Company
You have spent years learning to reach people. But a model now reads you first and decides whether those people ever hear your name. Almost nobody is marketing to it yet.
What we are for
If you deserve to be recommended, you should be recommended.
Measuring the gap is the easy half. Closing it is the product.

A letter from the founders
For twenty five years, search worked one way. You typed a question, you got ten links, and you picked one yourself. Everything marketers learned was built for that moment.
That moment is going away. Now people ask a model instead. It reads the web for them, picks a few companies, and explains why those companies are good. Most people never click anything.
So there is a new decision happening before your buyer ever meets you. A model decides whether your name comes up at all. When it does not, nobody tells you. You just get fewer customers and no idea why.
We watched this happen to friends running good companies. They were doing everything right and still going quiet. We looked for something that could show them what the models were saying, and then help them change it. Nothing did both. So we built Prefer.
Why this matters now
1998
The web opened up
Google ranked pages by who pointed to them. A stranger with a good page could beat a company with a budget.
2012 to 2024
We learned the rules
Thirteen years of watching what search rewards. First for our clients, then for our own companies.
2025
People stopped clicking
Answers replaced links. One reply, three brands named, and no page two to fight your way onto.
Now
The answer is still soft
Models have not made up their mind about most brands yet. Whoever shows up now helps write it.
Next, models will not only recommend products. They will use them and buy them on your customer's behalf. Brands that are readable to machines, and not just to people, get chosen twice. That is what we are building for.
See how answer engines describe your brand today, and where the openings are to outpace the competition.