I’m Giving YouTube Another Shot. This Time, AI Is Helping Me Edit

When I was 12, I sang a solo in front of 8,000 kids (it was a kids choral event, and I played Thomas Edison, haha). I also gave a speech at my high school graduation, so it's probably not a surprise that being in front of people has never scared me, and I've been told my whole life that my personality is suited for it (for lack of a better word).

So it's a little funny that it's taken me this long to give YouTube a real shot. Life got in the way, the way it does, and then I needed time to figure out where I actually wanted to go with my business.

I'm in a different place in my life now. AI lit something up in me, and it feels like my entire business career has been preparing me for this moment.

And if you've spent years building a business too, I think the same might be true for you. Everything you know about your customers, your work, and what actually sells is exactly what AI needs from you. It's the part no tool can do for you.

But the personal reasons are only half of it. I chose YouTube very intentionally, and this time there's a real YouTube strategy behind it.

I can't rely on one platform, and neither can you.

If I could get one idea across in this whole post, it's this one. Not SEO, not any single social platform, and that includes Substack. Algorithms change all the time, usually to keep investors and platform owners happy rather than the people creating on them. When your whole business lives in one place, someone else decides how many people see your work.

Unless you have the budget to run everything on paid traffic (in which case, have at it), depending on one platform is a risk. YouTube is one more place my work can live, and one more way for the right people to find me.

Video lets people actually get to know me.

When I had my podcast, the people willing to listen to me talk by myself for an hour were my ideal customers and clients. Video works the same way. When someone can watch you, they connect with you, and if who I am resonates with them, they're far more likely to stick around. That's also why I'm not making an AI avatar of myself. If it isn't me on camera, I'll use my Pixar character, on purpose.

There's another piece of this I keep coming back to. More and more of what I see in my feeds is written, read, and even commented on by agents. My friend Finn Tropy wrote a piece called "I Don't Read Your Newsletter Either" about how much of what we publish gets skimmed or summarized instead of read (including by him, with Claude doing the reading). And when I was talking with my friend Dheeraj Sharma, he told me he can tell when comments and notes are generated by an agent, and he's just not going to bother engaging with them anymore.

As someone who builds with AI every day, I get the irony. But it makes a real person on camera, talking to you, worth more than ever.

It's why my videos are going to end with a blooper reel whenever I have bloopers, which will probably be every time I record. I get to bring in more of my personality and my sense of humor, and the SaySo video below has one, so stick around for the end.

My friend Karen Michaels, who helps people with YouTube and social media, also told me YouTube is really pushing women over 40 right now. She's running a free workshop on November 13th, and if you're thinking about YouTube yourself, she's worth knowing.

YouTube works without me living in a feed.

I'll absolutely respond to comments. But YouTube runs on people watching, not on me posting, commenting, and engaging all day in a way that pulls me off the work I'm trying to get done. I only have so much bandwidth, and I want to spend it on the work.

And yes, YouTube can pay. Monetization and sponsorships aren't the goal, but they're a real bonus, and I don't hear people say that out loud often enough.

The goal is getting in front of my ideal subscribers, customers, and clients. Which brings me to the video that kicked all of this off.

I thought I was going to record a video about SaySo and have it up on YouTube in a day or two.

Instead, I built a teleprompter.

Then I recorded the talking-head portion, recorded separate narration for the product walkthrough, worked with Claude on the edit, went back into the script to change things, and watched more versions of myself than anyone should have to. (I also ran it through Descript's eye-contact correction, which is the one feature keeping me on Descript until my plan ends on November 3rd...)

It took a lot longer than I planned. 😂

And I loved it. I learned an enormous amount, and I still can't get over the fact that I can do all of this myself. But somewhere around the third version, it hit me that making one video and making videos consistently are two completely different problems.

The video is live now, and I finally get to show you SaySo.

If SaySo looks like a fit for your business, you can join the waitlist at getsayso.io.

Why does this video matter so much to me?

If you've been reading here for a while, you know I've built a lot of things with AI over the last 18 months. I was excited about every one of them at the time, and every one taught me something. But SaySo is different, and I knew it almost right away.

It started with a coupon.

I wanted to give a friend a discount on a workshop, and the checkout platform I was using told me I'd need to upgrade my account to create one. A coupon! Around the same time, another friend ran into almost the same wall on a different platform, with a basic selling feature locked behind a higher plan.

That irritation turned into a question: why is it this complicated and this expensive to sell your own stuff online? And once I started building, I could see how much simpler it could be for the person doing the selling.

SaySo is a storefront for digital products, workshops, coaching, and services. You connect your own Stripe account, and you manage your store by talking to Claude or ChatGPT. (Think of it like a WordPress site: your customers see the storefront, and you get the admin... except your admin is a conversation.) No features are gated behind a pricier plan, and SaySo doesn't take a percentage of your sales.

I haven't had this kind of knowing about a product since the Content Creators Planner, which did nearly half a million dollars in sales. SaySo feels like it has even more potential. It started out of frustration, but the moment I saw what was possible, I just knew, and now I want to put it in front of the people it was built for.

Which is where YouTube comes in.

