89 posts
  • Has been part of the Envato Community for over 4 years
  • Has collected 1+ items on Envato Market
  • Member of the Envato Team
  • Won the Envato Football Cup
+1 more
damirkotoric Envato team says

(The aliens are making me write this people so please forgive me)

I understand some of you are frustrated when we remove an element without knowing the reason. We have a bigger plan in mind to give you greater control of the search and browse experience so you can easily drill down to find exactly what you need. The goal is to find the right item at the right price and the previous solution was to offer an up / down arrow to let you do that. We believe that this isn’t a great solution and it is no longer in line with our goals for the browse / search experience. Considering only an average of 3% of our users actually make use of this sort arrow, we chose to remove it.

I also understand there are other frustrations with the site but I’d like to keep the focus here on the three features that I talked about in the first comment of this thread. Otherwise we will get completely off topic :)

Some changes may not make much sense when looked at in isolation. You guys are absolutely right when you say that removing the up / down sort arrow has made a negative impact on usability. However, it’s only true when looked at in isolation. Our goal with the search / browse experience is to make a significant and positive long term impact. As with any change, the transition period is going to be difficult with some aspects of the user experience. I promise you, we don’t just wake up in the morning thinking “let’s remove some things for no reason”. I agree we don’t always get things right the first time but we’re trying our best and we think carefully before making any changes, especially when removing interface elements.

Finally, we use a data driven approach to measure the impact our changes make. We will test the old and new interfaces against each other. If the new interface performs worse than the old one then we will have to rethink our approach.

2297 posts aspire to create - create to inspire
  • Elite Author: Sold more than $75,000 on Envato Market
  • Has sold $125,000+ on Envato Market
  • Made it to the Authors' Hall of Fame
  • Had an item featured on Envato Market
+9 more
Creattive says

If only 3% of the users are using that arrow, then that means you have to redesign it to make its function more clear to the user, not to remove it completely without any alternative. Instead of an arrow icon, it could be “Newest first” or “Oldest first” when choosing to sort by date, “Lowest Price first” or “Highest Price first” when using to sort by Price etc.

I cannot see how removing something essential can be good, even when you sayyou are going to make more changes, you are not talking about alternatives. And by the way, why can’t you tell us what you are about to do in the end? I mean, you are outputting incremental changes that partially look like crap but you keep on telling us that in the end everything will look fine. When people do something like this, they normally have a preview of the finished design somewhere lying around…

2276 posts Bird is the word..
  • Located in United States
  • Elite Author: Sold more than $75,000 on Envato Market
  • Has sold $500,000+ on Envato Market
  • Has been part of the Envato Community for over 6 years
+8 more
jonathan01 says

“You guys are absolutely right when you say that removing the up / down sort arrow has made a negative impact on usability. However, it’s only true when looked at in isolation. Our goal with the search / browse experience is to make a significant and positive long term impact. “

So if you agree that removing the arrows has reduced usability, then why did you guys do this when the rest of the change that will remove this concern by introducing new things isn’t ready to be shipped? Basically it sounds like this:

1 – we remove something to make it worse 2 – eventually the new shiny thing will fix it

Surely it would be better doing:

1 – remove something, introduce shiny new thing at same time and therefore we have an improvement.

I do understand things have to be done in stages, but why hurt things in the process – make the change a little bigger so the hurt is not an issue.

I appreciate the efforts being made by Envato, I’m just questioning the methods.

Jonathan

202 posts
  • Has been part of the Envato Community for over 3 years
  • Provided great feedback to improve the user experience on Envato Market
  • Provided great contribution to our forums
  • Has collected 100+ items on Envato Market
+2 more
FastNOC says

I feel compelled to respond here.

I understand your response, yet 3% doesn’t give an indication of the value of those features because it doesn’t exaplain (nor do you owe us this info) the amount of commerce that’s performed by the people included in that 3% and in all likelihood, people that do a lot of business on the site would be much more likely to be using controls like that. But this is really just picking at unknowns, the point is taken.

what bothers me, if that if you restrict the conversation to just the topic, where does someone go to discuss the other areas? the reason i brought up things that were not included in the topic is explicitly because I’ve tried without success for months to give feedback and so far it’s fallen on deaf ears.. I get frustrated because I think I’ve got opinions that have a solid foundation for several reasons. primarily because I do a lot of business here (though i will admit, my purchasing ramped up drastically in August, and it’s only since then that i have been purchasing so much). The reason for that is demand when we opened the revenue stream dictated those additional purchases, which is what led also brought my attention to the frustration I find when trying to locate items. I believe the first time I made mention of this to staff was in possibly October, and at one point I did have someone in discussion with me, then for some reason it jsut stopped. So I’ve been trying to find ways to reach the people that could use the information.

