CoG/HG Gifting

I would be happy to help!

This is exactly the type of thing I like to do for fun. lol

5 Likes

Welp, that’s done. More or less. While we reached our ten-book goal (and then some), execution (my fault entirely) was sloppy (understatement). Results still trailing.

If, in the future, anyone wants to try something like this again, you might PM me first. I may be able to save you some heartache with “lessons learned” and all that. In short, there is recipient demand and giver willingness, but quite a lot else needs to be sussed out.

Thanks to the staff and community for letting me try (the project and your patience) and for trying with me. Cheers.

9 Likes

I think lots of folks would be interested to read your thoughts on lessons learned; posting those thoughts here would make a lot of sense.

6 Likes

Hmm, okay. Sorry, I was trying to spare everyone the gory details. Keep in mind that these oversights are complicated by the time between messages (hours, days, time zones) at each phase.
Here’s some stuff off the top of my head. I will edit it endlessly and add more when I look at the data.

Some technical oversights on my part.
  1. I feel pretty ashamed that I didn’t get a clear go-ahead on all aspects of the execution before starting. So, some people had agreed before the conditions got changed. This is my fault; I got excited. Originally, I’d offered to be a purchase intermediary to cut down on comms, Steam literacy, and preserve anonymity. This led to belabored process, backpedaling, and some strained communications/relationships.
  2. As far as I could find, you had to become friends with someone on Steam first, in order to send them a gift. This is more intimate, I think, than some people (on both ends) intended to be. Even given the option to un-friend someone after the transaction. Friends have access to lots of information about you.
  3. A Steam username is different than a Steam profile. The latter is more useful for pairing people. I should have asked recipients to provide a direct link with their game request to cut out steps. Especially given time between posts. And I should have known to ask for the latter.
  4. Someone cannot send a friend request on Steam without having first spent $5 with them. I’m taking this on hearsay, having heard it from a few would-be recipients.
  5. People coming in from Reddit didn’t have the Trust levels to PM me directly. Neither did people who came to the site from other sources in order to participate. Lots of solves for this, but nothing I had in place.
  6. One giver was able to send me purchase keys which I then passed on for redemption to their recipient. This was a smoother transaction. AND having either a stock or pinned thread of these codes might enable ongoing giving. E.g. on VtM:OfB release, posting 10 codes for the public to redeem and striking through used codes. There would be exploiters, but this can be managed.
Some human oversights on my part.
  1. Because I waited two weeks to begin giving, 2-3 users got tired of waiting and acquired the games they’d wanted otherwise in the interim (parent, girlfriend/boyfriend purchase). So, a win for the company in terms of sales. Not for the person who’d been ready to give.
  2. As some users may have created CoG and/or Steam accounts in order to participate, I’m not sure they were used to checking them. So, process may have gotten unintentionally stalled.
  3. Because of recipient bail out, I ended up hassling and then disappointing people who’d wanted to give. Two in particular. One giver had been lined up to gift out the Sordwin Saga. However, their recipient got someone else to buy it for them in the interim without telling me. I had a backup lined up, who also bought all three books in the interim. So, six sales spurred for CoG, but 3 unfulfilled gift requests. Another giver had been lined up to give three distinct titles and, due to a combination of #5 and process fatigue, ultimately opted out.
  4. More givers were interested in supporting specific titles/authors. Originally, I thought givers would give what recipients chose. In addition, I had many messages from recipients wanting recommendations from the list of available titles. So, BOTH parties seemed to want to be specific gifts. Because I had both specific and unspecific gifts, my back-end organization kept changing with incoming requests, making more work for me.
More on process fatigue.

Given the above factors, some gifters had to PM me >10-20x only to end up gifting nothing. Others had to do a lot of leg work in order to do something nice. I could sense both gifters and giftees growing frustrated with my inefficacy. At one point, I had to tell a dozen or so people to check their Steam friend requests, but had to wait–the forum thought I was spamming, and I couldn’t think of enough different ways to say it. :upside_down_face: I don’t mind at all if anyone wants to share their own bad experience. For many people, too, it went seamlessly.

Here's what I would do differently, barring a dedicated pinned thread or forum gifting process.
  1. Begin with fixed demand. Say, if a user had a high school class of 30. And we wanted to gift each kid a book. That goes better. Methodical. There are lots of ways to do this. Happy to elaborate or not to. Originally, I’d only planned to match ten people as a proof of concept. My fourth giver wanted to give ten books all alone. So, I decided (as much for data as effect) to let it run its course. We ended with total pledges north of 40. All reipients will write reviews and buy a book for someone else in the future. And so forth in perpetuity. Promotional activity would be another way to generate fixed demand. Have givers sign up in advance; authors promote on their own sites pre-release or pre-whatever (e.g. anniversary, pride month, . . .)

  2. Fixed demand helps w disproportion. More givers than recipients. But I know from reading reviews there are people who want games but can’t afford them who I didn’t reach on the forum, Reddit, Discourse, or Steam. There are more givers within the area where I cast the net. But, if I asked a freshman coding class at community college if they all wanted one . . . There are also book reading clubs online–it might be interesting to sponsor a group of people to read a title together. And discuss. Etcetera.

  3. Organization. I could have saved a lot of steps by having a Google sheet which a couple of mods could access. This would also have spread the load. Ideally, it would have columns for each phase of fulfillment (including giftee app review and eventual regift). I would make only some of those sections viewable to the general public as read-only. Helps w accountability, enthusiasm, momentum–an actual chain.

  4. Labor. I probably should have asked for help, but didn’t want to subject anyone to my learning curve. In some ways, that was self-interested too. Takes more time to explain necessary pivots. Knowing what I do now, I would be able to anticipate a clear path forward and delegate sub-tasks. Like follow-up. Info sharing would be helped by the shared document.

  5. I would (dreadfully) add at least one step. Giftee confirmation. I should have asked at the end if giftees still wanted to receive their pledge. I didn’t to spare myself 50 or so messages. Plus, some had only been paired the day before, where others were paired weeks before. I think the shared document could help with this too. I could mass-message all participants. I know the source thread functions that way, but not everybody read every message. Nor should they of had to in order to give or receive.

There is something here that, if it got ironed out, could be huge for the company and community.

Company

A customer usually only buys a title once. Say a $10 cap. But they may support that author, people like themselves, or the readership enough to pay $40. But they can’t. They might pay $3 for a DLC in 6mos. But the CoK uncaps their spend. They can pay $40 by paying $10 4x. Plus, it generates reviews and further purchases. Each giftee gifts a giftee who gifts a giftee . . . I would like to run a controlled experiment on the VtM fandom, as it already exists in a large way and on established platforms. Plus, they’re content-hungry progressive humanists and very supportive of the LGBTQ community. E.g. LA by Night. I’m betting a controlled chain could go thousands long.

Community: same thing re: engagement. On forum, Reddit, and Discord . . .

The truth is CoG is on the forefront of the future of reading entertainment. We’re all early adapters. I think it’s about digging the right channels. You’re also onto some things with licensed IP. I’m betting you could raise $ for licenses on Kickstarter and still sell books to your funders. But, one thing at a time. Thanks for reading, everyone. Sorry for how strained things got with many of you. I ended up fielding so many messages that, at times, I may have seemed short or unspecific. Nothing but love from me.

15 Likes