What's different about my YouTube strategy this time?

This time, I know what I want to talk about. I've been on YouTube since around 2009, mostly posting WordPress tutorials, and I've drifted back more times than I can count without a clear answer to that question. The last year and a half of changes in my business finally gave me one. I know who I want to reach, what I want to help her build, and I have real things to show her, starting with SaySo and the tools I'm building for SPARK Lab.

And a video keeps working long after the day I publish it, so I can change the title or thumbnail on what isn't working and let the good stuff keep finding people.

Where is my channel starting from?

A small, inconsistent channel with some very clear signals about what works. Before I planned anything new, I ran my YouTube researcher on my own channel and the space around it. It pulls from the YouTube Data API and VidIQ, and the snapshot I'm working from is dated October 24.

The starting line: 1,660 subscribers, 29 long-form videos, and nothing published between March and September of this year. (Consistency is clearly the first thing to fix.)

The data also showed me a pattern I couldn't argue with. My best-performing videos are specific builds and specific tools, like my WisprFlow review at 376 views and the SaaS I built with Claude Code. My mindset and brand videos sit at around 20 views each. And my best short ever, at 1,173 views, is the one with "midlife" in the title.

So people find me when I show them something specific. Those mindset ideas still matter to me a great deal, but they belong inside a build video, not in its title.

The research surfaced two other things I'm building on. First, in all the channels it looked at, I didn't find a single woman over 50 building real business assets with AI on camera (at least in my research). The building tutorials come almost entirely from young men, and the AI-for-over-50 channels teach consumer basics. That's a pretty clear lane.

Second, one question keeps showing up in the comments on other AI-building channels without a good answer: where does my build actually live, and how do I own it? Hosting, domains, payments, and keeping it running. I've shipped websites, directories, a SaaS, and my own business hub, so that one is going on my list.

This snapshot is what I'll compare against at 90 days and again at a year.

What is each video supposed to do for my business?

Every video gets a job before I record it: who it helps, what it shows, and where it sends people next. That's the biggest shift from every other time I've come back to YouTube, and it's the core of my YouTube strategy. I'm deciding where each video fits in my business up front, not after it's live, and I'm wondering why it didn't go anywhere.

For the SaySo video, the job was clear.

I wanted to introduce the platform and tell the story of why I built it, but I also wanted people to see it working. Telling you that you can sell digital products and manage your store with AI is one thing. Showing you the actual store and the Claude connection is something else entirely.

The next step was clear, too. If someone watches and SaySo looks like a fit for their business, they can join the waitlist at getsayso.io and follow along as I get ready to launch (next week, fingers crossed!).

That gives me something far more useful to pay attention to than a view count. Did the video reach people who want to sell their own products? What did they ask? Did any of them want to know more?

The same thinking works for every other kind of video I'm planning. A tutorial can solve one specific problem and introduce someone to SPARK Lab. A walkthrough can show what's possible with a tool I've built. A behind-the-scenes video can give someone a feel for how I work and what they can expect from me.

If you're thinking about video for your own business (or podcast episodes, or articles... this works for all of it), these are the three questions I'm asking before I record:

• Who is this for, and what will they walk away with? "Women who sell digital products and are tired of paying to unlock basic features" gives me a lot more to work with than "entrepreneurs."

• What can I show that would be hard to explain? If I'm only talking, it might work better as an article. Video earns its place when someone can watch me actually do the thing.

• If they want more, where do I send them? That could be another video, a free resource, a waitlist, or an offer. It depends on the video and the person watching, but it has to be the most relevant next step for that video.

I've been publishing online long enough to know how easy it is to get caught up in making content without ever asking what you want it to do. This time, I'm asking first.

What am I updating on my older videos?

Titles, thumbnails, descriptions, and links, but not the videos themselves. Before I make a batch of new videos, I'm going back through what's already on the channel (even if some of them aren't that relevant anymore).

Some of those videos have been sitting there for a while. My business has changed, my branding has changed, and in some cases the descriptions and links needed real attention. A few feature projects I'm no longer actively building, and I want the descriptions to say that honestly without pretending I never built them. (They're still things I made, and someone might find a useful idea in them.)

I'm updating titles, thumbnails, and descriptions with a handful of questions in mind:

• Does the title say what the video is actually about? The research made it clear that people find me through specific builds and tools, so the title needs to name them.

• Does the thumbnail make the topic obvious and look like my channel today? It's the first thing someone sees, and some of mine were made before my current branding.

• Are the links still accurate and useful? A dead link in a description is a door that opens onto a wall.

• Does the description reflect where the project stands now? Especially for the projects I've moved on from.

• Is there a relevant next step for someone who finds this video today? Even an older video can point someone to SaySo, SPARK Lab, or a newer video on the same topic.

I'm not re-editing the videos themselves. Some of those topics just aren't as relevant today, and I'd rather put that time into making new ones. Updating everything around a video is quick, and it gives the existing library a better shot at reaching the right people.

If you already have content out in the world, try finding one of your own pieces the way a stranger would today. Where do the links take them? Does it still sound like your business? Is there something useful they could do next? You don't have to fix everything at once. I'm starting with a few videos and seeing what I learn.

