The 2nd Half of the Campaign: Flogging Schrödinger's Horse
In the Darkest Dungeon Kickstarter Post-Mortem Part 1, I talked a bit about the dreaded mid-campaign trough and how we, up to that point, hadn't been able to successfully jolt the daily raise. We had an undeniably awesome start to the campaign (and a decent end, as you'll see). However, this doesn't mean we didn't want to hit a home run in the mid-section, too!
Being an engineering analyst in a former life, I looked at a ton of campaigns and compared to ours. We were getting a pretty steady $3k/day during the middle period. But there were campaigns like the esteemed Hyper Light Drifter, or Banner Saga, that did $10k+/day in the middle. Banner Saga even had a two-day spike of $25k/day in the middle. Why not us? :) At least, that's a good way to think ambitiously.
In the end, we were unable to move the needle much. However, the $3k/day was still respectable and one thing not to be discarded is that perhaps our efforts, while unsuccessful at causing spikes, actually contributed towards a nice, steady baseline.
Aside from fan engagement (social media), press haranguing, and backer updates, we also experimented a bit with advertising.
Specifically, we purchased ads on BoardGameGeek.com (because we felt this was a good interest overlap and we might reach people we otherwise weren't reaching via videogame press sites) and Kicktraq.com (because we figured spreading the word to other project creators might amplify our reach, as they talk about it). It is tough say with absolute clarity, but I think both were positive ROI. The reason it is difficult to say is because we could only get click-thru numbers, but not conversions into pledges.
But if I apply 3-5% conversion rate estimates (SWAG), it showed that our BGG ads probably yielded around $3k on an investment of $500. But there is a big assumption there (conversion rate) and that wasn't enough of a warm-fuzzy at the time for us to want to increase the spend. We might have left a lot of money on the table there by not doing so, who knows. Truth is, we got busy in the middle of the project and didn't really do the analysis until it was a bit late.
I also got 80% though the setup and curation of an AdWords campaign, but never got around to pulling the trigger. Online advertising requires a certain amount of expertise and knowledge to avoid accidentally spending a ton of money ineffectively (like showing RPG banner ads to middle-aged women in Nairobi, for example). We just didn't have the stomach to accidentally flush a few grand down teh toilet due to inexperience, and never got around to fine-tuning things.
This is all to say, I think advertising can be helpful. If we had the bandwidth, we might have been able to do more of it profitably.
It is also hard to confidently spend a bunch of advertising money you don't have while you are running a campaign to raise money to actually be able to survive development. :P It's the old "gotta prime the pump in the desert" scenario.
Press Coverage and "Newsworthiness"
In a recent discussion, another developer hypothesized that you shouldn't expect a a 2nd round of press coverage during a crowdfunding campaign. Meaning, a decent looking promising campaign will probably get an initial round of coverage that amounts to "Check out this game X which has cool hook Y and is currently Kickstarting." Although press gets tired of all the KS submissions, they still love to cover good projects, in all honestly.
However, in absence of a hugely newsworthy event, it is very unlikely to get repeat coverage during the 30-day timeframe. 30-days seems like an eternity in your head when the campaign is running, but the reality is not enough newsworthy stuff generally happens to make the big press outlets cover you again if they covered you early.
We did have a strategy of three news pushes during our campaign (aside from the 15 or so backer updates, livestreams, social media campaign, etc.). But really we only got covered once by most big press. Fortunately, that was quite a good set of coverage and we are not complaining!
The first push was very successful as it was more about releasing the "House of Ruin" trailer than it was about the "we are Kickstarting!" message. (Although the "we are Kickstarting!" message was actually the important message.) House of Ruin was more or less a trojan horse carrying the KS campaign news. We got coverage from pretty much every major site except for the Gamespots and IGNs, which really don't cover Kickstarters much, from what I can tell. In particular, we got some awesome coverage from GameTrailers.com. This came about because they liked our "Terror and Madness" trailer that we released in October 2013--for full details of how important building an audience BEFORE you Kickstart can be, see Part 1 of this article.
Our second big press push consisted of our simultaneous announcement of a 5-minute combat walkthrough video ("Tyler Fights - Episode 1"), DRM free confirmation, Humble Widget/PayPal launch, and a recap of the character classes revealed up to that point. This was done just past the midway point of the campaign. I figured a few sites might post something, depending on which part of that news burst they felt was interesting. Actual significant coverage from this push: ZERO. This was a surprise--I figured, a bit naively, that a few big sites would run at least a piece of the news.
