Evergreen AI

Context: Evergreen AI is a competitor-driven SEO tool we launched publicly this week after running it internally at our agency for months. You enter your site and three competitors, it finds the keyword gaps, writes articles against them, publishes to your CMS. $149/mo flat, up to 30 articles.

Flat is the part that was hard. Almost every tool in this category sells credits, and having built one, I understand why.

The first version was going to lose money

We built the original on SpyFu. Good product, the data was fine, but two things killed it for us.

Their cost scaled with our usage in a way our pricing didn't. At any real volume we were looking at roughly half of revenue going straight back out as data cost. You can't run a business on that, and you certainly can't run one at a flat price.

The second problem mattered more. SpyFu hands you a packaged view, effectively a top-10-keywords answer, which is built for a human doing research once. We needed to re-evaluate every competitor continuously, because the premise of the product is catching a gap the week it opens instead of the quarter it opens. A rigid packaged answer can't support that at any price.

So we pulled it out and moved to DataForSEO's raw API. We pull rank, traffic, volume and difficulty for up to 1,000 keywords per competitor and run our own logic on top.

Data cost went from roughly 50% of revenue to roughly 10%.

The uncomfortable part is that the expensive version wasn't a mistake. It was a reasonable v1 and it let us validate the idea on real client sites before we'd built any infrastructure of our own. Buy the packaged product first, replace it with raw data once you know the thing works, is probably the right order. Nobody warns you the replacement is coming.

Model routing was the second lever

The article pipeline is 32 steps. Early on most of them called the same model, because that's what you do when you're just trying to get the thing working.

Most of those steps don't need judgment. Deduplicating new calendar entries against what's already scheduled is a semantic comparison. Pre-selecting which of a client's existing pages are relevant to an outline is a filtering problem. Both run on Haiku, and between them they're the bulk of the call volume.

The steps that actually decide something get Sonnet: what type of article this should be, target length, the competitive angle, and the outline itself.

Splitting the pipeline by which steps need to be smart, rather than running one model end to end, was the difference between LLM spend being a rounding error and being a second data bill.

Idempotency is a margin feature, not just a correctness one

Everything runs as scheduled background jobs on their own cadences. Signals get enriched, plans get updated, entries get identified and researched.

The bit I'd underline for anyone building something similar: every job locks its row to a processing state the moment it picks it up. I originally did that because concurrent runs were double-writing. The real cost showed up later. A double-processed article is double the API spend for output you throw away. Once you're paying per call, a correctness bug and a margin bug are the same bug.

Related: each job takes one project per invocation and always picks the most overdue item across all projects. That started as a fairness fix so a busy account couldn't starve a quiet one. It also flattens spend into a predictable hourly rate instead of spiky bursts, which is what makes a flat price forecastable at all.

Why this matters commercially

Credit pricing isn't really a pricing strategy. It means the vendor hasn't got their unit costs under control and is handing the variance to the customer, who now has to ration articles and think about their bill every month.

Getting to 10% is what let us stop doing that.

Try this before you read anything else I have to say. Open Google Search Console, look at your top queries, then go pull up your three closest competitors and ask yourself what they're ranking for that you aren't. If you can answer that in under an hour, you're better at this than most business owners.

That gap is the entire product.

We built Evergreen AI to find it and close it automatically. You enter your site and up to three competitors. It maps what they actually rank for, finds the keywords and pages where they're winning and you're absent, builds a content plan out of that gap, writes the articles, and publishes them to your CMS. No keyword research on your end. No briefs to approve. No SEO knowledge required at any point.

Tagline version: outrank your competitors without touching SEO.

Let me get the objection out of the way

Another AI content tool. I know. The category has earned its reputation, and if your instinct is that this is a slop machine with a competitor-analysis skin on it, that's a reasonable instinct.

Here's why it isn't, and it has nothing to do with prompts.

Evergreen AI didn't start as a product. It started as the engine we run SEO on at our agency, Tactycs. Everything it published went out under our name, on client sites, to clients who would have fired us over a single obviously-AI article. That constraint is the only reason this thing is built the way it is.

The article pipeline is 32 discrete steps. Not one generation call with a nice prompt. Thirty-two, including a slop filter that kills output before it reaches a client site. That level of paranoia is expensive and slow and nobody builds it for fun. We built it because our reputation was the thing on the line, not a churn metric.

An AI content tool built by people with no downside risk produces exactly what you'd expect. This one was built by people who had to answer for it on a call every month.

The proof I actually trust

TubTabs is a Canadian hot tub product company and one of our clients. They ran on this engine for twelve months. From their Google Search Console export:

• Organic clicks up 650%

• Impressions up 1,257%

• Average position from 42 to 7

• 9.5x the number of keywords ranking

One client, one vertical, one year. I'm not going to pretend that's a statistically meaningful sample. It's the reason I believed the engine was worth productizing, and it's real data I can show you rather than a marketing number I made up.

Why we opened it up

Evergreen AI powers the SEO we deliver at our agency, Tactycs. We opened it up so smaller businesses can run the same engine themselves, without needing us in the room.

That's the whole motivation. Tactycs isn't a past life I'm using as a credential, it's the thing still running, and the product is the entry rung. If a business outgrows it and wants people in the room, that's what the managed side is for. Most won't need it.

What you get

$149/mo flat. No credit-based pricing, no per-article metering.

• 3 competitors tracked continuously

• Up to 30 articles per month

• Up to 10 page enhancements per month on your existing pages

• Publishing to WordPress, Sanity, Shopify, Webflow, Wix and Ghost

• Google Search Console connected for real performance data, not estimates

• Webhook integration coming for everything else

There's a free tier, and I'll be precise about what it is so nobody signs up expecting more: you get a couple of keywords and one article, written. That's it. It isn't a trial of the full engine and it won't go find your competitors for you. It exists so you can read the output before you take my word on any of this.

Today is the first day this is available to anyone outside our agency.

Where this is at

This is day one publicly. Up to now it has only ever run on sites we manage ourselves, where I could see every article before it went out and fix anything that went sideways. It has never run on a site I don't have my hands on.

That's the honest gap, and I can't close it on my own.

So let me make it a trade instead of a favour.

I've spent this whole post telling you this isn't a slop machine. The free tier writes you one article. Read it, then tell me where it falls down, in the comments or by email, and be specific. In return I'll do the thing I normally bill for: I'll look at your site and your three closest competitors myself and send back what I'd actually do about the gap. Not a generated report, me. Whether you ever pay for anything or not.

I can do that properly for about the first ten people, so I'd rather be straight about the number than quietly stop replying.

If you'd rather skip all that and just have it running against your competitors, that's the $149. I'd rather have five customers arguing with me about the output than five hundred free accounts sitting idle.

Happy to get into the pipeline, the slop filter, the agency-to-product transition, or the pricing math in the comments. Including how it stacks up against Outrank, if you've used it.

tryevergreen.ai