Why I built a company around being recommended

Visibility is easy to count and easy to sell. I built Subscribe PR around the harder thing, being recommended, and here is how to find your own version.

Caleb Hellinger seated on a rooftop ledge at night

People sometimes ask me what Subscribe PR actually sells. The honest answer is that it does not sell press, and it does not sell rankings. It sells the moment a stranger says "you should talk to this person," whether that stranger is a friend, an editor or ChatGPT.

That is a strange thing to build a company around, because you cannot count it easily. Press clippings are easy to count. Impressions are easy to count. A recommendation is a sentence someone says out loud in a room you are not in. My view is that it is still the only outcome worth building around, and the rest of this post is why, and how you can work out what yours sounds like.

Visibility is a number. A recommendation is a decision.

I will start with a number, because it shows the trap. Across client campaigns we have seen 183M+ Google impressions. I am proud of that figure, and I also think it is the least important thing we do.

An impression means a page was shown. Nobody decided anything. A person scrolled past a result, and maybe they noticed a name and maybe they did not. Nothing in that number tells you whether anyone trusted anyone.

A recommendation is the opposite. It is a person or a tool putting their own credibility behind your name. Someone who says "go with her" is spending something, even if it is only a little of their reputation. That is why a recommendation closes business and an impression usually does not.

I am not saying impressions are worthless. They are how people end up in a position to recommend you. I am saying that when you build a business around the count, you tend to get very good at the count. You get a report that goes up and a client who still has an empty calendar.

The market is now two kinds of recommenders

For most of my career the recommenders were people. A colleague, a client, a friend of a friend. Subscribe PR is built for a world where there are two kinds, because the second kind arrived fast.

People now ask ChatGPT, Claude, Perplexity and Google AI Mode who to hire, which clinic to book and which brand to trust. These tools answer in a sentence or two, and the answer reads like a recommendation because it is shaped like one. I think of that sentence as the new front door. I made a related argument in AI tools don't rank you, they vouch for you, and it is the idea this company is organized around.

What the two kinds of recommender have in common is the thing they lean on. A friend vouches for you because of what they have seen you do. An AI tool vouches for you because of what independent sources have written about you. Neither one takes your own description at face value. Both want proof that sits somewhere you do not control.

That is why we combine editorial press with AI search optimization, the work often called AEO. The press creates the proof. The AEO work makes sure the proof is stated clearly enough that a tool can find it and repeat it. Take either half away and the other half does less.

Our numbers reflect that pairing, and I will describe them with care. We have recorded 27M+ AI citations across client campaigns, and 253.5K AI citations in 60 days for one client, starting from zero. I am not telling you a result like that is a promise for anyone else. I am telling you it is a counted proxy for the thing I care about, which is a tool choosing to name someone.

What I got wrong about my own offer

For a long time I described what I did in the vocabulary of the industry. Placements, coverage, reach. It is the language buyers expect, and it was easy to slip into.

The trouble is that it describes the work, not the result. When I say "we place you in publications," I am describing a task. The result a client actually wants is that the next person who looks for someone like them lands on their name and feels sure about it. I now try to write every description of the work from the point of view of the person doing the recommending.

That shift sounds small, and it changed how I judge what to do. A placement in a well-known outlet is only useful if it makes the sentence easier to say. If I cannot picture the sentence, I should question the placement. I wrote about a version of this in Why I stopped caring about vanity placements.

I should be plain about the limit here. I cannot guarantee what a particular person or tool will say about anyone, and I would distrust anyone who claims to control it. What I can do is make it more likely that the true, specific, flattering thing is the easy thing to say.

The question to ask about your own business

Here is the exercise I would run this week, and it works whether you ever hire anyone.

  1. Write the sentence. Finish this line the way a happy client would say it to a friend: "You should talk to ___, because ___." Do not write the version you wish they said. Write the one you honestly think they would.
  2. Ask three real people to finish it. Pick past clients or colleagues. Do not coach them. Compare their sentences to yours.
  3. Ask AI the same thing. Type a plain question into ChatGPT or Perplexity about who to hire in your field and city, then ask what it knows about you. Notice whether it can complete the sentence at all.
  4. Look for the gap. If the sentences differ, your message is unclear. If people can finish it but the tool cannot, your proof is stuck somewhere a tool does not look. Those are two different problems with two different fixes.

The first problem is about clarity, and you can fix it for free. The second is about independent sources, and it is slower, because it means getting other people to say it in places that count.

Why it matters to me

I lost a business a month after moving to Arizona, and I have written about that elsewhere. I will only say here that it left me with a durable respect for who vouches for whom. Skill did not rebuild my footing. Other people willing to say my name did.

That is the whole reason this company exists. I would rather help someone become easy to recommend than help them look busy.

So here is my question for you. If a stranger asked your best client about you tomorrow, what is the one sentence you hope they say, and would they actually say it?

Want people to find you when they ask AI who to trust?

Keep reading

More from Caleb