Our last big press push was the last 48-hours. (See End Rush, below). We got effectively zero stories, aside from a Polygon opinion article that I collaborated with Ben Kuchera on.
To be fair, our 2nd and 3rd press pushes had some cool stuff but nothing truly big-time newsworthy. So I think Darkest Dungeon, if nothing else, at least helps validate the general assumption that you can hope for only 1 article per major news site, unless you do or announce something really exciting.
There are things we *could have done* to amplify our newsworthiness. For example, had we confirmed other platforms (e.g. console, handheld, iOS), that might have done it. But we didn't want to do that yet, so I'm glad we didn't fire that bullet yet. It was a strategic choice not to.
Another thing we could have done to increase coverage is to release more gameplay footage or given demo builds to Let's Players...but we also didn't want to do that yet, again for strategic reasons.
The In-Between Pushes
All the millions of things we tried to do during the campaign had almost no discernable effect on the mid-campaign "trough". Our daily haul was incredibly steady (in terms of deviation, not magnitude) compared to nearly every other campaign I looked at. This seems to mean that nothing we did (backer updates, announcements, social media, etc.) really changed it at all day to day. Of course, maybe what we did was enough to create that steady earn and in fact our efforts were remarkably effective. But I think it's the reverse.
The End Rush
Almost all Kickstarter campaigns have an "End Rush", which means that in the last 3-5 days, the daily pledges pick up substantially. There are a number of catalysts, including the 48-hour reminders Kickstarter sends out to anyone who favorited your project. But I think the bigger effect is the "Ending Soon" filter on Kickstarter. People like to jump on things at the last possible minute, so when it's about to end, they finally make their contribution decision.
END RUSH: WHEN OURS STARTED
I was hoping pledges would really pick up as early as 5 days from the end, but they really didn't until day E-3.
Our ending week was not ideal: every game with a "Fall" launched that week, including Towerfall and the lesser-known Titanfall ;) Not to mention Dark Souls 2. It's hard to know for sure, but I always figure that if three promising games launch that week, a fair amount of discretionary money will be sucked out of the ecosystem, and you're bound to lose some purchases from people who would've liked to pledge, but were tapped out in the wallets. $20 is a lot of money for many people, and we always feel incredibly grateful when somebody backs or pre-orders and does their small part to help us make DD what it can be.
Our starting week timing was great and we wanted to end the campaign before GDC, so our finishing week was pretty much set in stone before we launched.
END RUSH: HOW MUCH WE GOT
(Figures Courtesy Kicktraq.com)
We got a total of $50k over the last 3 days. That represents about 16% of the total haul. $50k is a lot of money, but 16% is a relatively small end rush.
For example, games like Shovel Knight brought in 53% of their haul in the last few days. Planetary Annihilation brought in 28%. Rimworld did about 30%. Banner Saga did 32%. Hyper Light did 22%. Delver's Drop did 27%.
This brings up one of the things I am most fascinated with about KS campaigns: the earning curve shape.
END RUSH: EARNING CURVE SHAPE
In studying a lot of KS campaigns, it's interesting to normalize the data (ignore order of magnitude) to look at the shape of the fundraise.
At the extremes, you have projects that are "discovered" very late (e.g. the awesome Shovel Knight) and projects that rocket off to a fast start but then have spent all that momentum and don't have an amazing end (like Darkest Dungeon). See this image:
Neverending Nightmares had a nailbiting but lifesaving end rush:
Night in the Woods (another Vancouver-based dev) and Massive Chalice had shapes similar to ours: they leapt out onto the scene but their end rushes were unexeceptional :
Humble and PayPal
Immediately when the campaign launched, we started getting a steady volume of messages asking about alternative payment methods. Kickstarter processes pledges via Amazon payments, and some people have trouble pledging that way because they don't have a credit card, or are morally opposed to Amazon, and all sorts of other reasons.
The requests were steady enough that we were thinking that there might be a large amount of pledges to be collected if we could provide alternative payment methods via PayPal or the equivalent.
Two of the most popular options are offering PayPal pledging buttons directly or using Humble Bundle Widgets, which in turn facilitate payment through Amazon or Paypal.
PayPal pledging has the advantage in that it can be used for both physical goods or digital goods, and is quite flexible in what you can offer: purchases, donations, recurring donations, etc. However, I read a number of horror stories involving situations where project owners had trouble getting PayPal to release the funds, and things like that. I knew we could get it done, but...
Humble Widgets are not as flexible in that things are not set up to handle physical goods at all. However, they are easy to configure and set up, and the Humble name is well-known within gaming circles. In the end, we decided to set up Humble Widgets for all the digital packages, and we intend to add straight PayPal buttons down the line for physical purchases.
