I built a requirements traceability tool for regulated industries. It runs on your laptop. It connects to the spreadsheet your engineers are already using, reads your BOM and SRS and ATP output, parses git blame to map requirements to source code, and runs an MCP server so you can talk to your traceability data in natural language. It handles AS9100, ISO 13485, DO-178C, IEC 62304, ISO 26262, and Automotive SPICE. It costs less per year than a single day of a Jama implementation consultant’s billing rate.

It works. I know it works because I’ve been shipping software in regulated environments for more than a decade, including FDA-regulated insulin dosing algorithms at TypeZero and Dexcom (hello ISO 13485), and now at Allocortech (hello AS 9100D). I built RTMify.io, the thing I wished existed every time I watched a quality engineer spend three weeks manually wrangling an RTM in Excel before an audit. RTMify doesn’t replace the spreadsheet. It embraces the spreadsheet as the source of truth it was all along.

But, Houston, we have a little problem. I have no way to tell you about it.

I am altering the deal. Pray I don’t alter it any further.

Google Vader, you old rascal.

The old web had a deal. You build something useful, you publish it, search engines index it, people find it. The deal was reasonably fair: you provide valuable content or a service, the infrastructure provides the discovery, users provide the traffic, and everybody eats. A gal with a bright idea and a domain name could stand up a website and get customers.

That deal is stone dead. Pour one out.

Google answers in the SERP now. Your product page that you sweated over for a week, your painstakingly honest comparison with competitive tools, your landing page that explains exactly why a should care, and how your product kicks ass: Google reads it, extracts the answer, and serves it in a blue box at the top of the results. The user gets the information.

You get nothing. No click. Bupkis. Google ate your lunch and didn’t even leave the ketchup packet.

Chatbots are worse. At least Google leaves a link somewhere below the fold. ChatGPT and friends ingest your entire value proposition, paraphrase it into smooth BCG-inspired corporate paste, and serve it to the user with no attribution, no link, no awareness that a specific human built a specific product to solve a specific problem. The user gets a summary. You get scraped. (I’m told this is “democratizing information.” Sure.) And NOW… god this is just the icing on the cake… NOW Cloudflare wants to sit at the permissions layer and govern which machines can access your content, which would be hilarious if it weren’t also a toll booth on top of the toll booth.

The publisher still bears all the cost. Thinking, building, writing, hosting, maintaining. The intermediaries strip-mine discovery, synthesis, and access. The economics have inverted. The people who make things subsidize the people who index things, and the indexers have decided they’d rather keep the audience than share it.

Four options, all bad

So you’re me. You built something real. Currently, I need 200 quality engineers at Tier 1 aerospace suppliers who are drowning in Excel RTMs to know I exist. Maybe you have a similar problem.

Here are your options in early 2026:

Pay the SEM toll. Fund the extraction layer directly. Pay Google to send you the click they used to send for free, because they broke the thing that used to send it for free, and now they rent it back. Google Ads B2B cost-per-lead is north of $800 and climbing, with customer acquisition costs across B2B SaaS up 60% in five years. Pay the arsonist to let you back into the charred remains of your own house.

Play the increasingly surreal engagement game. Write LinkedIn posts about “the future of requirements traceability.” Fuck, actually no, it’s worse, circa early 2026 on LinkedIn. Go ahead, post a CAT VIDEO. Then, add a hook, add a listicle, add a CTA. Get impressions from people who will never, in their entire careers, need your product or service. The algorithm rewards you with more of the same non-audience, and you spend your evenings optimizing for a metric that has nothing to do with selling software. Go on, pimp yourself out to the engagement grift. It’ll be fine. Really.

Retreat to meatspace. Conferences. Handshakes. Trade shows. Rotary Club meetings, literally. Maybe this works in some industries because the buyer pool is small and trust-dependent. But it doesn’t scale. Rotary clubs, even “online communities” are not a go-to-market strategy. They are nonscaleable lifestyle-adjacent parasocial tar pits.

Give up. Pack it in. Fuck it, just be a wage slave.

These are the options. All of them. I have looked for a fifth. There isn’t one.

And before you say “product-led growth”: Yes, thanks, I’m aware. You can’t have product led growth if the top of the funnel is dependent on Google’s scraps of traffic. The RTMify product ladder is as sharp as it possibly could be. I designed it as a wedge-and-expand funnel because I’ve been shipping B2B software for twenty years and I know how this is supposed to work.

It’s still not enough. The ladder needs a first foot on the first rung, and the first rung is discovery, and discovery is what broke.

The Silence Where the Fifth Option Should Be

I searched. Seriously. Every “how to market indie software in 2026” article recommends the same four things I just described, wearing different hats. “AEO blogging” is just SEO after AI ate the old version. “Community building on Reddit and Discord” is Rotary Club with a keyboard and a six-to-twelve-month runway before it generates a single lead. “Build in public” is engagement bait that works for developer tools sold to developers who are already on Twitter; it does not work for compliance software sold to quality managers at medical device companies who have never heard of indie hacking and never will. “Referral programs” require having customers to refer. That’s the chicken. The egg is distribution.

Every channel has collapsed into one of two failure modes: monopoly rents or engagement death spiral. SEM is monopoly rents. “Social” (quotes deliberate) is the death spiral. Email is somewhere in between, with deliverability cratering and every inbox drowning in AI-generated slop. SEO is gone. App stores take 30% and bury you. Review aggregators like G2 are consolidating into their own monopoly extraction layer. There is no channel left where the relationship between “build something valuable” and “people find out about it” is direct and fair. Oh let’s not forget Cloudflare’s cynical play.

Sit with this for a second. You’ve got a specific product for a specific customer. It solves a real problem.

The mechanism by which a customer would have discovered it, ten years ago (heck even five), is gone.

Yes, this essay is a complaint.

I’m pissed.

You know what the internet looks like now? A mall where all the stores have been replaced by vending machines that sell photographs of the items that used to be for sale. The photographs are generated by AI. The vending machines are operated by Google and Meta. The store owners are still paying rent.

I don’t have a lesson here. I don’t have a “here’s what I’d build to fix it.” The web is fundamentally broken for people who make things, and nobody with the power to fix it has any incentive to, because the breakage is where the money is.

I’m going to keep building RTMify. I’m going to keep going to conferences and writing on a blog that gets 90% of its traffic from LinkedIn, which is itself an engagement-optimized feed that will happily show my post to 10,000 people who don’t need compliance software and suppress it from the 50 who do. And I’m going to be pissed about it, because being pissed is the correct response to a system that rewards making things less than it rewards capturing the attention layer above the things.

If you’re a quality engineer at a regulated company and you’re reading this: rtmify.io. Fifteen megabytes. Runs on your laptop. Your data stays local. Works in a SCIF. I couldn’t afford the Google ad to tell you about it, so I’m telling you here, in the only way I have left, which is to write an angry essay and hope the algorithm doesn’t bury it.

That’s where we are.