Acknowledged
over 1 year ago

Address recent tidal wave of SPAM / low-quality watch faces in the CIQ store

Recently, a handful of developers are publishing dozens of low-quality watch faces each week. 

As a customer, it is making it impossible to find new, quality watch faces and sift through this junk. Why even look if it's either stuff that was launched in 2020 and has 150k+ downloads or spam that's new?

As a developer, it is making me not want to continue building high-quality faces if my products are going to get lost in the mess. Why even put in the effort if it'll get buried?

They all follow the same pattern: a digital watch face with a background image. They're all for sale via KiezelPay.  In this grouping of 24 watch faces that got published recently, there are just 7 different developers.  And if you research these developers, some of them are "linked" (just using different accounts).

See image:

https://www.dropbox.com/scl/fi/bbtpxmomn1lojvjgnquct/Screenshot-2023-12-17-at-9.19.57-PM.png?rlkey=b3txaoyceikum4kwduxj9pjt2&dl=0
(I can't upload natively, since I'm getting errors about file quota restrictions)

This problem has existed for the last few months, but it seems to have gotten really bad lately.  

There's an active Forum conversation here with a number of thoughtful contributions:

The Connect IQ Store is getting worse by the day by a few developers who dump a shitload of *** watch faces

I'd argue that adding 780+ faces to the store should be considered abuse. Here are some of the developers.

See this example of a developer with 782 faces.
(I can't upload natively, since I'm getting errors about file quota restrictions)

These developers all seem to be taking advantage of a "loophole" in the store: 

Newly uploaded watch faces get prominent positioning so they can be discovered, but the rest of the store makes it hard to discover content.  One of the only ways to "survive" economically in the store is by publishing a constant stream of new content so you're always getting in the "NEW" rankings.

I don't have one fix in mind, but here are my proposals as ideas:

1)
What: Update the ranking system to factor in the number of platforms supported
Why: Incentivize and elevate products that utilize modern technologies.

A listing with 100 downloads that supports 50 products should be considered differently from a listing with 100 downloads that supports 5 products.  

For example, imagine someone made an app that's special for the diving watches and uses a special feature only available on them.  Right now, that app will get lost because it's going to compete with everything that targets a wider range. 

"Gross downloads" are a major factor in the rankings, yet this disincentivizes the development of watch faces that can only target modern devices with newer capabilities or specialized devices.

Garmin's doing a great job giving us modern features (like radial/angular text), but it's not viable to build niche apps that target modern devices.  Some face/app concepts don't have an alternative for graceful degradation on platforms without the special capabilities.  But if you just target the portion of devices that support your required CIQ feature, then you'll get drowned out by faces built to a lower requirement.

The rankings should be updated to incentivize the adoption of these compelling new features by factoring in the number of platforms supported into the ranking calculations.  Otherwise, there's a strong barrier to the adoption of newer capabilities and advancement of the ecosystem overall.

Apple solves this by creating categories to highlight apps using new features

2)
What: Allow "infinite" pagination through the categories
Why: There are probably hundreds or thousands of watch faces that can't be found except via search

We get just 4 pages of results per category.  If a watch face isn't in the first 4 pages, the only way it can be discovered is via searching. 

Why limit customers to only seeing what's ranking highly right now?  These rankings tend to be stagnant towards the top and volatile towards the bottom.  Some great products are getting lost off page 4 if they can't compete against the apps that were published ages ago and have accumulated so many downloads that their top ranking is assured permanently.

3)
What: Enforce new-app upload limits
Why: No developer is producing 10 quality watch faces each week.

It should be harder for SCAM & SPAM developers to flood the store with low-quality items. This doesn't mean prevent developers from updating existing apps.  But developers who aren't spamming tend to upload at a much slower rate, because real quality takes time to produce.  There's no reason any one account should need to upload 10+ faces each week.

There's no great solution to prevent them from setting up multiple accounts to circumvent this.  But it might be a start.

4)
What: Give new / small developers more exposure
Why: It can be especially hard to get started here; we should welcome newbies

If someone's gone and created their first (or second or third) face, the community should give them a little extra exposure vs. a developer that's on app 750.  The community would be better served by having more developers, so we should make new developers want to keep making new things.

5) 
What: Update the terms to block SCAM & SPAM
Why: This can't purely be automated; spammers always find a way around

Garmin should be empowered to give out warnings and put blocks on accounts that have uploaded too many apps that are too similar. This would also help them weed out "bot rings" (developer accounts that are shared by the same person / entity). 