By the time we deployed the Humble Widgets, the campaign was around halfway through and we had collected on the order of $200k via Kickstarter.
We licked our lips a bit and waited for the new pledges to roll in...
...we are still waiting!
By the end of the campaign, we had collected around $5,000 in net pledges from Humble Widgets, whereas the main campaign ended at $313k.
It was very satisfying to be able to offer the option and reach more fans and supporters. But I'm not sure it was worth the distraction and time to try to get it all set up in the middle of the campaign.
We have continued to collect pledges via the widgets, and see nice little bumps whenever a new article runs. We saw our biggest bump after PaxEast, which makes sense. But still, total revenue collected to date via the widgets is an order off mangitude lower than our Kickstarter haul.
This brings up a general piece of advice about all things Kickstarter:
LESSON 156: IT'S VERY DIFFICULT TO TELL WHEN AN ISSUE IS VERY WIDESPREAD OR VERY NARROW
During the campaign, you will hear from many backers, prospective backers, fans, haters, odd ducks, other project creators, potential business partners, press, and more. This is awesome.
When dealing with fans/backers/prospective backers in particular, it can be very difficult to tell just how big of an issue things are. Someitmes a vocal minority can drive an issue way out of proportion, or make requests that you believe might affect hundreds or thousands of people...and end up applying to a few.
If you try to please all of them and satisfy all the requests, you will go insane. But if you don't try to address as many issues as you can, you run the risk of alienating your backers or missing an issue that actually *is* quite large.
In general, though, I think it pays to "measure twice, cut once" during the campaign. The middle of a KS campaign is not a time for really hasty moves, because...
LESSON 157: ONCE YOU PROMISE SOMETHING, IT'S INCREDIBLY HARD TO UNPROMISE IT
Once you commit to something, it really creates havoc to change that commitment. We're fortunate in that we haven't had to do this yet. But part of this is because we have been very careful in how we promise or state things.
Unless we are positive about a course of action, we try to use phrasing with a bit of wiggle room. For example, "we hope to be able to do that", or "currently we are planning on doing that but it depends on how X goes."
I think this straightforward and measured approach has worked out well so far, and it also helps us think carefully about our strategy. It's very easy to get carried away in the enthusiasm and promise the world. Bottom line, our #1 priority is to deliver an awesome game. The details may shift and change, but this goal is constant. It is our "North Star", to use hip business terms.
Logistics: BackerKit, FulFillRite, Etc.
My advice to any aspiring crowdfunded project is to think about logistics before your campaign ever launches.
Logistics entails a few things:
- Handling of your backer database...especially address updates
- Accessing your backers with backer surveys, etc
Kickstarter is amazing for launching and maintaining your campaign. However, there are often things you need to do post-campaign that are made a lot easier by using some third-party services.
Long story short, we elected to use Backerkit to assist with post-campaign fulfillment. Backerkit does a number of things, but most important to us were:
--Good database handling and coordination for delivery to fulfillment house
--Ability for backers to self-update their snail-mail address, email address, and other order details during the long pre-fulfillment period
--Allows backers to purchase more add-ons at checkout!
--Front-line customer support for order/charging problems
--Good data access, for example Item Counts. Yes, you can do this yourself, but it's nice to have things automated.
Fulfillment of physical goods
This could be a giant article in itself and we are far from experts at it, given that we haven't shipped a single art book or poster yet. However, we have been planning for this since before we launched. If your campaign ends up with a small number of orders, you can brute force the mailing of items. However, we always planned optimisitically, hoping that we would get more orders than we can brute force. We did! (Yay / Oh Shit.)
Along with the rise of crowdfunding has come the rise of crowdfunding suppliers. Why hand-package hundreds of art books when you can send your database to a company that specializes in packing and shipping (fulfillment). Well, the only reason I can think of is cost. But if you value your time at any significant hourly rate, this cost pays for itself and more. Aside from $ cost, though, endless hours spent folding, labelling, and shipping boxes means we are not actually working on the game. This is really, really bad for that North Star goal we mentioned earlier.
Hence, we are planning on using a fulfillment house. We settled on FulfillRite, after doing some online comparisons, getting some price quotes, and also talking with our other service provider, BackerKit. So far I have been happy with the communciation with FulfillRite, but we haven't actually shipped anything yet so I can't give a postmortem review...ask again in a year!
I sleep a bit better at night knowing they are out there, though. We will probably hand-ship our Canadian packages because we offered free shipping to Canadian purchasers (yay Canada!) and it will cost us an arm and a leg to fulfill from Fulfillrite's NJ location given we have to eat the foreign shipping cost. The number of total Canadian packages to ship is a number that can be brute forced.