So I will apologise for bringing up issues that do not directly relate to this thread, though the topic of feedback shares a common ground, but I have been feeling more and more frustrated trying to get info, and in some cases it’s caused me to look into writing my own code to try finding things within your database. I’ve done so using Google to a point, but i heard mention at some point of an API, but can’t seem to find much info on that.

Anyway, my point is just that while I understand minimizing resource usage, I’ve yet to think of a valid reason to remove any feature that would help someone find things when the fact that it’s onlly 3% that may use something point to it certainly not being a problematic issue, so why eliminate ANY tool that allows us to find things?

it was mentioned that someone would be speaking to me at some point, I really hope that happens as there is a pretty big list of things I have not spoken of yet, because I don’t want to cloud the most important things I am looking for.

I do want to also mention something. i keep saying I can’t get responses. This ONLY pertains to two things. 1. I have asked where to find a person to speak to regarding suggestions, and 2. when i asked about the possibility of some assistance because I am doing pro bono work for non-profit groups throughout the world, and I pay out of pocket at retail prices for the pieces I need to do so. Neither requests have ever been responded to.

Other than that, for support, though it does take a couple days, I have found that I’m very satisfied with the responses I get to support from the staff on the purchases i make, and have had an overwhelmingly positive response when contacting the authors of the projects I purchase. there are isolated instances when the author may not respond, but it actually shocked me how attentive they have been most times, even when I am fully aware they have no responsibility to even respond to my questions, let alone assist. So regarding support for product, I would say on a scale of 1 to 10, 10 being the best, I would give envato a solid 9.5 on support.

I won’t respond again in this thread to subjects off topic, but I did want to explain why I did in the first place.

Thank you

6025 posts
  • Sells items exclusively on Envato Market
  • Elite Author: Sold more than $75,000 on Envato Market
  • Has sold $125,000+ on Envato Market
  • Has been part of the Envato Community for over 6 years
+8 more
VF says

Finally, we use a data driven approach to measure the impact our changes make. We will test the old and new interfaces against each other. If the new interface performs worse than the old one then we will have to rethink our approach.

Personally I have strong belief on these approaches. But as you say the impacts being measured, none of the staff considered about the fact that “Exactly at the time of these changes, why the Microlancer Ad should be added above and close to the pagination buttons?” Doesn’t that affect the process of measuring the impacts? In fact this the exact time, the additional stuff like Ad banners should NOT be added.

No, you guys are not careful about watching the impacts of these kind changes.

89 posts
  • Has been part of the Envato Community for over 4 years
  • Has collected 1+ items on Envato Market
  • Member of the Envato Team
  • Won the Envato Football Cup
+1 more
damirkotoric Envato team says

@VF Yes we are. When A/B testing we make sure that both test A and test B are identical except for the changes that we’re testing. Things like the Microlancer box and any promo banners will be present in both test A and test B, and therefore won’t falsely increase the performance of one test over another.

6025 posts
  • Sells items exclusively on Envato Market
  • Elite Author: Sold more than $75,000 on Envato Market
  • Has sold $125,000+ on Envato Market
  • Has been part of the Envato Community for over 6 years
+8 more
VF says

@VF Yes we are. When A/B testing we make sure that both test A and test B are identical except for the changes that we’re testing. Things like the Microlancer box and any promo banners will be present in both test A and test B, and therefore won’t falsely increase the performance of one test over another.

But both the A & B sides are made with unrelated insertion of Ad banner. At the time of testing, you are not actually capturing the difference between “with sort order button” and “without sort order button” but with new unrelated Ad button above the pagination could interfere.

@damirkotoric, anyway, thanks for answering, the bitterness is not particular about your work but the long trend of inefficient changes happening here on search. No worries.

7604 posts
  • Has been part of the Envato Community for over 5 years
  • Has sold $40,000+ on Envato Market
  • Sells items exclusively on Envato Market
  • Located in Europe
+5 more
doru says

@VF Yes we are. When A/B testing we make sure that both test A and test B are identical except for the changes that we’re testing. Things like the Microlancer box and any promo banners will be present in both test A and test B, and therefore won’t falsely increase the performance of one test over another.

