Harlequin’s Guide to Cities and Poison Post-Mortem

Harlequin’s Guide to Cities and Poison got 470 backers and raised $6,659. That’s about 80% of what Celawyn’s Guide to Wilderness and Fey got by backer count, but I always knew I wouldn’t necessarily be able to consolidate the spike in interest from that book coinciding with the upcoming release of The Wild Beyond The Witchlight. If we assume the Witchlight spike concealed some amount of reliable growth, that indicates that both Celawyn’s and Harlequin’s got about 15% growth. Some people have suggested that cities and poison is a bit of a dud topic just like piracy was, so maybe it was 20% from Celawyn’s and 10% from Harlequin’s, but I don’t want to bake that kind of speculation into my stats (especially not when I’m already speculating about the effects Witchlight spike as it is), so I’m writing down 15% for each.

It’s not as straightforward as all that, though. Harlequin’s Guide started even stronger than Celawyn’s. For the first two days, it looked like the Witchlight spike had been consolidated and we were going to grow even further from there. Things always fall off after the first two days or so, but Harlequin’s fell off harder than normal, the gap between it and Celawyn’s closing pretty quickly. I did some math to try and figure out why. The reason for comparing the first 28 hours versus the last 68 hours is because I start my campaigns four hours before midnight according to Kickstarter servers, which means my first “day” is only four hours long, my last “day” is only 20 hours long, and the final 48 hours (exactly) are spread across the last three “days.” This means, in order to capture the final 48 hours, I also have to include another 20 hours extra. Since this is true of every campaign I’ve got data on, however, it shouldn’t impact our results much.


Natalia’s: 36.7%
Irena’s: 33.8%
Bianca’s: 38.1%
Brac’s: 33.2%
Thaemin’s: 39.5%
Celawyn’s: 26.3%
Harlequin’s: 40.5%


Natalia’s: 5.4 (3.4%)
Irena’s: 6.9 (3.2%)
Bianca’s: 8.6 (3.0%)
Brac’s: 10.4 (3.4%)
Thaemin’s: 11.7 (3.2%)
Celawyn’s: 22.8 (3.9%)
Harlequin’s: 13.6 (2.9%)


Natalia’s: 22.1%
Irena’s: 30.0%
Bianca’s: 27.6%
Brac’s: 28.5%
Thaemin’s: 24.8%
Celawyn’s: 29.9%
Harlequin’s: 27.5%

From this, we can see that Harlequin’s had a stronger than average start (40.5% of backers in the first 28 hours), but not much stronger than Thaemin’s (39.5%). In fact, Celawyn’s was an unusually weak start (26.3%, far behind the second lowest, Brac’s at 33.2%). The final 68 hours were extremely typical (nearly median, in fact – the median is Bianca’s with 27.6%, and Harlequin’s is only one tenth of a percentage point lower). The relatively much stronger performance of not just Harlequin’s but also Thaemin’s (the latest campaign that did not receive an obvious spike from uncontrollable, external events) could be indicative that I’m approaching a plateau, with more and more of my backers coming from people who know instantly that they’re backing the campaign because they already know my work.

But I don’t have to speculate about that. I keep track of which backers are new versus returning for every one of my books, as part of extending special thanks to those who back me consistently. There’s a minor flaw with this, which is that people who never fill out the survey or who wish to remain anonymous are not counted, but these are both pretty tiny fractions of the total population for any given Kickstarter, so it shouldn’t badly impact the data one way or another. There’s also a major flaw, which is that it relies on survey responses, which usually take a week or so to come in. This means I don’t actually have data on Harlequin’s yet, and I’m too impatient to put this post off until I’ve got it. The data on what part of the campaign backers come from (i.e. first 28 hours vs. final 68 hours vs. the middle) was already showing warnings signs with Thaemin’s, though, so let’s look at the percentage of new backers for each book up to Celawyn’s and see if Thaemin’s is a noticeable aberration:

Natalia’s: 100%
Irena’s: 67.6%
Bianca’s: 58.9%
Brac’s: 55.8%
Thaemin’s: 50.2%
Celawyn’s: 58.5%

Looks like no, Thaemin’s doesn’t stand out. The total number of people who’ve never backed before is going steadily downwards, but that’s to be expected. My audience is drawn almost exclusively from people who back things on Kickstarter a lot, and people who back D&D-related things on Kickstarter a lot are ever-more-likely to have backed at least one of my projects as the total number of my projects increases. Some amount of my growth is coming from converting people who catch my books whenever one happens to catch their eye into people who back all of my books because it is me writing them, and that was always going to be the case.

So that leaves the question of the weak middle. There are many potential explanations why the middle may have been so weak:

-Poor stretch goal structuring. I always mark out my stretch goals in advance, rather than doing the standard slow unveiling thing. The slow unveiling thing is definitely effective (no one is ever motivated to back for a higher amount or convince others to back my project because the stretch goal after the current one might be something cool), but it interferes with my “fire and forget” approach to Kickstarter projects. Bad enough that there’s no way to schedule updates so I have to remember to post them. In any case, I usually structure the stretch goals around encouraging people to try and make the current project more successful than the last. This time, I actually structured them around just matching the last project, since I knew that consolidating the Witchlight spike could potentially be challenging or even impossible. It turned out to be the second one, which means only the $5,000 new art goal was hit. The lack of steadily hitting stretch goals may have harmed momentum in the middle. I should definitely switch to $2,500/$5,000/$7,500 for the next project.

-Running out of add-ons. Related to the above, there are no longer any signed copies of Natalia’s Guide to Necromancy being offered as add-ons, as all 100 are now spoken for. This led to fewer people buying signed copies (it is now impossible to get a complete set from scratch), which meant a lower amount of money-per-backer. This means that me and my freelancers are getting paid less for our efforts instead of more for the first time, but we all knew that the Witchlight spike might be impossible to consolidate. More relevantly, it means the number that most people pay attention to, total money raised, looked pretty weak compared to previous campaigns, especially compared to the higher stretch goal amounts. I should probably give add-ons for unsigned copies, including digital copies, going forward. This might end up just shuffling money around, since I already have a link to a page showing all of my work so far in the FAQ, so maybe I end up getting less money from DTRPG and more from Kickstarter. Also, it’s kind of weird to ask people to wait until the end of the Kickstarter to buy things that are available immediately from DTRPG. Not everyone sees the FAQ the way they do add-ons, though, and adding a few extra clicks (click through to the list of my complete work, click through to individual books’ DTRPG links, click through to buy from DTRPG) might be driving down sales.

-Being a “project we love” is actually harmful. My project was marked a “project we love” by Kickstarter staff, which was a neat feather to have in my cap, but it doesn’t seem to have done anything (except maybe driven a higher early surge in backers?). In fact, there’s weak, correlative evidence that it may have harmed the project, but this is so counterintuitive that my current assumption is that it’s coincidence.

One final confounding factor is that a cross-promotion with another Kickstarter yielded at least 10 backers. The boost from that cross-promotion bled into the final 48 hours in a way that makes it hard to say exactly how many backers they brought in, but it could plausibly be as many as 50. It’s possible that actually steady growth is winding down, with projects getting only 10%, not 15%, and it only looks like 15% because this project had a cross-promotion spike that covered up data on steady growth just like the Witchlight spike. Unlike the Witchlight spike, this suggests an obvious course of action: Do more cross-promotions.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s