next »

[PatrickECS]

A lot has been happening with Crest in the past few months. This will be such an interesting game to write a post-mortem about once we're finished with it. I'll give you a bit of an insight what has happened from a marketing perspective during the last months.

Crest is now in its third year on Steam Early Access and while the second year only generated half the revenue the first one did, the third year is now about to surpass the first year earnings. November 2017 is our best-selling month since we launched on Steam - and there was no update released or any other outside influence from press, youtubers or sales.

After we had launched our last Module at the end of July, I dove deep into all the community feedback and refund notes. Based on this, I started optimizing our Steam store front in September and immediately saw a boost in sales. Unluckily, we received a couple of negative reviews in a row beginning of October, holding the sales back a bit. I continued iterating our presence, published a FAQ and started a wiki.

Since we were getting close to finishing our work on the current Module, the whole team took a day out of the usual gamedev schedule to come up with solutions for the current problems of Crest that had been pointed out through the feedback and we critically looked at the game ourselves. Afterwards I wrote an article about what came out of this meeting, including the reasoning for all the planned changes: The Future of Crest.

Beginning of November I finally got rid of the game's subtitle "Religion Simulator" (except here lol) and changed it to "Indirect God Sim". I had had a bad feeling about the original subtitle since I started working here, but left it unchanged since it was an integer part of the game and I didn't wanna change too many things right away. Seeing now how much it influenced the sales, I would have changed it immediately, but it also took a few months for me to realize that the god-game genre is now welcomed again by players, unlike the time when Crest launched in Early Access 2015, where Godus had still been a thing.

Shortly after, the negative reviews left in October were not shown anymore in the "MOST HELPFUL REVIEWS  IN THE PAST 30 DAYS"-list, giving another significant boost in sales. Trying to prevent more negative review from coming in, I placed the FAQ prominently in the Update Notes section on the store page. Another week later we announced the release date for our next module (28th November, announced it three weeks before) and placed the date visibly on most capsules, further preventing people from leaving reviews (positive ones as well though). Preventing negative reviews is very important for us at the moment, since we're currently sitting at a 72% positive rating (78% old rating) and a "mixed"-rating before launch of the new module would have been devastating.

In order to boost the reviews again, we're having a small discount on Crest a week before the launch to convert some of the whishlists, hoping they'll wait with leaving a review since the big update will be only a few days away and another week-long discount right after combined with a visibility round to maximize the launch buzz.

Here's this year's graph of Steam Units sold (daily average) to visualize the above mentioned changes. Noteworthy, the game has remained unchanged since beginning of August but we released two builds on the beta branch, last one on 6th October.



Tl;dr: Continuously optimize and iterate your store presence based on your current situation according to feedback & refund notes. It's never too late. ^^

[JobLeonard]

Expectation management is really important, and probably one of the few aspects of marketing that I'm not completely cynical about. Good work, Patrick!

[PatrickECS]

Thank you!

It was quite a delicate path to walk since I was aware of all the game's current faults but knew there would be no update coming for the next two months. Players were having trouble understanding the game, so I updated the Steam description first to reflect the game better, which resulted in more sales but also the refunds went up rapidly because the words, what we think Crest is or like it to be, didn't reflect the current state of the game. After some more iterations, the refunds went down again, which was quite a relief. ^^

I hope Steam will some day introduce A/B testing, it's a feature I'm really missing and that would make the whole optimization less troublesome.

[io3 creations]

If it ain't broke, don't fix it

Sometimes, it's obvious what is broken and needs fixing.

Based on your results, it seems like you also managed to "fix" some things that give better results. 


p.s. I'm wondering about having TL;DR at the end.  Someone asked it recently on Reddit and there seem to be two meanings:
-Executive Summary (essential quick takeaway. placed at the top)
-Summary (at the end)
Personally, I prefer the "Executive Summary" version at the top, since if I don't have time or interest to read it, chances are, I won't see it.  Also, on some sites (e.g. Steam reviews) only part of the post is displayed.  There it is even more useful to have it at the top.  The main benefit of having it at the end is in situations where the summary might contain spoilers and is best placed at the end.  But if it's meant to be "if this is the only thing that you read", then it's better at the top.

[PatrickECS]

I've only ever seen TL;DR at the end of a post and treat it myself as a summary. Usually you start reading, decide it's too long, doesn't capture you enough, start scrolling down to see if there's something else catching your eye, until you reach the end and are presented with the summary.

[Greipur]

Hey all! I return from the shadows to post our new trailer.








I try to push the envelope with each trailer, this will probably be one of the last ones we do, but I'm already eager for the launch trailer, hopefully we can make it even better. I made a quick summary of how I make one of the sequences on Imgur, it's for a general audience so the TIG-community probably don't find it that interesting, but maybe you've missed my meandering pieces.

I miss being able to post here more often, but right now duty calls, daily usually.

[JobLeonard]

Stylish as always

[Greipur]

Thanks Job! I got very inspired and will try to apply some of the shader know-how into the realtime version of Crest, AKA the game. We'll see how that goes.

