Imagine discovering that AI cites your brand often — and celebrating. Now imagine reading what it says: "Brand X is an option, but users report slow support and above-average prices." You appeared. And maybe it would have been better not to. That's why measuring presence without measuring sentiment is telling half the story, and sometimes the wrong half.
Presence and Sentiment Are Two Different Axes
Every time AI mentions your brand, there are two independent questions. The first: did you appear? The second: what did it say? The two don't necessarily move together. You can have high presence with bad sentiment — AI talks about you a lot, and talks badly. You can have low presence with great sentiment — appearing rarely, but always as a praised reference.
Treating presence as the whole story leads to wrong decisions. A brand can invest to appear more and, without realizing it, be amplifying a negative narrative. More reach for a bad message is a loss, not a gain.
How AI Forms That Tone
A response's sentiment isn't invented by the model; it reflects what's in the sources. If most of what's been written about you is positive, AI tends to reproduce that. If there's a pattern of complaint — about price, support, a known issue — that pattern leaks into the answer, even if you resolved it long ago.
There's an important subtlety here: AI sentiment is often lagged relative to reality. An old crisis, a rough patch already overcome, a criticism that went popular at some point — all of it can keep coloring answers long after it stopped being true. AI speaks of your reputation as it was recorded, not necessarily as it is today.
The Tones That Matter Beyond "Good" and "Bad"
Reducing sentiment to positive or negative loses valuable nuance. In practice, a few tones deserve specific attention.
The hedge. "It's a good option, but..." The "but" is where the information lives. AI often cites with a condition attached, and that condition is what may be costing you conversions.
The unfavorable comparison. Always appearing as the cheaper alternative, or as "similar to the leader, but smaller," shapes perception even without being technically negative. The framing is a kind of sentiment.
The association with a problem. When your brand is cited mainly in the context of a controversy or a limitation, even factually, the effect is negative by context. It's not what's said; it's where you appear.
Lukewarm neutrality. Being cited without enthusiasm, as one more name in a list, isn't negative — but it doesn't help either. Sometimes the problem isn't bad sentiment; it's the absence of positive sentiment.
Why This Is Manageable
The good news is that sentiment isn't destiny. Because it reflects the sources, it responds to changes in the sources. Genuinely fixing the issue that drives complaints, building new positive presence, ensuring the correct and current information is available in the sources AI consults — all of it, over time, shifts the tone.
But the first step is knowing. You can't correct a negative narrative you don't know exists. Many brands operate with an AI image far worse than they imagine, simply because they've never read what AI says — they only checked whether they appeared.
Measure the Tone, Not Just the Frequency
Mature AI-presence monitoring tracks both axes together: how often you appear and in what tone. It's that cross-reference that tells you whether more visibility is a good idea now or whether, first, you need to fix what AI is saying.
Tracking sentiment over time also reveals movement: is the tone improving after your actions, or is a negative narrative gaining strength without you noticing? That's the kind of alert worth gold, because a reputation problem caught early is a cheap problem to fix.
Genoma measures sentiment alongside presence, precisely so you don't fall into the trap of celebrating visibility that is, in fact, spreading the wrong message. Appearing matters. Appearing well matters more.