Why is it a problem that old items show up? After all, isn’t that kind of the purpose of a library? Otherwise we might as well just delete all items older than 3 years. Are you saying that no quality energetic tracks could have been made in 2009?
What we need is a healthy MIX between old best-sellers, old forgotten tracks (randomization), new tracks, and new best-sellers (trending).
You’re right, for the most part. Having the right mix between new and old tracks is optimal - for example try a search of “Indie Rock” at the moment. There are mixed results; mostly new items (which is good), sprinkled with a few high sellers (also good). However as CommercialMusic pointed out, the ‘updated’ terms such as “corporate”, “upbeat”, “inspiring” are completely biased towards old, top selling items, leaving no space for new and upcoming items (some of which could inevitably become new top sellers if they get enough visibility).
I think, if a person wants to take a look at 2009 or 2010 year items, they can apply some filters in the search. But in my opinion usually people come to envato to find present-day modern items.
There’s no problem with 2009 track showing up, the problem is it shows up first, leaving no chance for 2015 new item.
For a seller point of view - there’s no motive to produce a new track under that name
For a buyer point of view - if he already looked that phrase before, there’s no point in giving him the same item at the first search, it is not good for Envato as well, because it’s showing luck of new tracks.
Just to chime in, yes I think the “mix” is the thing to look for in the long run. Maybe it’s time to look outside the box when it comes to search listings. Instead of just having the one long list, maybe divide the page into different sections. First list the 10 newest items, then the 10 best sellers from the last month, and then 10 best sellers of all time.
I don’t know, but it looks like it’s kind of difficult to make everyone happy with just one single algorithm that will please new buyers, seasoned buyers, new authors, established authors, not to mention all across the marketplaces. Different keywords have differing levels of supply/demand and the only way to make it “fair” across the board is to make the subdivision of exposure PRIOR to the sorting of the list.
With all the facets and search functionality already in place, people looking for only new or only best selling items will still have the option to do this, but what we’re really talking about here is the opportunity for exposure when a simple, default, one or two word search query is performed. As I see it, these author groups deserve a nice shot at the first page, regardless of market and keyword:
To solve the identical name issue, further subdivision would be required, as in:
1a. New items with similar title
1b. New items - title omitted from search algorithm
2a. Hot new items with similar title
2b. Hot new items - title omitted from search algorithm
3a. Best sellers with similar title
3b. Best sellers - title omitted from search algorithm
Just give these groups equal exposure on every search page and everybody wins. I wouldn’t know too much about the technical difficulties of implementing this but it doesn’t seem at all impossible to group the items into their respective category prior to performing the searches (to avoid duplicate listings etc)
Cheers to the team working on this and good luck with the new year 
Well said! Fully agree. It’s important that there is that crucial initial exposure for new items, but it’s also important to reward those tracks that go on to sell well.
To add to your proposal, a balance between new and old item groups is a great idea. I still think a 1-column results page would work just fine though, if nothing else to maintain consistency across all the marketplaces. The results page could feature the all the groups in one list.
In my view, 50% should be new tracks (close to the previous algorithm), 25% hot / trending tracks (less exposure as they will likely have reached the popular files list), and the remaining 25% for those select “hit” tracks that proved their ground.
Something along these lines, yes!
One other thing that can be improved is how many items show up per page. Right now, only 30, and it’s pretty inconvenient to go through more than 2-3 pages… Buyers get tired quickly. Check out the biggest competitor to see how quick and easy it is to go through 100-200 tracks and instantly see the waveforms in a nice way to see if the intro/outro/midsections could fit.
I know AJ can show waveforms but I can only see 6 at a time before I have to scroll (compared to 25) and it’s way too slow and clumsy.
To sum up, with this much music, you need to at least double the amount of items on each page or introduce infinite scroll.
Can’t we just have the title excluded completely, so that creative titles can be seen again. I thought that was the whole purpose of our plea to Envato!!!
If categories and tags are used correctly then we should be half way there. Of course with so many files I imagine other facets need to come into play but ultimately the first thing that is needed is original titles so the site looks professional and enticing for buyers. Oh and a bit of a spend on Adwords.
Eliminating title entirely from search and forcing the customer to type in 5 descriptive keywords will do the trick for “randomness”.
Just test search the term “Rock”…the playlist delivered is all the popular items from elite authors and top selling tracks.
However if you test search 5 terms: rock, upbeat, happy, fun, energetic
The results are more varied, random, and give more authors a shot at being heard. I happen to use these terms as keywords to describe several of my tracks and what do you know…my tracks actually show up. I do think about what a customer would type in to find my tracks. I think we all do. Coming up with keywords is thoughtful, creative, and strategic work, but when the site ranks keywords 2nd over title…things do not work out for the author they way they thought it would.
What is the point of making us think of 30 keywords if title is all that matters?
So I encourage everyone to do some test searches using 5 descriptive keywords…See if your tracks start showing up on pages 1 or 2. Do The test a few times. This really may be a quick fix to this titling mess.
1, Eliminate Title from search
2. Force the customer to type in 5 descriptive keywords…OR…force them to chose genre, tempo, then a minimum of 3 descriptive keywords if you think 5 is too demanding. I personally do not think 3 descriptive words is too demanding…Maybe even display a list of mood words for them to check off.
Anyway, I just want to report that when I type in 4 or more keywords…I do see my tracks showing up more often…and I do see more random delivery of playlists to chose from. So encouraging the customer to type in more descriptive words may help this problem go away forever.
I do realize that one word for Photodune is best. If I need a picture of beaches, I will just type “Beaches” and get the picture of beaches I’ll need. Ditto for Skyscrapers, Cars, Mountains, Lions, Flowers, etc.
Let’s remember that 1 descriptive word NEVER describes a track. Think about it…“Rock”…well gee there are only about 200 flavors of rock. So please let’s get the customer doing 5 seconds more work by forcing them to type in or check off a few words. Everyone’s portfolio now looks like this: I am selling:
corporate
happy
fun
positive
inspire
inspired
inspirational
motivate
motivational
epic
documentary
rock
indie rock
background
cheerful
happiness
upbeat
corporative
the happy
the upbeat
optimistic
Is this really the way forward?
Music is always evolving, when you go to a website looking for music for your project you don’t purposely go to the very last page to find the oldest stuff do you? No one does because you want to see the newest, most updated stuff.
Well my latest tracks are named:
Corporate
Piano
Christmas Ident
Christmas
Corporation
Cinematic
The Inspirational (cringe)
Inspirational
Technology
Travel
Children
Inspirational Piano
I have some lovely titles for my PRO music it’s such a shame 
I agree with flossie. I’ve had buyers using my music in videos which had no real connection to my music - except the title. It really gets annoying and even I came to an point where I’m considering naming my next items “Inspiration” or “The Epic”… I can absolutely understand that a search engine can be quite delicate to deal with, but other sites get this right, so why not audiojungle?
I liked previous search algorithm more than this. Because of the new algorithm I have drop in sales.
I feel you Flossie … still holding on to the old names for my old tracks … but new ones are title ‘neutered’
Agree! A dramatic drop for me too.
…in any case, the consumer client will choose good music (modern sound, creative ideas, style, genre)…
A great impact on a client’s decision will give amount of sales.
I still wonder why those monsters elite sellers who earn thousands dollars each month, complain about the Search Algorithm ?
I am not a monster, first. Second, because i earn here money i want it to keep working, and i care about how it looks now and what impression it makes on buyers in the long term
Soundroll you are not a monster and AJ authors should respect those who have provided music that is useable and of top quality to the market. We should all be thankful for this platform and praise those authors who are reaping their just reward for their hard work…IMHO. 