Progress Report 2025 Week 25

My plans for Artificial Intelligence (AI)-based audiobook narration took an interesting turn this week. One possible source of AI narration hit a dead end, and I suspect I should have known earlier. Another continues to stumble forward. That aside, my hopes to publish Dreams of Deucalion in June are fading fast. One might say they’re faded completely. Also, I am faced, once again, with the implications of not having a better grasp of marketing. Despite a year to trying to think positive, I’ve come to Yet Another Realization (YAR). In addition to that non-writing stuff, I actually got some writing done. I’ll talk about how much after we take a look at the key performance indicators (KPIs).

Last Week’s Progress By the Numbers

I certainly can’t complain about the number of words that I pushed through the keyboard this week!

Purely in terms of writing, this was an outstanding week. I didn’t have to handle any Real Life Family Events (RLFEs), which is a welcome change of pace! Not sure how long it’ll last, but I was ready for a week of relative peace.

The characters continue to freely speak to me. Gerhard Wimmer in particular really stepped it up this week. He took off in his chapters, and I honestly had no conscious foreknowledge of where he was heading. The content worked, too!

Owen Payne and Moritz Lehner didn’t want Wimmer to get all the credit. Wayland’s Hammer Book 2: Resistance Movement has less action than Wayland’s Hammer Book 1: Conventional Forces, but the characters are, I think, making up for it with suspense and drama. That should set the stage for a cataclysmic confrontation in Wayland’s Hammer Book 3: Lines of Operation. I hope. Fingers cross, if the creek don’t rise, and all those cliches.

State of Flux as an Understatement

I Can’t Use What With ACX?

I’ve mentioned that I’m working with Amazon’s Virtual Voice. I plan to use it to make Evolution’s Hand Book 1: Executive Action into an audiobook. When Virtual Voice leaves beta, I may have some insights to share about it, but until then, I’ll just say this: its result is better than I could produce on my own. Plus, I can’t afford any better.

The latter hit me in the face this week.

My progress post from 2025 week 14 talked about my experiments with ElevenLabs. Its output sounds the most natural to me of all the platforms I’ve tested. It’s a paid product, but given how affordable it was compared to a human narrator, I pulled the trigger for a year’s subscription.

It occurred to me that I wasn’t sure of the procedure to take the output from ElevenLabs to Amazon. Would I need to go through ACX? Another distributor? Was that even possible? I had researched recording an audiobook myself, so I had a vague idea of the technical requirements, and I wondered how ACX (or whatever) would check the output from an external source like ElevenLabs.

It was surprisingly hard to find a clear answer. I found some sources saying you can upload directly to ACX, but more sources disputing that. I found one saying ElevenLabs had a deal with Spotify (who purchased FindAway Voices), who could then distribute to Amazon. Then I found sources that said no, that was wrong.

Telling Generative AI to Show Its Work

I even asked ChatGPT to use only recent sources, link to primary sources, and give me an answer. Its answer self-contradicted. Not surprising — it can only operate on what it can find. Then I came across this video called Where can I self-publish my A.I. narrated audiobook? UPDATE – ElevenLabs and Findaway Voices by Mary Kate “M.K.” Williams. I took away two important points:

1. The AI narration industry is in a state of massive flux; the main players, like Amazon, Apple, and Google, are fighting hard to establish themselves as the leader, which means there’s not a lot of collaboration
2. There is no direct path — or any path at all — from an ElevenLabs-recorded audiobook to Amazon

That second point hit hard. I thought I had read there was a path, back when I decided to buy a year’s subscription. But looking back, I think I read what I wanted into inconsistent information. The bottom line is that if I want to sell audiobooks through Amazon, and I do, then I either have to use Amazon’s Virtual Voice or find a human narrator. Yeah, I could go through ACX, but I can’t afford to pay up front, and the other options don’t appeal to me (rights wise).

I’m not sure what letter I’m down to now. For audiobooks, maybe E?

Some writers use ElevenLabs to publish on Spotify, but I have a sense that market isn’t huge. Other writers use ElevenLabs to publish their works on YouTube, and there are possibilities there. But realizing those requires time, and I have even less time than cash. And I don’t have a ton of cash. So that’s out.

I learned something while using Virtual Voice. I catch more typos when I’m listening to the draft of the audiobook. So, I’m using ElevenLabs to read the first chapters of Dreams of Deucalion Book 1: Special Recon back to me. I’ve already made a few minor changes based on how it sounded. But if that’s the only use I can get out of it, I won’t be able to justify renewing it.

