ideas, they will falter. FB has been on my target list forever. Being fair, social media is the entity on my list. Remember MySpace anyone? The day will come when the darling is no longer the darling. RIP FB.

  5. Lies, damn lies, and statistics by Anonymous Coward · · Score: 1

    Never trust ad platform metrics that you didn't measure yourself.

  6. Honing math... by U8MyData · · Score: 1

    Is this a question of honing or manipulation?

  7. Reliable numbers by fph+il+quozientatore · · Score: 1

    You mean my ad hasn't been seen by 10 billion people in the last month?

    --
    My first program:

    Hell Segmentation fault

  8. Lol, ooops by JustAnotherOldGuy · · Score: 1

    Facebook: "Lol, ooops, we overcharged you, our bad, sorry, no refunds."

    --
    Just cruising through this digital world at 33 1/3 rpm...
  9. "Measurements" by sexconker · · Score: 1

    It's not a measurement, it's a count. And it's fucked up for 1 of 2 reasons (or both):

    A: Facebook lied to sell more ads and make more money, or was incompetent with the queries but didn't care (because they want to sell more ads and make more money).
    Solution: LOL.

    B: Facebook uses shitty bigdata nosql non relational horseshit instead of a proper relational database.
    Solution: Use something sane, like SQL. You can trivially get a count of how many people have seen an ad and how many have clicked it, with whatever parameters (date range, region, gender, age, race, etc.) the evil advertisers want. Your results will be deterministic, complete, and correct. Yes, even for a distributed, web-scale database. Modern shit makes all the syncing and log shipping transparent. At worst you query (the full, complete, correct) database as of x minutes old, where x is your syncing window in minutes. Even if the queries are too hard for you to figure out, you can export your shit to many statistical analysis tools to do it for you.

    1. Re:"Measurements" by chrism238 · · Score: 1

      Nothing you've described cannot be performed accurately using nosql. Are you demonstrating an ignorance or a bias?