6) 
What: Add an upvote / downvote capability (Reddit-style)
Why: Users should be empowered to have a voice without needing to download a watch face

We can recruit the community to help stamp out spam and low-quality stuff.

The current rating system is limiting because you have to download the app to rate it.  Some things are obviously spam or low-quality, but you're not going to download it just to downvote it. The community is missing a LOT of value from average users being able to vote on what they want to see (and not see).

There could be a secondary system (alongside app ratings) that allows everyday users to upvote cool things and downvote obvious spam.  It could be one factor, not the main or only factor.  

7)
What: More Garmin editorialization
Why: There's a reason there are food critics and tech writers; average people want help identifying "extraordinary" products

The Garmin team should have a way to help clever or useful items rise to the top more, and they should rotate what they highlight to keep the community getting exposed to a wide variety of new things.

Apple has the "Editor's choice" for this

myaro
  • 10 months ago

    This is my "low overhead effort" proposal for what Garman should do about this.  These junk faces all have the following characteristics:

    1) Have single static graphic image

    2) Show digital time and a few simple data values such as HR, battery level, steps, calories and date,

    3) Don't have any other features other than the above, certainly nothing obviously requiring some effort to program

    4) Request payment via Kiezel Pay, sometimes via a bit.ly link

    5) Offer a bundle price

    6) Ask for a review

    Any apps that fit any 4 or more of the above 6 points should be put in a new category on the store like "Image faces", and not be allowed into the Hot & Fresh Apps category.  If necessary, set up a "New image faces" category for new faces that fall into this classification. Plus it gets easier than that because once you see more than 2 apps from the same developer that fit this filter, it is 99% guaranteed that all the apps from that developer also fit.  Their name can be put on a list and alll their apps would just be directed into the "Image faces" category.

    Garmin has set up new categories before, and already needs to do a manual review so this should be very doable with minimal to no added effort.  

    Furthermore, all such apps already in the Store should be reclassified into this category.  I see a lot of these junk faces spamming categories where they obviously do not belong.  The developer list determined using the method that I mentioned above can be used to easily do this, probably even in an automated way. 

    From what I've seen every one of the junk faces will fit the criteria above.  I've only seen 2 with an added animation and one with analog hands, but they still fit the criteria.  Other things that a lot of them do is link to their catalog of other faces, and put tags in their app description.  None of Garmin's apps or any other serious contributing or even newbie developer's work should fall into this category.

    -------------------------------------------------------------------------------------

    How can we deliver our request to Garmin?   Do we need a petition?

    Doesn't jim_m or flocsy work for Garmin?  They seem to answer almost every post.

  • 10 months ago

    I'm going to post a free watch face, put the unlock code in this post, and direct users to vote. I also call on other developers who are affected by this to launch a campaign to guide users here to understand the seriousness of this issue. Hopefully this will get the attention of the Garmin authorities. Let them solve problems faster.

  • over 1 year ago in reply to jim_m_58

    I would still like to have a life instead of manually having to flag all new automatically generated spam watch faces in the store as 'Inappropriate' every day.

    This shifts the problem to us as users who can't do anything about a few developers abusing the store. Well thought out Garmin, great!

    With all these kind of reactions I get the feeling that for Garmin this isn't an issue at all, they are probably even happy with all this spam so they can fill more categories and show off the amount of 'nice apps' in the store . I'm still quite disappointed.

  • over 1 year ago in reply to myaro

    I'd still try the "Flag as Inappropriate", as you can include a message as to why you flagged it.

  • over 1 year ago in reply to jim_m_58

    Thanks, Jim.  When I think of "Flag as Inappropriate," I think of items that are distasteful or otherwise objectionable (e.g. pornography).  That seems like a tool for a different issue, so I'm not sure that's a viable option here.  But perhaps a "Mark as Spam" option next to it might help? 

    But then again, I'm not even sure if that's going to be the right term.  Some users might indeed want 20 different versions of Santa and a Christmas tree.  This doesn't seem to be an issue at the watch face / app level.  I think the issue is more at an ecosystem / abuse level. 

    The issue is really flooding the store with low-quality items at such a high pace as to monopolize the "NEW" category each week / crowd out other developers.  It's not that any one watch face is objectionable.  It's the way developers are publishing so many of them that are nearly identical to game the system.