What's with the new variable product column widths???

Discussion in 'Shapeways Shops' started by draw, Apr 14, 2015.

  1. draw
    draw Well-Known Member
    So I would imagine that this is a new feature that is thought to be a good thing. Depending on the width of the display window the number of columns of products appears to now be variable going up to four columns as opposed to previously having three fixed columns.

    Of course this change comes after I put a lot of work into stabilizing my store with groupings of product elements into groups of three columns (including the use of placeholders) throughout my store and all product sections. Now the views can be all mucked up depending on the user window size. Is there a store setting somewhere where we can option into the new mode or stick with the previous 3 column style?


    ETA: The new format also means that the product title lengths are variable, showing fewer characters as the number of vertical columns increases. How about showing the full product titles on multiple lines regardless of the number of display columns?
     
    Last edited: Apr 15, 2015
  2. stonysmith
    stonysmith Well-Known Member Moderator
    It's a bit worse.. when it switches to 3 columns, the page count at the bottom doesn't change.
    That means 1/4 of your shop is totally hidden from the potential buyers. In two column mode, HALF of your models are not visible for sale.

    I don't mind 2, 3, 5, or 20 columns.. I just wish the page didn't have so much wasted white space left and right on my 1600x900 monitor.
     
    Last edited: Apr 15, 2015
  3. draw
    draw Well-Known Member
    The same number of products (12) are maintained on a page as the columns change so there's no need for the number of store pages to change at the bottom. However a horizontal scroll bar will show up and you might need to slide that over to see all of the products on the page. (At least when using Chrome...)


     
    Last edited: Apr 15, 2015
  4. Bathsheba
    Bathsheba Well-Known Member
    The problem I'm having with it is that the smaller column widths the thumbnails are enlarged past their resolution limit. Jaggies are clearly visible. I see this in Firefox.
     
  5. stonysmith
    stonysmith Well-Known Member Moderator
    I have 39 pages of models when in 4-across mode.
    A you can see below, with chrome Version 41.0.2272.118 m, the scrollbar won't go any farther to the right.
    Image1.jpg
     
  6. draw
    draw Well-Known Member

    Look at your vertical scroll bar, it is not at the top. f you scroll it upwards you'll find six more products higher up on the page. Your Bucyrus product is the 12th model on the page no matter how many columns there are. The number of rows are increasing as the number of columns decrease.
     
  7. stonysmith
    stonysmith Well-Known Member Moderator
    I'm happy to stand corrected :)
     
  8. draw
    draw Well-Known Member
    Here's some info as to how my store is being affected by this. With the old fixed 3 columns approach I set up my store as shown below to group similar objects into multiples of 3 and use placeholders to fill out missing spots. Not only did this approach keep store sections to stay grouped but items also remained grouped properly when viewing all elements of the store. A store section would always start in the 1st column and end in the third column. I worked around the recent policy to not display not-for-sale items by uploading a dorky model to the "placeholder" elements and making them for sale. Not ideal but it kept things grouped properly. The placeholder elements also allow me to set aside space for associated models to be added later when I find the time.

    napkin_ring_display_example_3col.jpg

    Now with the maximum of 4 columns plus variable columns everything is again mucked up as it was when the not-for-sale no show element policy was instituted. So now the groupings and organization techniques I was using are no longer valid and in addition the title lengths are smaller making it harder to distinguish differences between models. I suppose I could shorten model titles but it makes more sense to me to provide a capability to have enough display room under a model to display the the full maximum of the possible model title length, probably at least one more line or perhaps 3 lines when things are squooshed down to narrow 4 column widths.

    napkin_ring_display_example_4col.jpg

    As far as fixing things now I could adjust groupings and placeholders to reflect the 4 column maximum and take a hit if someone views on a display that forces them down to 2 or 3 columns. (I assume this change is being implemented to improve store viewing and shopping on smaller tablets or cell phones.)

    Another solution would be to use something like a model container that can hold all the variants of a design within it, sort of like a single model container element that would allow purchase for several ring sizes without having to display 10 different ring size files. Such an element would dramatically simplify store displays for things having multiple variants related to wall thickness, size, multiple hanging loop sizes, etc. It's not clear to me that there is a good solution for this type of thing right now.

     
  9. stannum
    stannum Well-Known Member
    But the preview size seems to increase, forcing a scrollbar (and as pointed by Bathsheba, bluring) much like if there was no rearrangement. Doesn't make much sense in the end to have 4 columns of ~200 pixels (800) vs 2 of ~300 (600 instead of 400).
     
  10. draw
    draw Well-Known Member
    I agree, the resizing doesn't seem to be a good thing unless they are targeting a better shop interface on a particular device such as an iPhone, iPad, or other tablet. The resizing may not happen in all situations, or perhaps it does and it is inadvertent. Hasn't most e-commerce moved to mobile platforms on a percentage basis? Relying on a generic PC sales interface is probably a money loser in 2015 so I can see the need for improvements.