How much of the video did AI actually make?

A lot of it, but not the parts that make it mine. AI can help me make the video, but I decide what makes it good.

I've been deep in AI-assisted video for weeks, which is how "I need to record a video" turned into "I should build a teleprompter" before the SaySo video was even finished. That's just how things go around here.

Here's how the work split on this one:

• What I did: recorded the talking-head footage, recorded separate narration for the product walkthrough, rewrote parts of the script, watched every version, and made the calls about how the finished video should feel.

• What Claude did: helped build the product demonstration, handled a big chunk of the edit, and created the graphics.

• What Descript did: eye-contact correction, so it doesn't look like I'm reading off a teleprompter (even though I was). I also edited some of the audio since Claude cut too much, lol.

There were bumps along the way (and it was kind of a long couple of days, lol).

At one point I had to move the whole edit into a different Claude environment. I tested it with Claude Code in VS Code, it botche it, then I went back to where I'd already built out a structure and a workflow in Claude Code desktop (I was using it for something else at the time).

Some of the time went to learning things for the first time, and some of it went to caring about the result. This is the first video introducing a product I believe in this much, and I wanted it right.

But if I'm publishing two videos a week, I can't put that much into every one, and I don't need to. A quick tutorial doesn't need the same editing and graphics as a product introduction. I also need to find the happy medium between planning ahead, but letting it be a little more organic.

For the next video, I'm making those decisions up front: what needs to be on screen, what I can simply show and talk through, and which parts actually benefit from a heavier edit. I'm also paying attention to the tasks I keep repeating, because those are the ones worth building into a workflow soon-ish.

I've been building a bigger video production system, along with a visual editor called Scene Studio, and I've put the bigger system on pause. I want to make more videos and find the real sticking points before I spend another week perfecting a workflow I haven't used enough yet (plus, creating a bigger system isn't as important as publishing right now).

How will I know if my YouTube strategy is working?

By tracking what happens after someone watches, not just how many people watch. A video can get views and still leave me wondering whether it reached anyone who cares about what I do. So for every video, I'm keeping a short record:

• The topic and who it's meant to help. One line, specific enough that I can tell later whether I hit it.

• What the viewer gets from it. The answer, the walkthrough, or the thing they can go do afterward.

• The next step I'm offering. A related resource, the SaySo waitlist, or another video.

• The title and thumbnail I started with. That's my baseline.

• Every change I make, and when. If I swap a thumbnail on day ten, I want to know that when I'm looking at day thirty.

Then I'll look at the YouTube numbers next to the business results I can actually track: visits to my sites, email subscribers, SaySo waitlist signups, and interest in SPARK Lab.

The SaySo video makes this easy to picture. The next step is the waitlist, so I'm using a trackable link in the description and comparing those visits with signups. That won't explain every decision a viewer makes, and a bump in signups the week a video goes up isn't proof the video caused it. But it's a much better starting point than staring at views.

A tutorial has a different job. If I make a video answering "Where does your AI-built app live?" I might point people to a related resource or the next video in a series, and the result I care about depends on what I offered.

And no, I'm not tracking any of this in a spreadsheet. I'll pull the data into Claude, have it give me a markdown summary, and drop that into the project where I'm working through my video strategy. That's the whole system for now (and I'm trying very hard not to build an app for it before I've published the next video). 😉

What will I look for at 90 days?

I'd love to see signs of traction in the first three months, but I'm giving this a full year before I expect solid, tangible data. Two videos a week, every week, is the commitment.

At 90 days, I want to be able to answer these:

• Which topics brought new people to the channel? New viewers and subscribers, not just total views.

• Which videos started conversations with the people I want to reach? Comments and questions from the women I'm actually building for.

• Are viewers following the links? To my resources, the SaySo waitlist, and my offers.

• Which videos were fun and manageable to make? I want to know which formats I'd happily make again.

• Where did production take more time than the result justified? That's where any future workflow building will go.

• What do I want more of, what do I change, and what do I stop? The honest version, not the hopeful one.

I'll have the October 24 snapshot, the videos I've refreshed, and everything new I've published to compare against.

If the channel gets monetized along the way, that's the bonus I mentioned. I think it's possible, since the channel's been around a while and already has some views, but the bigger goal is getting the work I'm doing in front of the right people, giving them a reason to stick around, and seeing whether that brings them to SaySo, SPARK Lab, and the things I make.

Questions about my YouTube strategy

What happens next?

If this post gets you thinking about one thing, I hope it's where your own work lives. One platform isn't a business plan, and the more places the right people can find you, the less any single algorithm gets to decide your future.

I'll share what I learn here as I go, including the experiments that work and the ones that send me back to the drawing board. I honestly don't know what the next year will look like, and that's a big part of the fun.

For now, the SaySo video is out, the next one is getting planned, and my YouTube strategy for the next 12 months is officially underway. If you want to follow along as SaySo launches, the waitlist is at getsayso.io.

If you're using YouTube for your business (or thinking about it), I'd love to hear what you're trying.

Let's see what we can do.