Fulfillment of digital goods
Fulfillment of digital goods will be handled mostly through Humble--distributing Steam Keys and DRM free download links.
Refunds, Order Updates, Address Changes
This stuff makes me want to tear my hair out. It's a good problem to have, but with close to 10,000 backers, there is a constant trickle of people wanting to change their orders, up their pledges, reduce their pledges, change their address, change their email addresses, and so on.
This brings up several issues: what are your policies towards these things? And how logistically will you do it all?
We've tried to have very flexible policies. In one case, a backer had overextended themselves and had unexpected family expenses pop up right after the Kickstarter charge had cleared. We worked with them to refund them via PayPal. It's nice to be a small enough company where we can remember the human element. Not trying to pat ourselves on the back for this so much as to say that I think it's helpful to remember that each person has contributed to make this dream game possible. Our goal at Red Hook is to acquire fans not just for Darkest Dungeon, but for life. If we can help someone out with a little extra effort, we do try. Unfortunately, the scale means that it's not always possible. But we do try.
As for logistics, Backerkit helps here. The nice thing about Backerkit is that backers can update their own information (snail mail address) by editing their own order at any time. This is one of the reasons we decided to use it. (I think that's a bit harder to do on Kickstarter directly, at least as of when we chose to go Backerkit.) So when someone emails with an address update, I direct them to their Backerkit checkout page.
Still, having Kickstarter issues, PayPal/Humble issues, and Backerkit issues can be a bit much sometimes. My biggest fear is having people slip through the cracks, and I fully anticipate that we will face some of the issues that every KS project faces down the line: a few mixed up orders, people not receiving their emails, outdated snail mail addresses, etc. I think it's unavoidable to have a few of those things happen.
These are times, though, when I wish we could afford to pay a person to deal with all of this stuff! However, our entire budget is reserved for development, and honestly the volume of administrative stuff isn't high enough to support a person yet. But when you are 5-person company, any time dealing with customer service means a monster isn't getting designed or balanced! One of the joys of independent game development. :)
The Joys of Taxes
Our tax situation was not simple and I won't go into it all here. However, one helpful thing is to be able to determine backer residency (country, state/province, etc.). You actually can't do this easily via Kickstarter itself because Amazon is who processes the payments and stores the customers' locations. We ended up using Backerkit survey questions to determine the pieces of information we needed, along with Backerkit's database segmenting. In the end, it was quite easy for us to figure out who resides in WA state via this survey functionality.
Other Kickstarter Projects - Paying it Forward, Curation
It's exciting and exhilirating to have a successful Kickstarter project. If you do, you can expect to become inundated with requests to do a shout-out for other campaigns.
This is a fun way to meet other project creators, mutually pat each other on the back, and pay your success forward. We take this very seriously, as we benefited from campaigns that went before us, and we would love to see tons of other successful campaigns.
However, being suddently "popular" can be hard at times because we just can't shout out everyone that asks us to. Well, I guess we technically could, but two things interfere: time and curation standards. Time, because we are trying very hard to focus on development, and it just doesn't always work out to shout out a project if we don't have a Backer Update planned or aren't heavy on social media at that moment. The second issue, curation, means that sometimes there are projects that we don't really want to shout out, for any number of reasons ranging from poor subject overlap (see below) to not personally being excited about the project. We don't want to become a megaphone of daily shout outs, so we'd rather shout out the projects we are genuinely excited about, as opposed to every project that asks us to.
Potential backer interest is important to us, most of the time. We don't really want to promote things to our backer base that we think might be totally outside their tastes, or our focus. It doesn't make much sense for Red Hook to promote an interpretive dance KS, even if the project is awesome...or unless it is interpreting the Necronomicon, of course.
That being said, there are at least a half-dozen projects I can think of that definitely deserved to be shouted out, but timing just didn't line up or we forgot and dropped the ball. I feel bad that we couldn't meet some of those requests from really solid teams and projects. So if you asked us, and we didn't do anything...sorry!
But in general, I think we've done a good job of paying it forward and working with other creators, and we look forward to doing more of it! It's an awesome part of the crowdsourcing and indiedev process, and we've met some really cool people this way.
What's Ahead: Harnessing the Power of Backers!
This article is already long enough, but there's so much more we could touch on. In the eventual Part 3, I hope to recap how our selective crowd-sourced design efforts went (some content creation tied into KS reward tiers), fulfillment process, and more. But that will be some months away at the earliest.