This week, we're kicking off the "Simplifying Slokie Initiative " with the goal of allowing our customers to order books using a single picture with minimal requirements around it using a tool that lets them immediately know whether the image will work with one of our books or not.
The first stage of this initiative is now live. We're pleased to announce we now only require two images to generate a copy of Nightlight (down from five), and if all goes according to plan, our second book will only require one.
We hope this is welcome news for grandparents, aunts, uncles, and anyone else who's been wanting to order a copy of Nightlight, but hasn't had the right pictures readily available, or those who just don't have the time to do a quick photoshoot.
Overview of the Initiative
We launched Slokie with a system that was a bit more complicated than we would have liked, requiring at least five uploaded images that met a very specific list of requirements. From doing some beta testing, we knew it at least worked, and that the end product was generally well received by those who went through the ordering process. So, we considered it to be at the 'minimal viable product' stage and launched. But we recognized that there was certainly room for improvement, especially with respect to our ordering process. And so we've been quietly working on this issue in the background since launch, and now finally have the first major improvement live on our servers.
And we're not done yet. Over the coming months, we plan to test and deploy even more changes to simplify and generally improve our ordering process, from providing a quick preview of the book, to giving our customers the ability to quickly and easily self-edit that book and send it immediately to the print queue.
Roadmap/Timeline
- Reduced number of required images (now live!): cut the number of required images needed to generate a copy of Nightlight down from five (5) to two (2). Note: we still recommend at least 3-4 for best results, but results are generally pretty good with just two.
- Auto-Generated Downloadable PDFs (tentatively coming in September): customers will have the option to buy a PDF version of the book at a much lower price that will let them read the books to their kids on their tablets. Illustrated books don't lend themselves to the dynamic e-book formats, so this is best solution we have found. Customers who purchase these PDF versions will receive a code that will offset the cost of the printed versions by the cost of the PDF version, so they can more easily see the results. Also, we will offer customers a free revision of these PDFs if anything did not turn out as intended.
- Instant Feedback of Uploaded Images (September/October): initially this will take the form of a photo validation tool that will simply reject images that will result in subpar books,
- Long-term/Ongoing: (2023 and beyond)
- A live preview tool accessible during the ordering process, with image edit features to give customers the ability to create books without any reliance on our staff for manual reviews or approvals.
- Customer accounts with galleries for easy ordering of new book titles using previously uploaded images. We currently delete images after a while both to protect our customer's privacy and also because we can't currently link them to future orders, so they really just waste online storage space.
- Alternative methods for illustrating heroes. We currently use a background removal process to extract our heroes from uploaded images. Instead, we're working on an alternative set of solutions that would allow a single image that meets no specific list of requirements to work. It could be a portrait or a full-body shot. There are several machine learning-based approaches we're experimenting with here, so we're not ready to make any announcements on the timing or other specifics of this feature, but it is in the works, and should be done either Q1 or Q2 of 2023.
- Add additional customization options for books, such as our hero's favorite animal in their favorite color, using DALL-E. Timing is completely dependent on the release date of DALL-E's API, which is not yet known.
None of this mentions our upcoming book titles, but some of these features will coincide with new releases. For instance, we plan to further reduce the image upload requirements from 2 to 1 for our second book, but we may need to keep the two image requirement for Nightlight due to decisions that were made in the plot and artwork for that book when it was created. We may update Nightlight accordingly down the road to account for that.
As always, please subscribe to our newsletter at the bottom of the page to remain informed of any future updates.