Acknowledged
CIQQA-4580

A Critical Integrity Flaw in the Connect IQ Store Ranking System

Subject: Concerns and Suggestions Regarding the Connect IQ Store Ranking and Discovery System

The Connect IQ Store currently lacks an effective recommendation system and sufficient safeguards against ranking manipulation. As a result, developers can potentially improve the visibility of otherwise average-quality watch faces by artificially increasing their ratings, reviews, downloads, or purchases.

Because the platform does not appear to have strong mechanisms for detecting or limiting this behavior, the most prominent positions have remained dominated by a relatively small group of developers for many years. This makes it extremely difficult for new or independent developers to gain meaningful exposure, regardless of the quality of their work.

The store’s discovery experience also feels highly repetitive. Watch faces that users have already viewed or downloaded should not continue occupying the same prominent positions indefinitely. Video platforms, for example, generally avoid repeatedly recommending content that a user has already watched. However, when users return to the Connect IQ rankings after one or two months, they often see nearly the same products in nearly the same positions.

The store would benefit from a “Show me something different” or “Refresh recommendations” option, as well as a more dynamic rotation system for featured and highly ranked content.

I would suggest considering the following improvements:

  • Introduce stronger detection and prevention measures for artificial reviews, downloads, and purchases.
  • Give greater ranking weight to reviews and activity from verified, genuine users.
  • Reduce the influence of unusually large review counts, so that a product with 10,000 reviews does not automatically receive an overwhelming visibility advantage over one with only a few dozen legitimate reviews.
  • Use actual engagement signals, such as whether a downloaded watch face remains installed and actively used after a certain period.
  • Consider retention rate, active usage, uninstall rate, and long-term user satisfaction as ranking factors.
  • Personalize recommendations by reducing the visibility of products a user has already viewed, downloaded, or dismissed.
  • Reserve part of the front-page exposure for randomized or rotating discovery, giving newer and lesser-known developers a fair opportunity to reach users.
  • Regularly refresh ranking results instead of allowing the same products to remain permanently at the top.

A healthier ranking system should reward product quality, genuine user satisfaction, and long-term usage—not simply whichever developer can accumulate the largest number of reviews or downloads.

Without meaningful changes, the current system creates pressure for honest developers to participate in the same artificial competition simply to remain visible. That is not a sustainable environment for developers, users, or the Connect IQ ecosystem.

I hope Garmin will review these concerns seriously and consider building a fairer, more diverse, and more manipulation-resistant discovery system.

Thank you for your time and consideration



8206990

  • I have been regularly reviewing the reviews of some particularly popular apps and there is a pattern of 1-2 word reviews, which when compared with something else very popular like Spotify (for example) give a very different pattern of response that should be easily identifiable in a mutual duty of care between Garmin and developers.  Potential special methods for reviews also give rise to other odd patterns such as the example below:

    “I love this watch, I wear it every day…” and “Hopefully it will work and be for humanity” … both 5* reviews, clearly unrelated to app functionality… what did reviewers think they were reviewing - the app?

  • in reply to flocsy🤠

    Just a guess, but it could be that in general there are many more DFs than Device-Apps so the device-app is more visible.

    I think many people start by developing a watch face or data field and not a device app, resulting in higher numbers of WFs and DFs.

  • I'm not doing anything outside editing the app description from time to time. Not even that any more, since we can't see the real daily download numbers. But to counter most of what others write here, here's a strange thing: I have a datafield, and an app. They're essentially the same, the goal was to allow older watches with not enough datafield memory to be able to use it as an app. My efforts were more or less put into the datafield only. Also the app was added much later. Still somehow the app managed to become more popular. I can't explain it.

  • in reply to 8206990

    LOL, my experience with Reddit is completely the opposite. I have quite a few five-star reviews "best watch face ever", but on Reddit I got 20-30% upvote ratios just within a few hours without a single comment, and my posts became virtually invisible to others. My posts followed the rules, such as having special title, badges, etc., and weren't removed for self-promoting (as if I could earn something from a free app), etc.. Meanwhile, other developers' watch faces, that even haven't been released yet, or are very close to generated "simple watch face" template and barely customizable, like they show battery level and steps, but can't be changed to show heart rate, get near zero reviews in the store, but receive a lot of "I want this" and "Looks great" comments on Reddit.

  • in reply to jim_m_58

    It is not that I have never promoted my products. In fact, some of my promotional efforts were highly successful—and that is precisely why I believe there is something fundamentally wrong with the Connect IQ ranking system.

    A few years ago, one of my Reddit posts received more than 400 upvotes and generated over a thousand additional downloads in a single day. Even with that level of genuine traffic, my product remained near the top of the rankings only briefly. On the day when downloads peaked, it still received fewer than 20 new ratings or reviews.

    By comparison, many watch faces and apps gained only one-tenth as many downloads as mine, yet their rating or review counts increased by nearly 100 every day. This growth continued day after day without interruption.

    Such patterns are difficult to explain as normal user behavior. Genuine external promotion and a large number of real downloads appear to have far less influence on ranking and visibility than these unusually persistent increases in ratings. This is exactly why I believe the Connect IQ ranking and discovery system is fundamentally flawed.