Hello kind Night Readers,

It’s January 11th, and as promised, here’s a brief update on Saji, which has (despite appearances) not yet fallen into the void of forgotten projects:

I had originally intended to sell the notebooks direct via a Shopify. This plan largely failed because I had no time to devote to care and feeding (marketing) of the project in 2022.

I did sell some books to family and friends, and while the quality was very well received, the main piece of feedback I got was that they were too expensive. Indeed, $63.00 is a lot to pay for a notebook, however, after surveilling the marketplace for similar hand-made and editioned “objet d’art” products from independent designers, I believe the price to be in the right range… for the right customer. My interpretation of the feedback from friends and family who did purchase notebooks was that most were not the type of customer who would buy this sort of thing, were it not sold by yours truly. This data point proves both the enormous generosity of community and the dangers of relying on said community for business intelligence.

This begs a question: how do I find and connect with customers who WOULD buy and love such a product? I have been adjacent to brand builders, content marketers, and advertisers for most of my professional career as a designer, and yet I have learned that such things do not come easily to me. Building a demand chain on the internet is extraordinarily time-consuming, and I believe this to be especially true in the “designer goods” category, which relies heavily on brand, presentation, and narrative. To summarize neatly: The pile of work is large, and I have very little interest in tackling it right now at the expense of other, more important projects.

Thus, I am beginning to lean towards using indirect sales channels (i.e. brick and mortar shops), which have the benefit of already having built-in customer bases, allowing me to begin selling and gaining exposure faster. Because I intend Saji (for now) to be a small side-business and not a primary source of income, anything that reduces my initial time expense is a gift.

There is another approach to explore, which is lowering the cost to appeal to a larger base of consumers. This might involve digitally printed covers instead of screen-printed art, less expensive binding methods, and possibly different paper. While the resultant product would certainly still excel at being beat up and written on/drawn in, I envision these being companions to the handmade-editions for more price-conscious customers — sort of like making prints of original artwork. This, of course, comes at the cost of doubling down on my commitment and ordering more inventory before I have found a market for my existing inventory.

No matter which path (or paths) I pursue, there is likely some work ahead of me. I am doubtful, even in a retail setting, that customers are likely to buy an “objet d’art” product without sufficient narrative. I would likely need to come up with some sort of branded collateral, better imagery and retail displays… perhaps even a dedicated, higher-touch web presence than what I have cobbled together thus far. It may be prudent once I am underway with these efforts to hire help in areas where my own skills fall a little short — product photography comes quickly to mind.

All of this adds up to a hefty chunk of effort that will likely need to be spread across a number of months as I bumble my way towards traction.

There is, as ever, a line to be walked, carefully and thoughtfully, between not enough effort, energy, resources, time, and so much of these that I have none to devote to anything else. This is Goldilocks zone, and the holy grail of any successfully lean enterprise (aside: in the first recorded version of the Goldilocks story, the three bears throw Goldilocks into a fire then impale her. How’s that for a grizzly ending?).

Ghost 5 and other interesting technical gobbledygook

You may notice that this edition of the letter looks a little different from prior editions. If you’ve never thought much about where newsletters come from, it may come as a surprise that they require dedicated software (as well as services and servers) to create and deliver. Newsletter delivery is an industry unto itself, and many millions of dollars have been made facilitating the delivery of emails to willing (and unwilling) recipients.

Since the beginning of this enterprise, I’ve used a service called MailerLite to perform the magic of mass message delivery, which — combined with some custom code — has allowed me to create and send these emails to you while keeping my own costs manageable.

This comes at a expense of a lot time and energy, however, as the workflow for creating and publishing is extremely onerous; it involves not one, but three separate platforms (I publish each letter to The Desk and your inbox at the same time), lots of one-off edits and code adjustments, and is highly susceptible to human error… all the more so because this particular human is prone to rewriting prose long after he has finished writing and letters have been committed to their respective publishing platforms.

So when Ghost (the wonderful open-source CMS behind The Desk) announced a native newsletter integration that would allow me to publish letters to both the web AND to your inboxes in one streamlined step, I jumped at the chance. I am already hosting and maintaining my own copy of Ghost in order to run the desk, and the added newsletter functionality only costs as much as its underlying service layer, Maligun, which is both affordable and scalable. In theory, it is the perfect solution. The only downside is that since this functionality is new, it lacks robustness and features that I have come to appreciate while using a dedicated email service. Things like welcome emails, list segmentation, and theme customization/CSS injection are simply not possible here (the latter is particularly painful to give up, as I am quite fond of some of the little design details in the old newsletter template), however the gains in efficiency and publishing workflow more than make up for these losses.

In order to make all of this possible, I had to perform what I thought would be a simple and quick Ghost version update... not so; before I knew it I was in the midst of a multi-day linux-updating marathon replete with broken binaries, misconfigured mail servers, and missing dependancies. Ahh, the joys of server administration — they make a body want to shave his head and become a monk.

In any case, the work continues, albeit slowly. Eventually, we’ll get where we’re going.

Until next time,


P.S. Would you like more “under the hood” content like this? Did it make your eyes glaze over? I’d love to know what you think. Just hit “reply” and send me a note!