Boston • New York • Detroit
Boston • New York • Detroit
Visibility + Engagement = Opportunity
You become the answer it gives back. And the name a colleague would have offered anyway.
The Definition
Recommendation Marketing is the practice of building a business that earns recommendation from the people who advise your buyer and from the AI systems your buyer now asks first.
Awareness asks to be seen. Recommendation asks to be vouched for, and a vouch costs the person who gives it something. They put their own credibility behind you. That only happens when they can describe you accurately in one sentence and defend the choice if someone pushes back.
That used to happen in a conversation. Now it happens in a search result, a model’s answer, a Slack thread you were never in, and a procurement shortlist you never see. The requirement is unchanged. What changed is how many places it gets tested without you in the room.
The Shift
You still have to be worth recommending. What changed is who does the reading, and how little patience it has for a message that takes work to parse.
Then
Now
A model cannot ask you a follow-up question. It works with whatever it can find and verify, which is the same position a buyer is in before they call you.
The Mechanism
A person recommends you when they can explain what you do without hedging, point to work you have actually done, and trust the referral will make them look good. A model runs on those same three inputs. Build for one and you have built most of the other.
Fig. 01 · Two readers, one set of signals
Both paths draw on the same three inputs. A person needs to defend the referral and a model needs to verify the claim, so one body of work serves both and the effort compounds.
Research basisAggarwal et al., “GEO: Generative Engine Optimization,” ACM SIGKDD 2024, which tested how source-level signals such as citations, statistics and quotations affect visibility in generative engine responses. Reported gains vary by domain.
The Pillars
Recommendation Marketing works in three stages. Skip one and the next has nothing to stand on.
You can make the best pencil on the market. If seven people see it, that is the end of the story. Attraction earns attention by showing you understand the problem before you mention the product.
Its job: Get seen by the right people.
Everyone has peers and competitors. Articulating why a buyer should choose you is the whole contest. Experience, expertise, and a solution you can evidence move someone from aware to convinced.
Its job: Make the choice defensible.
An ask is fine once the buyer already believes you care. Time it well and a direct next step is something they were waiting for.
Its job: Turn interest into a conversation.
What It Includes
Each component is assigned to one of the three jobs and scoped to what your engine actually needs. None of it is sold on its own.
01 Attraction
02 Consideration
03 Conversion
Every component serves one objective: building a business people recommend.
The Fit
It compounds, which makes it powerful for some businesses and the wrong instrument for others. Worth knowing which one you are before you spend anything.
A strong fit if
A poor fit if
The How
Five parts, one system: find the angle, build the engine, create and place, farm the leads, compound it. Including which sources each AI engine actually reads, and what it takes to appear in them.
Questions
SEO optimizes for a ranked list of links. Recommendation Marketing optimizes for being named inside an answer, whether that answer comes from a search engine, an AI system, or a person. SEO stays in the mix because some engines track organic rankings closely. It does one job inside a larger system.
Answer Engine Optimization and Generative Engine Optimization are techniques for improving how AI systems retrieve and cite you. Both are tactics inside this discipline. Recommendation Marketing covers the human referral and the machine citation together, since they run on the same underlying signals.
No, and be skeptical of anyone who does. These systems are probabilistic, their source mixes change, and no vendor controls them. What you can influence is whether you are clear, verifiable, and present in the sources a given engine draws from. We report on movement in those signals and we do not promise placement.
Message clarity lands immediately, often in the first strategy session. Content and placement compound over quarters, since third-party sources run on their own publication timelines. If a quarter-end number is the only thing that counts, this is the wrong instrument.
Usually not. Most of what is already running works fine on its own and simply is not connected to anything. The first move is normally to point existing activity at one message, then fill the gaps.
Scope varies with how much of the engine already exists. The strategy session is thirty minutes and free. You leave with your angle and your first three moves whether or not we work together.
The Strategy Session
Thirty minutes. Bring the business. Leave with the one thing worth building everything else around.
30 minutes · No pitch deck · No obligation
