Help us by filling in your choices.
For me none of those options, the asset store is full of those.
If i may ask i’d like to see
-
Good FPS Hands with animations (different weapons reload, grenade throwing, knife, punch, etc) there are already high quality FPS hands but with little or no animation.
-
A complete modular art pack for a Tower Defense Game with models having 5 or 6 upgrade levels.
I am an asset store seller and 3D modeler, but also supportive to anyone else who wishes to sell. Something that I’ve heard requests from (and what I’ll eventually be working on, but your welcome to beat me to it!) 
are Vehicles… and of that, would be especially spacecraft. If you make modular spacecraft parts for people to create their own ships, I believe you may make a pretty penny off of that!
From above post ^
I agree with b4c5p4c3 that more tower defense models would be a great buy! There is one great pack already released on the asset store, but the more the merrier, as everyone enjoys a good TD Game here and there!
+1 for FPS hand animations. I think a big library of animations would do very well since there’s a lot of demand and little to no competition.
Other than that, specialize. Choose a genre and build a lot of art packs for it. Once developers start building their games using, for example, your 2D isometric fantasy village pack, you’ll be their first choice when they need matching 2D isometric fantasy city packs, dungeon packs, townsfolk packs, orc packs, wild animal packs, etc., to fill out their game.
One of the challenges for game developers using the Asset Store is that it’s often hard to find enough assets that have the same art style and level of detail.
But this means your team may be working in a particular art style for a long time. Rather than following what customers vote for in a poll, decide yourselves on an art style that you can still remain passionate about for months at a time.
^ this. Great observation, Great advice.
I too absolutely agree with this.
If I were you, the other big question would be whether to produce assets in familiar styles, which (apparently) have broad appeal, or to create assets in less common styles. For example, I’ve been looking for environments for a sci-fi game. I am not a big fan of the Mass Effect art style, but apparently everybody else is, since virtually all sci-fi environments use that style. (Well, that, or industrial.) I don’t really expect to find something exactly like what I want (which is currently something like the sets in Forbidden Planet), but it sure would be nice to have a greater variety of styles to choose from. But maybe those uncommon styles just don’t sell. I don’t know, but it’s my biggest frustration with the asset store nowadays.
2 Likes
Mine is with LOD, new techs, texture formats, taxomony…
Meh, I’ve started so I will rant:
It’s one thing to find 5 models in matching styles.
Quite another to get them all with nice LOD and matching skeletal rigs even in some cases.
Some have curvature data, some lack even normals, some are a complete mesh, some are parts
Suddenly you have a great idea … but … “Now I find I need 10 variations of the NPC, 20 of the actual story characters,” et cetera… Oh boy.
I would say this is not an atypical situation for most beginner game designers and even the more seasoned get Spec Doc creep (although they likely don’t have this issue).
Thank goodness for /some/ standards peeping up, please adhere to them and try to make a “Standard” that people can expect from you. Like:
Best Practices?
Textures - decide on the sizes and formats you will support. stick to them.
Include linear and gamma versions of things if required/appropriate, consider substance support, embed tweaks where you can.
Materials and their naming conventions - choose a working standard,
also limiting multi-material definition/variations as much as possible to reduce final PBR shader build size (Don’t go down the “Here it is in Red, Blue, Green, Orange, Pink, Purple, Lilac…” road that many authors try to bulk out their work with)
Sizing, go with 1:1 - unless you have a really really really good reason.
Formats, choose one (fbx is popular) - DO NOT include meshes that require a -90 x rotation to make sensible (maya export bleugh), it’s careless and non-helpful in many situations, consider making your workflow also produce stuff for U4 and other engines as a result of your work (rather than extra effort)
Auto-naming skeletal and texture conventions (use them), the Allegorithmic tools will use that info, include a material_id map if you can. (That makes re-working SOOOOOOOOO much easier)
Folder taxonomy, pick one. Stick to it.
DO NOT USE ANYTHING FROM STANDARD ASSETS TO DEMO THE WORK (future nightmares), ever. period.
Exception being maybe the character controller code. (but don’t expect it not to change)
Avoid 3rd party shaders or assets in demos, it’s just pain, people won’t have it, it becomes unavailable or bit-rots.
use your own stuff.
always include clear animation timing readmes(when applicable) and general model construction readmes, Don’t expect anyone to know ANYTHING about rigging, modelling, unity, or even 3D graphics in general.
Include a default 1st person and 3rd person controller for your character rigs, it’s not hard and demonstrates you might actually have tried it IN GAME. Plus you can just say “Use the included controller” 1 million times when supporting instead of trying to explain how to implement a character controller 1 million times. I know which I would prefer if I was a support worker.
finally about style:
To be able to create assets that “Work” in a sci-fi game, and assets that work in a top-down rpg, and assets that work in an ancient castles tower defense game is eminently “doable”, however if you can make all of those assets immediately “Look like your work” and still fit in all those genres you are on a winner.
Since that will help “merge” together your assets better when someone is Jamming the castle asset, set in the Year 2567, featuring the barbarian from your RPG assets 
Good luck.
2 Likes
Useful tips. Thanks guys.