10 ms·

Pretty sure Twitter has an ad metrics problem too. As I wrote last year: > When I ran a Twitter Ads campaign for one of my clients, within two days I noticed t

• ndespres 8y ago
  I do exactly that behavior all the time. It's difficult to tell whether an image will expand when I click it on Twitter app for iPhone. When it opens a browser window that I wasn't expecting, I close immediately.
  • alsetmusic 8y ago
    > It's difficult to tell whether an image will expand when I click it on Twitter app for iPhone. Tap and hold on a link. A sheet will pop up asking how to treat the link. The link address will be listed at the top.
    • mbel 8y ago
      Which takes approximately the same amount of time as clicking, going to browser and returning back to the tweet.
• pavanlimo 8y ago
  Can't disagree with Twitter's reasoning there, anecdotally I myself have done that many times.
  • wongarsu 8y ago
    The explanation is solid, but that doesn't excuse Twitter. If you charge for clicks you should do your best to make sure that the clicks are genuine, not design a user interface where 95% of clicks are caused by bad interface design.
    • MiddleEndian 8y ago
      I spoke with a someone recently who uses gmail on his Android device. An ad at the top that resembles an email appears about a second after his real emails appear, timed perfectly to interrupt his attempt to tap and read his first email. Of course he immediately goes back if he makes the mistake, but google probably still charges the advertiser for a click. (This doesn't happen on my phone but I am more likely to have configured it than your average user.)
      • creato 8y ago
        For me, that first "email" is not the ad itself, but the "promos" folder, tapping it only expands the promos folder, not any of the ads. It does appear a ~second after opening the app though. However, a ton of other things cause the same shifting-as-I-tap problem (not just ads) causing me to mistap on the wrong e-mail all the time. So it seems like garbage UI design more than anything else. UIs moving around while trying to interact with them is one of my biggest pet peeves.
      • nwienert 8y ago
        I lived in Spain for a summer and my roommates there were definitively non-techy and had relatively slow internet. I watched them every single night often multiple times accidentally click the ads that load above on youtube (about 1 second delayed, perfectly as they went to click). They would of course always click back once they noticed it, which with slow internet takes more time too and probably doesn't look as much like a bounce. It's the perfect cover for Google. All the techies who would get outraged by something like this generally have fast internet and ad-blockers.
      • nkrisc 8y ago
• emilsedgh 8y ago
  A billion dollar business relying such a shallow dark-ui technique for it's main source of revenue. The state of marketing departments in companies must be a joke if this is how Facebook and Twitter are milking them. I imagine billions of dollars are being wasted in companies with no return value in the name of "marketing". And if anyone ever thinks Facebook and Twitter are actual competition to Google in ad market they must be out of their minds.
  • cm2012 8y ago
    I am a direct response marketer, and only manage ad campaigns that provably generate revenue. Over $10,000,000 spend across more than a dozen companies. I can tell you that for many products Facebook absolutely outperforms Google.
    • murukesh_s 8y ago
      I too can support with anecdotal evidence for that (my wife). One of her habits is browsing Facebook ads and she end up actually buying stuff from those sites.
      • da02 8y ago
        What products has she bought?
        • murukesh_s 8y ago
          Mostly fashion related as those are something you don't search for but an impulse purchase. Fb is mostly like window shopping or street shopping - you have no idea what you want but you like something and take it. I on the other hand if I have to buy something will google it or go to amazon directly.
    • cstejerean 8y ago
      I’m not surprised. Google shows me ads for things I’m searching for, but Facebook shows me ads for things I wouldn’t even know to look for. I can definitely see how one or the other would be better depending on the type of product.
• baybal2 8y ago
  It is near impossible to properly meter ad shows on the web. An ad can load somewhere in the footer of the page and never seen, and it will be logged as a show. If some scripting is involved to log scrolling, and logging it over ajax, it will not help for the fact that when the page is being rendered, footer may for a split second appear within the viewport. The later, is what makes the lion share of contention around video ads.
