OuterciteOutercite

Prompts

The questions you are measured on. Prompts are grouped into topics, checked on every surface your plan covers, and read as a grid rather than a list.

A prompt is a question you want to be the answer to. It is not a keyword, it is the whole question, phrased the way a buyer would ask it, and your own name is never in it.

What you'll learn

  • What makes a good prompt, and why your name stays out of it
  • How topics group prompts, and why the grid matters more than the list
  • Where prompts come from if you do not want to write them all yourself

Landscape, topic, answers

The Prompts lens descends like every other lens. The landscape is your topics. Open one and you get that topic: the prompts inside it, how each is doing, and on which surfaces. Descend again and you are reading the answers themselves.

The grid is the part people underuse. Prompts down the side, engines across the top, one reading per cell. A list tells you which prompts you are losing. The grid tells you where you are losing them, and a prompt losing on one engine is a completely different problem from a prompt losing on five.

What a good prompt looks like

Write the question, not the phrase:

WeakBetter
melbourne physioWho is the best physio in Melbourne for a running injury?
cloud accounting softwareWhat is the best cloud accounting software for a five-person firm?
commercial solarIs commercial solar worth it for a warehouse in Queensland?

Three rules that matter:

  1. Never name yourself. A prompt with your brand in it asks the engine to talk about you, which it will, and the result measures nothing. Branded prompts have their place for reputation monitoring, and they are classified separately for that reason.
  2. Match a real decision. The prompts worth tracking are the ones a buyer asks before choosing. See intent classification.
  3. Be specific enough to be answerable. "Best software" gets a generic answer that names nobody. Add the buyer, the constraint or the place.

Where prompts come from

  • Write them. Add a prompt directly, assign it to a topic.
  • Discovery. The platform proposes prompts from your site, your category and the answers it has already seen, and you accept the ones that match your buyers.
  • From what you are losing. A contested prompt that a rival wins is often the best next prompt, and the Competitors lens surfaces those.

Start narrow. Fifteen prompts that map to real buying decisions beat two hundred that cover the category, because every prompt costs coverage on every surface every cycle. See coverage and credits.

How often a prompt is checked

Checking runs on a cadence set by your plan rather than on demand, across each surface your plan covers. That is why the assistant answers from stored answers with dates attached: it is reading what the monitoring already produced, not asking the engines while you wait.

Try this in Outercite

Ask the assistant "Which prompts am I losing, and who is taking them?", then open the grid to see whether the losses cluster on one surface.

Was this helpful?

Last updated on