I don’t get how the dev/marketing team go against the guidelines set by one of the founder and owner of envato. He specifically said one week ago that ads are not a good thing for search. Why did you add ads if he said nope?

Why did you undermine his position in the eyes of the community? You don’t know we will hardly believe anything he will say/promise from now on? We trusted Collis, why one chose to kill this trust for some ugly banner ads to microlancer?

Those arrows (up and down) did little to the search results. If they get replaced with more powerful filters (think amazon or ebay or envato competitions sites that I can’t write the names here) it will be awesome.

But as jonathan01 said there’s no logic in removing them now and then wait for an indefinite period of time to eventually add the promised features that you can’t talk about.

It will make sense if a new powerful filter will be launched next week but we both know it will not happen since you are testing a version without those buttons and no additional features. It will be ridiculous to test the new interface and then use it one month. If the new interface passes the tests then is here to stay for a long time till we see new improvements that probably are not even in the planning stage

Some of the dev team may not know the history of this place. We asked for search to be improved for a long time. (years) We also suggested to check powerful marketplaces and envato competitors and create something equally or even more powerful of what they have there. We are not against improving the search we want to be on the same level as those other websites. Now it looks we are playing in the mud and is not fair, we have our pride.

71 posts
  • Has referred 100+ members
  • Has sold $40,000+ on Envato Market
  • Has collected 1+ items on Envato Market
  • Has been part of the Envato Community for over 2 years
+2 more
TitanicThemes says

I also noticed that the search is not that reliable, one day the item is at the top, second day he throws me on third page, one day the item name counts more than the tags, all these with the same ratings. Some time ago made some tests and i saw that the order in witch you put the tags counts as well and now i see that the tags are alphabetically ordered.

Did you changed also the algorithm of the search?

Now i see that the name of the item it is really important, the tags don’t mean that much or at all.
I think the ratings and the tags should be the first option, not the item title.

236 posts Marketplace Developer
  • Has collected 1+ items on Envato Market
  • Member of the Envato Team
  • Located in New Zealand
  • Has been part of the Envato Community for over 2 years
madlep Envato team says

Dropping in to respond as best I can :)

disclaimer I tend to stick more to the back-end dev tasks at envato – currently search, but we all move around and work on different parts of the site from time to time. I make sure data is on servers, and servers talk to servers, and those servers send data to users happily. I don’t do much front end – so… be gentle in that regard :)


The search team has been very interested in your feedback and has been thoroughly reading your comments. Changes are going to be ongoing while they work on constructing a more powerful search. The Community Team even received an email last night from the Search Team on your great feedback FastNOC :) Sometimes we may not update every thread with a “we’re recording your ideas and using it to refine our efforts”, but let me assure you that the Search team is heavily invested in this project and very much values any feedback you can give.

we’re recording your ideas and using it to refine our efforts ;)


There should be some very granular control over searches. Some things, to me seem very obvious yet don’t exist. What I don’t understand is the tools sometime do exist on other areas. For instance, on AudioJungle I can add vars to the query to limit my search by length of the track. Just three days ago I had a requirement for projects that were under 5 seconds in play on Videohive. now, I realize that I could edit the cameras, and effects and scenes to reduce the time, or raise it if needed, but most times these projects are geared to be static in time, they aren’t dynamic (and I’m not sure how they really could be) but buying a project that is built for 20 second run is complicated to edit down to a 5 second result, and maybe not even really feasible. I gave up looking. and when I asked in the forum for suggestions…... most of the replies were simply by authors of projects, but the fact that nobody had any real suggestions is telling.

+1 definitely agree. This is something we’re looking at doing. In fact I’m working on a change right now to store attribute key/values in the search index properly to make them more searchable. Part of the challenge is that there are a lot of item types, with a lot of different attribute types, with… variable… data quality. Some great, others not so great.

Attribute searches on exact matches are relatively easy, but examples like searching for length of audio/video are trickier, as they are arbitrary values, and a range is more useful. Other attributes have a predefined set of values, others still are boolean yes/no etc. So yes, that would be awesome, but needs some more thought on the best way to achieve it.

I’m slightly curious about: “For instance, on AudioJungle I can add vars to the query to limit my search by length of the track.” I’m wracking my brain trying to figure out how/where this is done, as we haven’t coded anything in to handle that – not in the new search engine anyway. It’s probably some long forgotten part of the old one.