• snowwrestler 8y ago
  Did you exclude Twitter Display ads? Because those are absolute shit that skew every metric in the wrong direction. And of course they are turned on by default. We had a similar problem and it was all down to Twitter Display. Once we turned those off, the metrics improved immensely.
• knicholes 8y ago
  There's always going to be a ~30% dropoff between clicks and analytics because people click on ads, realize they've done it, and either push back or close the popped-up page. It's even higher for video. I work for a company that is a DSP, ad server, and analytics, and I've been investigating these click discrepancies. It's surprising how many places this can all go wrong. It gets especially worse when there are redirect after redirect to other ad servers, and you lose your tracking macros because the trafficker hasn't configured their landing page tracking macros correctly for each ad server redirect.
  • tripzilch 8y ago
    > It gets especially worse when there are redirect after redirect to other ad servers Yes this is very wrong, but not because your tracking loses accuracy. You realize that this sort of careless trail you send your users along is exactly how people get served malware from ads right? I mean you admit yourself you lose track of where they end up so who knows what they get served or redirected to.
    • knicholes 8y ago
      They're the ones sending themselves down that trail by serving their ads through various third party ad servers.
• ttty 8y ago
  You should check server hits not Google analytics. From my experience analytics didn't track many visitors. For example cloudflare reports 1000 unique users, GA only 300 to 500.
  • village-idiot 8y ago
    Anecdotally, even a lot of non technies in my life are starting to install browser extensions to stop ad spying.
    • ttty 8y ago
      Adblocks don't really blocks analytics.. I guess Google would ban them haha
• thaumasiotes 8y ago
  > they are legitimate clicks, but people are probably clicking the image in the tweet to expand the tweet or image, not expecting it to take them to a different site. As soon as they realize they’ve just clicked an outgoing link, they go back or close the tab before our site and Google Analytics script even has time to load. I do this a lot on Wikipedia on my phone. My workflow is to open a new browser tab (Wikipedia is the homepage) and hit the magnifying glass in the corner to search. But shortly after the page loads, it reconfigures itself so I end up hitting a link to somewhere else. (Wikimedia, I think.) It is incredibly annoying. The equivalent on a desktop is a badly behaved site (such as... gmail) which first loads a bunch of stuff and then, only then, focuses a text field, so you end up typing the second half of your password into the field that just got focused. I am not in favor of letting webpages suddenly adjust their own layout or move your focus once they've already loaded or you've already picked a focus.
  • sjg007 8y ago
    What’s the mobile vs computer demographics here? Me thinks twitter is telling a little twattle.
    • thaumasiotes 8y ago
      Just compare the other comments in this very thread: > I do exactly that behavior all the time. > Can't disagree with Twitter's reasoning there, anecdotally I myself have done that many times. > I spoke with a someone recently who uses gmail on his Android device. An ad at the top that resembles an email appears about a second after his real emails appear, timed perfectly to interrupt his attempt to tap and read his first email. Of course he immediately goes back if he makes the mistake > I watched them every single night often multiple times accidentally click the ads that load above on youtube (about 1 second delayed, perfectly as they went to click). They would of course always click back once they noticed it It happens all the time. This is an issue with dysfunctional webpage design/layout.
      • sjg007 8y ago
        These comments feel oversold.. but one could agree with the dysfunctional by design revenue generation. I don’t think you need to be dysfunctional here.
• cm2187 8y ago
  Also, I thought something like 40% of browsers run adblockers, and possibly a disproportionate amount in your demographics. That could explain the bulk if the difference.
  • andruby 8y ago
    Wow, that would be a lot more than I would expect. Did you have a source for this number?
    • tripzilch 8y ago
      I personally try to make sure that nobody in my circles, techie or otherwise, surfs without an ad blocker. This irresponsible 3rd party advertising bullshit has got to stop. We have people in this very thread openly talking about deploying third party ad scripts that send their users on chains of redirects that they admit to losing track of, and the only thing they see wrong with this is that they can't accurately track their users, not that they very probably sent their users down whatever rabbit/shithole serving malware, miners or drive-by downloaders. Online advertising is a disgusting wasteland and many people coming out of the woodworks in this thread, working in the industry, are knowingly, unknowingly or usually willingly blindly working to create this situation as it is. Dollar bills make for very good blinders, especially if it's your living.