    I'm just kind of exhausted by "improvements" being pushed out without warning, without apparent testing, or the ability to opt into them as opposed to just being stuck with them for months until they are repaired or modified. I can't imagine how older stores are mucked up by now if they haven't received attention over the span of the last few months or year or so because of poor backwards compatibility. Pricing, mark-ups, default images, and plenty more must be way different than how people may have left them.

     
  11. draw
    draw Well-Known Member
    It looks like the product images are no longer changing as display columns change, but the number of characters of the model title do change, with the maximum number of model title characters being present for the 3 column display state.

    I'm still stumped as how to maintain model positioning coherence with variable columns. Could someone from @Shapeways chime in and let us know what the longer range plans are for store page displays? I have to readkjust things around but I hate to do it if will only be a temporary stop-gap measure.

    One thing I'm fairly certain can be done right now is to set store sections to contain multiples of 12 items. At least that keeps related items on a single page which is a bit better for comparing similar yet different products by the customer.
     
  12. HenrikRydberg
    HenrikRydberg Shapeways Employee Design Team
    Hey guys.

    Thanks for raising the concern here. Your assumptions are correct: this change is part of the efforts to make Shapeways shopping mobile friendly. We rolled out this change last Friday relatively silently, giving ourselves the weekend to find and fix any surprise snags. I'm sorry the announcement side on this was weak.

    With this change, the page and the product grid is being optimized to any page width the customer might be using. On a mobile screen, you have 2 products next to each other, some tablets show 3. At maximum on wide screens we sport 4 products side by side. The total amount of products on the page always stay the same.

    Our tests show that more products on a single view results in better browsing experience (and adds more products being viewed). So this is where the rational behind the change lives. We're also standardizing this viewing experience across our browsing tools. Last week we updated our search view to show products the same way as in your shop. I hope you too find it much much more friendlier viewing experience ( https://www.shapeways.com/search?q=lightning ). In the near future, we will roll out a new Shop browsing experience where this theme will continue... and hopefully we can fill more of Stony's white screen space ;)

    As you might know, Google starts penalizing sites that aren't mobile friendly tomorrow (April 21st). This is why we moved fast on getting the change out. Google now accounts Shapeways.com as mobile friendly so you and we all will keep our SEO scores. (Even rise if competition doesn't respond to this requirement.)

    Draw: The idea of grouping products in a single page (via container is interesting). I'll voice that to our team. I see you're trying to solve variants. We've done A LOT of thinking in that area and believe we have an exciting solution in the works. But it's too early to talk about yet :(

    One area I'd like your input guys, is having 12 products per page. As I stated, more products per page results in nicer experience. What do you think if we'd increased the number to 48 products per page? Yes? No?


    Best,
    Henrik
     
    Last edited: Apr 20, 2015
  13. Bathsheba
    Bathsheba Well-Known Member
    Yes. 12 is not enough.

    Customer is already surfing an image-intensive 3D printing site, I think it's safe to assume they're not doing it through a 300-baud modem and would like to actually see the content they've requested, not a little teaser slice of it.

    Thanks,
     
  14. stonysmith
    stonysmith Well-Known Member Moderator
    I'd very much like to see 48 or more products per screen.
    Remember the days when the customer could pick 12 or 36 for themselves? <grin>
     
  15. draw
    draw Well-Known Member
    Thanks for the feedback. It makes sense to be multi-platform and Google friendly.

    The main issue I have with the store views is that store section views can arrange themselves differently in the full store view. The store sections can chaotically blend with each other in the full store view, meaning that a store section can start anywhere within a row. And that makes things a bit messy looking and unclear as to what is what. I'm not sure what can be done for this particularly since divisions of 2 and 4 can be made to work together with work arounds but having a division of 3 makes things more problematic. Here are a couple of ideas:

    1. In the full store view force a model "carriage return" between store sections. At least that resets the start of individual store sections to the the first column.
    2. Insert a thin horizontal banner between store sections to denote the new store section with a <store section> name. If a store section is being continued from a previous page the top of the page could start with a "<store section> continued" banner. This is essentially idea 1. but with the added banners.

    This is less of a problem for focused shops where almost all products can be grouped into one or a few sections, but for us attention deficit designers who design on a whim store sections can practically range between 1 and hundreds of models. As far as increasing models on a page to 48 or whatever I think it would be fine as long as something identifies and differentiates store sections on the page.

    The variant problem is mostly a different issue although it contributes to messiness and affects how store owners might organize their stores.

    Tonight I will make a valiant attempt to reorganize my store based on the 12 models per page style. (I will pretend I did not hear about 48 models per page.) So these are some of the methods I will be using.