Another problem, that is apparently not going to be fixed by the upgrades, is sorting by sales. I don’t know if you realize this, but sorting by sales is virtually pointless. Here’s why. If I search on themeforest for a WordPress theme, and I sort by sales, I’m going to find that the results don’t show me what I need. It shows me the overall sales. Well that’s not relevant. Most times you’ll end up getting results from themes that were built YEARS ago. Because they’ve accrued a higher number of sales than anything new. But what if I’m looking for something that’s really hot right now. I want to search and sort by the sales, but I don’t care about results from projects that are over a year old, because they are much less likely to have the updates of the newer technology. This is critical in this day because with HTML5 things are very different. Qualifiers should be much broader and I for one really need to be able to chain variables. I want to sort by sales in the last 6 months. Or maybe I want to sort by sales in the last 6 months, but only projects that have woocommerce. That’s not possible. And I don’t see any of that changing with these features. The sales and ratings sorting are pretty much useless because they cannot be qualified. Sort by ratings and you end up with pages of results from files that are year’s old, sales, same result.

That’s another interesting feature that would make sense.

We do do that with the default relevancy sorting already. There is a boost on “recent sales” for an item as well as “total sales”. It’s not exposed as a sort option though. Would make sense for it to be. (I have to talk quietly here… because we’re not supposed to talk publicly about how boosts work, and @michellerobinson might tell me off :P )


I believe there is a very large number of sales missed due to the simple fact that there are not the tools we need to be able to search. I don’t understand it because I understand database queries and db management and these are things that are pretty much basic level requirements. People want speed. but not even speed is the issue here, it’s not even a feature that’s possible. now I could understand to a point that length isn’t searchable on videhohive and it is on audiojungle, because it’s probably not a tag that’s specified. but shouldn’t it be? Sometimes I want a project that’s over 2 minutes. how would I find that?

Technically, this is all served up from our search engine (elasticsearch), not the main database. But yes, point taken. This stuff should be available – and the capabilities are there, they just haven’t been exposed to end users yet.


There are a couple issues with the process of the forum, and it goes to security. Knowing this without researching you might hav security research this

The forum has a time limit on edits right? That, i assume is done to prevent someone distorting a conversation by editing their older responses.

Did you know that this is easily circumvented? I haven’t done a security review by any means, but just by looking i figured out how I can edit any of my posts, regardless of whether they’ve past the time limit.

If you want me to prove this, I can edit a post from yesterday and show you

That doesn’t sound good. Can you send me more details to julian@envato.com? I’ll see it gets fixed up.


If you want to know about search buggs I will write many of them,just ask me and tell me that anyone from devs are interested in hearing me.

I found a new Bug: write logo in search look at the results. And now go to any logo item and go to its tags and now press on “logo” tag. Results are totaly different.

I’m interesting in hearing about bugs!

That one isn’t a bug though. As others have mentioned, it’s searching on different fields. General search is on name, description, tags, + a bunch of other stuff. Tag searches are just on the tag field.


I also noticed that the search is not that reliable, one day the item is at the top, second day he throws me on third page, one day the item name counts more than the tags, all these with the same ratings. Some time ago made some tests and i saw that the order in witch you put the tags counts as well and now i see that the tags are alphabetically ordered.

Did you changed also the algorithm of the search?

Now i see that the name of the item it is really important, the tags don’t mean that much or at all.
I think the ratings and the tags should be the first option, not the item title.

This is the effect of other boosts in action. New items get a boost to give them initial exposure and a chance to get traction. This only lasts a few days, and gradually tapers off. This is especially marked for highly competitive search terms where there are a lot of items closely ranked for the top places.

Tag order shouldn’t have any effect. They’re more like a set than a list – the search engine only cares if a tag is on the item, not it’s order.

Name, description, and tags all get separate boostings, and we are constantly tweaking these so search scoring relevance stays good. It’s a continual process.

In addition, it depends on the exact tag, and how often that tag is used throughout all other items in the search index. tf-idf is an important algorithm used by the search engine – one of it’s properties is that as more items use a tag, the lower the score that a match on that tag gets. So if every themeforest theme is tagged “responsive”, then a match on “responsive” as a tag doesn’t count for as much anymore, as it’s not very unique and doesn’t add as much value.


I think we need to buy some tickets to Australia and take the control over the headquarters and access the servers till is too late

Servers are in the US though ;)


I’ll call Will Smith…

by
by
by
by
by
by