[io3 creations]

Hey all! I return from the shadows to post our new trailer.

Ah, you mean the Underworld as depicted in the game.  So, it's not just a fictional place after all.

The trailer is looking good!
I noticed a few things:
At 0:34 when the camera is moving toward the islands, is it just my computer or is the video playback somehow not very smooth? (Could be due to many of the items being generated).
At a few places when there's text on screen (with dropped shadow as partial outline), if the background is bright then it can be a bit hard to read.  You've probably seen other trailers where the text sometimes has it's "own scene" with perhaps some limited visual background.  Or sometimes, there is some kind of visual that surrounds the text to make sure that there's always contrast.

I was going to mention it a while back, but this commercial always keeps reminding me of your game



Can you guess the reason's?

[Greipur]

Hey all! I return from the shadows to post our new trailer.

Ah, you mean the Underworld as depicted in the game.  So, it's not just a fictional place after all.

Ssssh! Not so loud. 

The trailer is looking good!
I noticed a few things:
At 0:34 when the camera is moving toward the islands, is it just my computer or is the video playback somehow not very smooth? (Could be due to many of the items being generated).

Yes, framerate issue, we were pressed for time and put it in hoping it wouldn't bother too much.

At a few places when there's text on screen (with dropped shadow as partial outline), if the background is bright then it can be a bit hard to read.  You've probably seen other trailers where the text sometimes has it's "own scene" with perhaps some limited visual background.  Or sometimes, there is some kind of visual that surrounds the text to make sure that there's always contrast.

Yes, I believe they are called "title cards", hearkens back to the silent movie era. I usually find them quite overdone in game trailers, you know the one "BECOME AN EPIC HUNTER WARRIOR!" and then you see 4 seconds of gameplay and then "SMASH THROUGH YOUR ENEMIES WITH A SPOON!" and then spoon action. It works, it's clean, but I want to explore other ways. You're right that it didn't quite work in all scenes, I will ponder about it for the next trailer.

I was going to mention it a while back, but this commercial always keeps reminding me of your game



Can you guess the reason's?

Interesting storytelling device, I see why. The newest trailer shares some similarities what with running through time, but it's more clear in this commercial. Maybe I should try the concept of visualising time in the launch trailer as well? I think it was a bit hard to read in the latest one.



---

Thanks for praise and feedback! 

[io3 creations]

Ssssh! Not so loud. 

Not to worry!  Most people will would believe it as much as an atheist believes in God.

The trailer is looking good!
I noticed a few things:
At 0:34 when the camera is moving toward the islands, is it just my computer or is the video playback somehow not very smooth? (Could be due to many of the items being generated).

Yes, framerate issue, we were pressed for time and put it in hoping it wouldn't bother too much.

That part wasn't a biggie.  You might have seen Time.captureFramerate in Unity https://docs.unity3d.com/ScriptReference/Time-captureFramerate.html    to export screenshots as images.  That way you can get a smooth recording ~ in case you need it for certain situations.  Though, it probably wouldn't have helped much due to the time constraint you mentioned.

At a few places when there's text on screen (with dropped shadow as partial outline), if the background is bright then it can be a bit hard to read.  You've probably seen other trailers where the text sometimes has it's "own scene" with perhaps some limited visual background.  Or sometimes, there is some kind of visual that surrounds the text to make sure that there's always contrast.

Yes, I believe they are called "title cards", hearkens back to the silent movie era. I usually find them quite overdone in game trailers, you know the one "BECOME AN EPIC HUNTER WARRIOR!" and then you see 4 seconds of gameplay and then "SMASH THROUGH YOUR ENEMIES WITH A SPOON!" and then spoon action. It works, it's clean, but I want to explore other ways. You're right that it didn't quite work in all scenes, I will ponder about it for the next trailer.

Ah, those "text things" do have a name!   Thank you for letting me learn something new.   Interestingly, wikipedia has it as intertitle (also known as a title card).  Though, I wonder if it's technically the same thing if you have video showing behind the text?

I agree that you don't always have to with the "tried and true" approaches ... as long as the new approaches accomplish the same objectives and not create issues.  Near the beginning of the trailer, in terms of timing it was fine when the title card appears after some on screen event (though one line had the contrast issue with the sea).  But later at 1:09 the issue is attention related because the text is shown only for a brief time and it's like as if you have to chose between looking at the title card or watching the on screen events.  Then, if I can't rewind (or don't want to due to time), I find that I tend to wonder if I missed something important/relevant ... which leads to more confusion and missing parts and confusion.  Another example would be a large amount of text (e.g. at 1:13) or in other trailers when multiple dialogue options are displayed on screen but not enough time given to read them.  So, the title card > action > title card > action > etc works because attention is directed by showing only one thing.  In a way, showing in-game interaction can have the similar "attention splitting" issue if title cards are also shown or there are way too many items on screen that the viewer sees the first time.  At least, these are my pet peeves.

Interesting storytelling device, I see why. The newest trailer shares some similarities what with running through time, but it's more clear in this commercial. Maybe I should try the concept of visualising time in the launch trailer as well? I think it was a bit hard to read in the latest one.