  • tomp 8y ago
    That’s a good point! Twitter’s tracking is almost perfect with their link shortener. I wonder if/when adblockers will take care of that as well (resolve the link once then cache it).
    • mnx 8y ago
      There's actually no need to resolve it. At least so far, they put the full url in the page source. I have a proof of concept extension already coded up for removing the redirects. I plan on adding a bunch of other sites (google, facebook, whatever else) and releasing it.
    • 8y ago
• xevb3k 8y ago
  Why do you (and why do people in general) not use server logs for analytics of this sort? Seems like it might show requests that failed to get to the point of processing the analytics JS as well as those running adblockers.
  • gk1 8y ago
    Not sure how I’d set up server logging on a serverless site. (Not being snarky; I really don’t know.) I accept that GA will be off +/-10% because it’s clientside. Don’t need total accuracy to make marketing decisions.
    • xevb3k 8y ago
      Yes but you’re saying twitter say their metrics are off by a factor of 5 because it’s not running the GA JS. If you wanted to properly investigate that looking at the server logs would probably give a better idea in this case. I guess with serverless you should be able to see this either with your own logging or perhaps on AWS with Cloudwatch? (no idea as I’ve not used it).
    • phyzome 8y ago
      "Serverless"... someone's running a server. What kind? For instance, if you're on AWS and an ELB is involved, you can set up ELB logging to S3.
  • lmkg 8y ago
    Your logs may be split across different services, and some of those services may be difficult to access. For example, if you're using a CDN, many hits won't go to your server. Off-the-shelf analytics solutions have a ton of heuristics and in-house knowledge built up over decades that help deal with inherently janky data. For example: over half of all internet traffic these days is bots. All modern analytics solutions know how to discriminate bot traffic (to an acceptable extent), you would have to re-create that or get heavily skewed data. As another example, inferring geography from IP: Google, Adobe, etc maintain their own in-house databases for this sort of thing. Analytics solutions can be configured to send data for interactions that don't involve a page load. There are a lot of interactions that won't put anything into the logfile unless you re-invent tracking beacons. The situation gets more complicated for Single-Page Applications for both types of tracking, but generally the solutions for off-the-shelf technologies are more understood and can be applied more generally. For another thing, cost and ease of access. Google Analytics has a free version. And for businesses, many of the people who want data work in marketing or UX, not the sort of people who have various types of expertise to dig around in logfiles. At the level of effort it would take for someone to build a logfile tool or learn to dig through the data, GA would offer more benefit at lower cost. For a solo dev or small team running a not-too-complicated website, logfile parsing may make sense. For a large-enough organization, a custom solution that addresses their specific needs may make sense. But aside from a few sweet spots, most off-the-shelf tools are going to give a better benefit at lower cost. This is especially true of companies that have different teams with different needs and skills, and that benefit from integrations with other tools (like testing or personalization platforms). Also, slightly-less related... analytics load time is not always a dominant factor in not recording clicks. Most clicks on ads go through a chain of redirects through a half-dozen different services that set cookies, read cookies, attach data, increment counters, etc. There's a large window where a user can cancel navigation before a request is made to the website. Analytics should run pretty fast, unless there is a long time in-between when your server receives the initial request and when the user begins to receive HTML.
• mbesto 8y ago
  Honest question - who here has done Twitter ads and got positive ROI? AdWords/Bing/Facebook all seem to "work", but I've never found Twitter to be fruitful..at all.
  • everythingswan 8y ago
    I've run lead gen campaigns there that were almost comparable to Facebook but can't they "worked". I think the problem is, as an advertiser, treating it like I do the other social channels. I don't feel like it's similar to Facebook at all and has become even more different over time. Pinterest has similarly fragmented. I see using content as remarketing ads to be effective there. But that's the hardest ad-use-case to prove ROI on, in my experience.
  • beerlord 8y ago
    I've found Twitter to be great to communicate with other professionals in my industry. Facebook to actually advertise to my customer base.