Guest Post: Dan Austin on Google Mapmaker Categories

Categories have been a mess in Google Map Maker for years, as well. GMM Issue Tracker allows you to add categories for potential inclusion into GMM (which hopefully bleeds over into Local): http://goo.gl/OT3VO and star for the ones you like the most. Unfortunately, GMM, like so many other ‘community’ initiatives they’ve sponsored, largely ignores the input of mappers in favor of big bug fixes, UI upgrades, pet projects, and of course, figuring out ways to lessen the people part of the GMM community, whether it’s through limiting communication with mappers (that seems to be a big initiative, right now, especially in the forums and on individual edits) or increasing the bots ability to manage the system independently. AKA havoc!

Anyway, mappers have been requesting that Google re-organize the categories into a better system, clean up the categories, add more categories, restore missing categories, correct buggy categories, and create a category hierarchy, which you can see reflected in this forum post: http://goo.gl/GddFb; this spreadsheet: http://goo.gl/M4XMX by a power mapper (not Google); and this Facebook discussion: https://www.facebook.com/groups/mapmaker/ which Rich Hintz manages, and who discusses GMM category problems a lot.

This is complicated by the fact that Local has its own set of categories, some of which are ‘invisible’ in GMM and vice versa, and some of which can only be edited in one interface or the other. Local adds its own categories automatically, for example, turning Park into Park, Parks. There’s a longstanding bug or ‘feature’ that appends gcid:[name of category] to categories that you add through Report a problem, which don’t appear that way in Maps, but are visible as such in MM. I assume that this is because the Maps categories, even the ones you select from the dropdown, are not recognized as such standard categories in MM, so it appends gcid: to indicate it’s a custom category. As happens all too frequently, the POIs in one database can become out of sync with another, resulting in islands of categories, and trying to get them to sync can be a real chore.

There has been an effort toward changing the UI for Maps community edits to be more like GMM, but it will also end up orphaning categories that are only visible in the current UI scheme for community edits, making them inaccessible to everyone but Support.

Additionally, GMM has a Primary category, and Local doesn’t (just like GMM has a field for suites, and Maps basically doesn’t), so arranging the categories can be a delicate affair, especially if you want one of two categories visible on the Local page. (Good luck with that, as whatever you change in GMM appears first).

Last, some categories are locked in GMM and Maps (like Locksmith and Military Area), locking the entire POI from any kind of editing on any interface except for the Dashboard. So basically, you have to use Report this in GMM and try to explain what you’re trying to do in order to ‘fix the categories’, or Other in Maps Report a problem. Sometimes this works, sometimes this doesn’t.

So, usually what I do is go through and try to rationalize the categories myself. If it’s claimed (and this is not always clear from GMM, since there’s no indicator that a page is claimed), I try to mess with the categories as little as possible unless there’s category spam, in which case, I’ll delete and/or replace with more appropriate categories, including custom categories. I try to use standard categories whenever possible, set a primary category that explains what the business primarily is, and then build out custom categories that exceed the five category limit if the listing mandates it. Since all those categories are searchable in Maps, even if they’re not visible, it can enhance a business listing’s ability to turn up. And even when I make all these changes, business owners of claimed listings may not approve of the changes, so that can end up screwing up the categories even more as the changes may not sync back to GMM.

Between the bots adding junk categories (that has been a big problem with hotels in Vegas, as bots aggregate tons of bad categories from unrelated features to the larger hotel POIs, resulting in a big mess of categories), clueless business owners listing product types rather than what they do (i.e. Swatch instead of Watch Store), and Google’s inability to rationalize categories and harmonize all the different databases with different category types, you have a lot of problems.

Please consider leaving a comment as your input will help me (& everyone else) better understand and learn about local.
Guest Post: Dan Austin on Google Mapmaker Categories by

6 thoughts on “Guest Post: Dan Austin on Google Mapmaker Categories”

  1. @Dan
    It sounds like you’re suggesting it’s a good idea to seek a few more categories beyond the ‘standard’ ones to enhance Google’s ability to to find a listing when the business offers a varied array of services? If so, what is the best process to have these mods implemented in MM?