    1) Set store sections to have multiples of 12 models.
    2) Break up or combine store sections to make it easier to target the 12 model multiples.
    3) Use my dreaded placeholder elements to pad out extraneous empty spots in store sections. (do a model search using "placeholder" and you'll see why I find them to be a dreadful work around)
    4) Edit default renders to include more information about model variants. This is especially important since I try to put variant information in the model titles, but since they are truncated the variant information needs to appear somewhere.
    5) possibly more to be added as I massage my store.
     
  16. Bathsheba
    Bathsheba Well-Known Member
    I don't think Shapeways ever intended to support organization of stores by rows of fixed size. They've certainly never said so. I'm not convinced it's reasonable to expect support for that going forward.
     
  17. draw
    draw Well-Known Member

    That's probably true, but since things were limited previously to a certain fixed row/column format that's what I leveraged and why I adapted my store to that format. Having no major talent for ESP I couldn't foresee future changes so that's why I am where I am now. It looks like in the near term at least the 12 items per page (or an integer multiple of that) will be maintained so I can work with that for now. I don't mind variable rows/columns as long as I know what's coming down the pike, and improvements in container elements (sort of like the variable ring size selector tool) or highlighting store sections would simplify things greatly. I can always be creative and make life more difficult for myself prior to every future format change. ;)

    Anyway in 2 weeks I will be gone for months on my walkabout (not Bali) so I'm trying to lock down my store in some format that may or may not be a good choice for when I return. It feels like stabilizing a patient before taking hoisting them down the side of a cliff in a rickety basket blowing in the high winds.
     
  18. Bathsheba
    Bathsheba Well-Known Member
    Maybe you'll feel more secure if you don't build your store around a number that no one has stated will be consistent, and that is in fact under discussion for change at this moment?

    I know full well it's a problem that sections don't work well and that organizing one's store in a friendly way is not really possible. But relying on fixed numbers of products per row, per page, per anything isn't a supported solution. Now that you don't need ESP to know that, why do it?
     
  19. draw
    draw Well-Known Member
    Already done! (well almost...) So let it be written, so let it be done!

    Having store sections with multiples of 12 models seems like a good near term compromise. 12 is divisible by 2, 3, and 4, so store sections will always begin in the first column (in full store view). If someone views individual store sections the placeholders will generally be at the bottom of the last page of the store section. Not ideal but not a killer. If they change format to 36 or 48 items per page that's divisible by 12 so positional formatting behavior is retained. If they go to 47 items per page view that's problematic but a highly unlikely change; it is likely that the number of items per page will always be a multiple of 2, 3, or 4. Anyway, all I need to do is to replace placeholders with some new crap over time and it all becomes a moot point.

    Ideally the system would automatically "carriage return" between store sections and identify sections in full store view. I won't hold my breath for such a feature to be deployed. The distraction of padding placeholders at the end of store sections could be eliminated with a new null model that would contain no model, does not display anything, but can be positioned with the shop inventory display order tool at the end of a store section to trigger a new display row if applicable.

    I'd consider doing a custom web site or somehow plopping stuff on a site like Shopify or equivalent, but those moves don't make any sense if yearly earnings are insufficient to cover the costs. I'm only throwing out ideas and playing with techniques just in case it triggers some brain juices at Shapeways for future options and improvements. That doesn't require ESP but I have a gut feeling that I should try to anticipate what might happen here over the next few months while I'm on my mega northern climate holiday tour.


     
  20. draw
    draw Well-Known Member
    With regards to showing the full model title under the models...

    Maximum model title length appears to be 50 characters.

    As the store window is widened and narrowed the columns range from
    2 to 3 to 4, the displayed title characters range as follows on my laptop:
    (note that displayed characters are also followed by ellipses)

    2 columns
    mode 1: 10 - 15 characters
    mode 2: 15 - 26 characters

    3 columns:
    mode 1: 16 - 23 characters

    4 columns:
    mode 1: 16 - 18 characters

    So if you want a better chance of the full model title being displayed it's
    probably best to limit it to 10 characters in length.

    If the full model title were to be displayed the worst case situation
    would require 5 lines under the shop preview image, although
    if the ellipses were dropped maybe it could be done with 4 lines.

    ETA: Another option might be to have 1 column of display in order to be able to display more relevant information for each model, in addition to a longer model title. Upon cursory examination it looks like in this mode Ebay has a minimum title length capacity of 64 characters and Amazon comes in with at least 80 characters.

    Is the idea behind putting more models on a page, potentially 48, related to the assumption that more eyeballs falling on more products results in more sales? One could theorize that the display of more complete descriptive information could also drive increased sales because it would prevent shoppers from having to needlessly click down into models and back out again to determine those specifics. That's wasted shopping time. There must be studies for this type of thing by now, or consultants that could help recommend the preferred approaches.

     
    Last edited: Apr 22, 2015