That's interesting.  Actually, I was thinking more in terms of the themes.

In terms of visualizing time, that will probably depend on how you want to do the trailer but a "start small, then show various things as you progress

So, the live-action trailer is coming for Crest 2, right?


Did you manually animate the randomized low poly faces (e.g. changing "low-polyness setting" ) or did you use some kind of scripting?

Have you considered using one of those online renderfarm services for rendering the videos?

Thanks for praise and feedback! 

You're welcome!

[PatrickECS]

So, the live-action trailer is coming for Crest 2, right?

Our Community is way ahead of you. :D

[JobLeonard]

A very interesting disclaimer under the video:

Eat, Create, Sleep did ask me to make this content, but I was not offered pay.

It's like saying "yes, I'm biased, but I'm not a sell-out"

[Greipur]

"attention splitting"

Yeah, good point.

Did you manually animate the randomized low poly faces (e.g. changing "low-polyness setting" ) or did you use some kind of scripting?

That's the decimate tool in Blender. I don't know if you've used it but you can basically animate 90% of all values in the program which is a really cool feature imo. I usually come up with fun experiments that way. You basically have a slider from 0% to 100% polycrunching and I just added keyframes to zero decimation to say, 50% decimation. Not super pleased with the oucome, but it was an interesting experiment.

Have you considered using one of those online renderfarm services for rendering the videos?

Not really, but we've 7 good custom built desktop computers which we're planning on linking together with a free Blender render farm solution in the future. If you leave them on during a night or two you can basically render 1-2 minutes without problems.

[PatrickECS]

A very interesting disclaimer under the video:

Eat, Create, Sleep did ask me to make this content, but I was not offered pay.

It's like saying "yes, I'm biased, but I'm not a sell-out"

I think when he uploaded the video the "sponsored content"-debate had been heating up and everyone was overcautious with that kind of stuff.

[io3 creations]

So, the live-action trailer is coming for Crest 2, right?

Our Community is way ahead of you. :D

So, the live-action trailer is coming for Crest 2, right?

Our Community is way ahead of you. :D

[io3 creations]

Did you manually animate the randomized low poly faces (e.g. changing "low-polyness setting" ) or did you use some kind of scripting?

That's the decimate tool in Blender. I don't know if you've used it but you can basically animate 90% of all values in the program which is a really cool feature imo. I usually come up with fun experiments that way. You basically have a slider from 0% to 100% polycrunching and I just added keyframes to zero decimation to say, 50% decimation.

Now that you mention it, I might have seen it when looking for something.  However, at this point I'll primarily use Blender to create models and use Unity for animations (e.g. cutscenes, character animations, etc).  I remember the time when moving from hand animation to cg keyframe based animation was a great step in terms of productivity.  But now, I prefer to use more scripting for automating animation as much as I can (especially for repeating animations) and use keyframes sparingly.  Another benefit of in-game animation is that you don't have to worry about using video files with the game (size, compatibility, wrapper, etc).

Not super pleased with the oucome, but it was an interesting experiment.

I like exploring various low-poly ideas, so for me it was quite interesting.  I'll probably use polycrunching for brainstorming or explorations ... until I set up my own script in Unity.

Have you considered using one of those online renderfarm services for rendering the videos?

Not really, but we've 7 good custom built desktop computers which we're planning on linking together with a free Blender render farm solution in the future. If you leave them on during a night or two you can basically render 1-2 minutes without problems.

Sure, if you don't need to do a lot, then that approach could work.  I was wondering if you have some cutscene animations in the game.

Thanks Job! I got very inspired and will try to apply some of the shader know-how into the realtime version of Crest, AKA the game. We'll see how that goes.

Looking forward to it!

[Al_B]

This is a very intriguing thought experiment. I'm reminded of Populous (I know I'm dating myself with this reference), except your game has no actual end goal. Kinda reminds me of that quote from Futurama where Bender met God. "When you do it right, it'll look like you haven't done anything at all."

[Greipur]

This is a very intriguing thought experiment. I'm reminded of Populous (I know I'm dating myself with this reference), except your game has no actual end goal. Kinda reminds me of that quote from Futurama where Bender met God. "When you do it right, it'll look like you haven't done anything at all."

Thanks! Populous from 1989 was my first PC game, my father bought a used version, one of the few DOS games I played.  However, the big difference with Populous and "normal" god games and Crest is that in most games of this kind you change the landscape to influence the people. But in Crest you influence the people to change the landscape. The interaction is reversed. I would argue that it's misleading comparing it to Populous, it shares the same genre name but not much else.  

Mechanically it's close to Civilization 5, The Sims, city building games and thematically close to Dwarf Fortress. But associations are spontaneous and subjective, just wanted to clear that out since I think you'd be surprised when trying the game if you thought it played like Populous.

Nice quote from Futurama, we want the player to feel a bit more involved though, hehe.

[PatrickECS]

We've been super busy the past few months and can finally announce our launch date! It only took us about four years of development. ^^

It's exciting but at the same time frightening as hell.