    This is regarding your point: …and then build out custom categories that exceed the five category limit if the listing mandates it. Since all those categories are searchable in Maps, even if they’re not visible, it can enhance a business listing’s ability to turn up.

    Thanks :-)

  2. Great stuff, Dan.

    I’m wondering the same thing Andy is wondering. Also not sure how your advice regarding custom categories dovetails with (what seems to be) Google’s retirement of them.

  3. You simply type them in, and GMM will approve them. I usually try to find a category that best fits using the current scheme, but there’s essentially no limitation to what you can type in there (and I can see that from all the POIs coming from Places, that Google Places approves almost anything). I also avoid keyword type spam, which as any good SEO can tell you, you know it when you see it, but it can be hard to explain. The only hard ‘rule’ you have to follow is that some categories are not available as Primary categories, so you simply type in Establishment/Point of Interest (which is invisible on Places, and is basically a non-category) and use whatever category you want as an additional category. Finally, use up that space as much as you can (with standard categories first, then custom categories), and if you need more space, add more categories beyond the original five.

    As for the sunsetting of custom categories, I don’t know if it means:
    *Your POI is going to be flagged and suspended (worrisome);
    *The categories won’t appear (possible, but at present the categories are still searchable from my tests, even if they’re not visible on the Google+Local page).

    What I don’t know is how exceeding the five category limitation affects what’s in the Dashboard under the current (or future) scheme. Some experimentation is in order.

    I also don’t know why Google is sunsetting custom categories if the current category scheme is so inadequate and if there’s no clear process for submitting new categories that Google can add to the present list. Without that process, SMBs and SEO experts will just find another way to keyword stuff the rest of their listing or website with the necessary content, and given the current state of Google’s spam detection algos, it would stand to reason that you have a good chance of not getting caught.

  4. Thanks Dan :-) *I’m in Canada
    I will continue to do this, although lately, when pointing out spammy listings, I’ve noticed several denials via “Google Automated Moderator”? These are legit ‘corrections’ that are clearly explained?

  5. Hi Andy,

    I should have prefaced, ‘if’ they approve them. If you can catch a human moderator, then the changes will (usually) get pushed through, but it can be hard to figure out what changes will be approved by a person, and what will be approved by the Mod Bot. The Mod Bot algos are not so good, even as Google has become increasingly dependent on it to carry out a lot of low level approval work. In Map Maker, if you don’t get the desired results by direct editing, using Report this will send the report to a human, and the turnaround times have gotten much more reliable than the recent past (their previous response has been sometimes to never). I can’t speak very highly of Report a problem on Maps, even though I’ve made extensive use of it–it’s just not a very reliable way of dealing with problems, for a variety of reasons–listings are buggy, edits can create unexpected changes, changes can be ignored or denied for the wrong reasons, etc., so I try to avoid it unless I can’t affect the desired changes in MM. There’s use case scenarios that make it easier for dealing with one UI or the other, and it’s dependent on the outcome you’re trying to achieve as well as how persistent you’re willing to be in reaching the end goal. I’ve found that with Google Maps products in general, you have to be willing to hack it because of the many bugs teeming there, and be extremely persistent in reaching your end goal–it’s rocky terrain, with many unexpected and unpleasant pitfalls. Knowing the general public’s attitude, if it doesn’t work the first time, it’s their last time.

    Incidentally, Canada is currently using a slightly simplified Map Maker UI for Maps edits (Report a problem/Edit Details), which solves some problems, but brings other problems, as it orphans some listing data that can only be fixed through Support or the old UI, since Maps uses a different database/dataset than MM. I haven’t figured out the solution to that problem yet, other than to hope that they unify and sync their databases to a greater degree so that changes in one will propagate to the other, and what you see is what you should see, not just one aspect of it.

  6. Take a drink every time Dan uses the word, “category” or “categories.” Bah ha ha! (42 times…)

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Comments links could be nofollow free.