How Orkie's pieces work together

Most businesses run on a stack of separate apps: one for the feed, one for email, one for reviews, one for reporting. Each is good at its own job and blind to everything else. The feed app does not know which products actually sell. The email tool guesses from its own clicks. Nobody measures the product suggestions. Orkie is built the other way around: every piece reads the same set of data, so what one piece learns, the next one uses. Here are the five connections that matter most, with an example of each.

1. The pixel decides what the feed pushes

The pixel runs on your own domain and counts every verified order once. That makes it the one place that knows which products actually sell, not which ones get clicked.

The feed reads that. Products that sell get pushed harder to Google Shopping, Microsoft, Meta, TikTok and ChatGPT. Products that only draw clicks get labeled and pulled before they waste more budget.

For example: a product had 900 clicks last month and two orders. A standalone feed app sees a popular product. Orkie sees a product that costs money and does not sell, so the next campaign leaves it out. You see the change, with the numbers, before it goes live.

2. Related products you can actually measure

On Shopify and BigCommerce stores, the "you may also like" and "goes with" spots on product and cart pages follow rules you set in plain words:

• Same brand only.
• Only parts that fit, matched on the attributes you choose.
• Never show a certain tag.
• Or the exact products you pick to go together.

Because the same pixel sees what people buy afterward, you can ask whether those suggestions sell, instead of guessing.

For example: you tell Claude "only suggest parts that fit the same model." It shows you the new rule and what would change before anything goes live. You approve it, and the pixel reports what the new suggestions sold.

3. Your customer lists become ad audiences

The CRM builds lists from real behavior: high-value repeat buyers, people who bought once 60 days ago, people who looked three times and never bought. Those lists go to Google, Meta and TikTok as audiences, with ads consent checked for every person.

That does two things at once. You show buyers the next thing they are likely to want, and you stop paying to show them an ad for what they already bought.

For example: your "bought once, 60 days ago" list goes to Meta for the refill ad. You approve the list before it goes.

4. A personal email about the exact product

The pixel sees what each shopper looked at and what they left in the cart. The Outbox uses that to send an email about that exact product:

• The cart they walked away from, with the items they left.
• The product they viewed but did not buy.
• A follow-up when they asked for help on your site.
• The item that goes with something they already own.

Every email goes from your own domain, consent is checked at the moment of sending, and each one is measured to the order, not to the click.

For example: a shopper leaves running shoes in size 10 in the cart. That evening they get an email with those shoes, in that size, and a way to ask a question. If they buy, the pixel credits the email once.

5. The company brain keeps learning

This is the connection that ties the others together. Every test, campaign and change ends as a lesson in your company brain: what worked, what did not, and why. Goals and facts get checked again as the business changes, so the picture stays current.

The brain only keeps what someone on your team confirms. The AI proposes a lesson, you say yes, and from then on every chat, every campaign and every person who works in the workspace starts from it. It lives in your own workspace and works the same whether you use Claude or ChatGPT.

For example: before the next spring sale, it reminds you that last year's white-tee test sold a quarter as well as black, so the budget stays on black. Nobody had to remember.

Why this matters more than any single feature

Every one of these connections is small on its own. Together they change how the business learns. A result in one place shows up in the next decision somewhere else, and the lesson stays after the campaign ends and after people move on.

You also stay in charge the whole way. Feed changes, recommendation rules, audiences and emails are proposals until a person approves them.

You do not need every piece on day one. Each connection works as soon as both of its pieces are installed, and most businesses start with the pixel. See all five on the How it works together page, or talk to us about which ones matter most for your business.

---

Want this for your business? Talk to us How the pixel works