Publishing Dreams of Deucalion

With The Sword of Sirius, I followed this path:

1. Finish the first two drafts (initial writing and read-aloud)
2. Let it sit until the next manuscript is done
3. Re-read it and push it through ProWritingAid
4. Order the cover
5. Send it out for an alpha read(s)
6. Incorporate the results
7. Send it out for a beta read
8. Incorporate results
9. Send it to my proofreader
10. Incorporate results

The key idea is that I’d finish each manuscript, write the next, then put the first through the editorial process. That way, I could let it sit long enough to read it more objectively.

I tried something different with Dreams of Deucalion. I held off on everything after step four until all three books were done. My reasoning was that hey, if I screw up and need to do major revisions to support foreshadowing for a later book, I’d have less editorial work.

What actually happened was that all three books hit a choke point. My alpha, beta, and proofreaders had to work through all three manuscripts one after the other. The alpha/beta readers have finished their work. My proofreader just finished Dreams of Deucalion Book 1: Special Recon. It’ll be at least early July before Flanking Maneuver and Bait and Bleed are done. I’ll be lucky to finish by the end of July.

For Wayland’s Hammer, I’m going to go back to the same process as The Sword of Sirius. Even if I have to make revisions, I can proofread a paragraph here and there myself. The end result will be publishing about a month or six weeks after the final manuscript is done — instead of two months to ten weeks.

Marketing Eludes Me

Here’s what I think I understand about marketing right now:

Potential customers might bail out at different spots, depending on whether they start with Facebook or Amazon Ads.

Last week, I talked about how well the BookBub Featured Deal performed for me. My sales and page-reads are still elevated over their usual levels despite that deal having finished a week ago. I’ll just be blunt here: I want those numbers all the time. I want better than those numbers. I want to fund my son and daughter’s retirement with my sales. And it’s not happening so far.

I’ve been getting more ratings for Red Flag Warning, and they’re fours and fives. Even my ratings on Executive Action are high. I’m confident in the stories. I just can’t figure to get them in front of enough people! I’m missing something very, very basic. I have a tendency to do that, too, and it’s frustrating.

The diagram above represents my understanding of the factors I can control; factors that might cause a potential reader to bail and not buy. I think my blurbs are okay. I think the first chapter’s good. I’m blaming marketing, but I’m going to be honest here. I’m afraid I’m paying for not writing to market. That’s particularly true for Evolution’s Hand. It’s hard to market a series that tries to be its own thing and in doing so, straddles genres.

Is writing that way selfish? Am I leaving money on the table? Am I putting myself at a disadvantage? Or do I need to work harder to find readers who like my stuff? Something like The Expanse (affiliate link) is hard to market, but once it found its people, it exploded. Or maybe its marketers figured it out before they published. I don’t know.

Hell, maybe I need to re-read Chris Fox’s Write to Market (affiliate link) and pursue that course for a while. I mean, I want to write what I love — but it doesn’t mean I won’t love it if I figure out a niche that’ll sell. Or maybe I need to keep cranking out novels and see if I can build a readership. The Facebook group is called 20 Books to 50K, after all. I have one esoteric (hard to market) series of six books plus one trilogy that’s selling okay. Maybe a second and third trilogy will help.

My Yet Another Realization? I know more about marketing than I used. But I still don’t know marketing.

Progress against Last Week’s Goals

Here’s how I did against last week’s goals:

1. Finish Gerhard Wimmer’s plot turn 1: Done!
2. Finish part III of Owen Payne’s plot turn 1: Done!
3. Finish Wimmer’s pinch 1: Done!
4. Finish Moritz Lehner’s pinch 1: Done!
5. Continue working on the AI-narrated draft of Evolution’s Hand Book 1: Executive Action: Continuing! I managed to proof another two chapters.

Goals for the Week in Progress Report 2025 Week 25

Here’s what I hope to achieve this week:

1. Write Owen Payne’s pinch 1
2. Write Gerhard Wimmer’s Midpoint
3. Write Moritz Lehner’s Midpoint
4. Write Owen Payne’s Midpoint
5. Continue working on the AI-narrated draft of Evolution’s Hand Book 1: Executive Action
6. Prepare Dreams of Deucalion Book 1: Special Recon for publication

What Do You Think?

Have you worked with any of the major vendor’s AI narration? Or are you opposed to it? I’d love to hear your experience in the comments!