Postprism is simple to set up, but it covers fewer networks than the larger suites.
Which social media scheduler supports the most networks?
All features
Brand perception
Brand perception. What AI says about your brand.
See what each answer claims about you and your competitors, good or bad, with the exact sentence it came from.
Opinions about you
23
Traits with a criticism
3 of 6
Discussed for others, not you
3
Where you are ahead or behind on each trait
2 ahead2 behind, of 8 traits
Platform coverage2 v 3.5Behind
Ease of use5 v 2.3Ahead
Scheduling4 v 2.0Ahead
Analytics1 v 2.7Behind
Pricing3 v 2.5Level
Customer support2 v 1.0Too few
AI captions0 v 1.0Too few
Social listening0 v 1.0Too few
Traits discussed for other brands, never for you
3traits with no opinion about you
- AI captions2
- Social listening2
- Agency workflows1
What they actually said about you
6traits, 3 with a criticism
Traitcontains
| Trait | Praised | Criticised | vs competitors | What they said |
|---|---|---|---|---|
| Platform coverage | 2 | 3 | Behind | It covers fewer networks than the larger suites. |
| Analytics | 1 | 2 | Behind | Its reporting is lighter than most rivals offer. |
| Ease of use | 5 | 1 | Ahead | The calendar view takes a little getting used to. |
| Scheduling | 4 | None | Ahead | Queueing a week of posts is quick. |
| Pricing | 3 | None | Level | Its plans suit small teams. |
| Customer support | 2 | None | Reviewers say replies come quickly. |
Platform coverage
In your favour 2Against you 3
Teams posting to many niche platforms may find Postprism too narrow.
Best scheduling tool for a brand on six social networks
Agencies with clients spread across many channels may outgrow Postprism.
Is Postprism a good fit for a social media agency?
Who else was judged on this
4| Brand | Praised | Criticised |
|---|---|---|
| NNorthline | 6 | None |
| KKestrel | 4 | 1 |
| EEverdene | 3 | None |
| HHalden | 1 | 2 |
Illustrative.
Being named is half the story
An assistant can name you and still steer the buyer elsewhere. It can call a rival the best for teams, or call your pricing high. Those sentences shape the decision, and a mention count cannot see them.
For example, two brands can each be named in the same number of answers. If the assistants praise one for its reporting and call the other expensive, the buyer leaves with a clear favourite. A mention count would score them as equal.
Two brands named in the same number of answers can be judged very differently.
Claims for and against two brands that were each named in 10 answers, in a made-up example.
- Brand A
- Brand B
6
1
1
5
Favourable claimsUnfavourable claims
How a claim is recorded
Every claim is tied to a sentence you can read. That is the rule the whole feature is built on.
- A model reads each stored answer and lists what it says for or against each brand it names.
- Each claim is filed under one of your topics, such as pricing or analytics, as favourable or unfavourable.
- Each claim must quote the sentence it came from. Agentcites looks for that sentence in the answer, and drops the claim if it is not there.
- Plain description is not recorded. "It supports nine platforms" is a fact, not a verdict, and would only fill the charts with noise.
This is the one judgement on the dashboard. Everything else is counted. So every judgement shows the sentence it rests on.
What a claim looks like
The rows below are made up to show the shape of the data. They are not from a real scan.
| Brand | Topic | Verdict | The sentence |
|---|---|---|---|
| Brand A | Reporting | Favourable | "Brand A has the clearest reports of the tools listed here." |
| Brand B | Pricing | Unfavourable | "Brand B gets expensive quickly as your team grows." |
| Brand B | Integrations | Not recorded | "Brand B connects to nine platforms." This is a plain fact, not a verdict. |
One sentence can hold a claim about more than one brand. When it does, it counts once for each brand it judges.
What you see on the screen
The Reputation screen, on its Praise and criticism tab, puts every trait on one table: what was said for you, what was said against you, how you compare, and the sentence itself.
Open a trait and you read every sentence behind it, praise and criticism apart, each with the assistant that wrote it. Under them sit the other brands judged on the same trait. A criticism only you got reads differently from one the whole field got.
Favourable and unfavourable claims for you and each competitor, by topic.
Each claim opens to the sentence it rests on and the answer it came from.
A panel that sets this scan beside the last one, by assistant, each count beside the number of answers it came from. It shows two counts and no arrow, because a difference of one claim is one answer worded differently.
What it does not do
It does not record neutral description, which is most of what an answer says.
It only judges brands the scan already tracks. A brand nobody listed is not judged until it joins your competitor list.
It is a model’s reading of a sentence, not a measurement. That is why the sentence is always shown, so you can disagree with it.
It does not say whether a change between two scans is real. One scan’s claims can move by a few on wording alone.
Answers from the assistants your buyers use
Each question goes to each assistant you choose, with web search on, the way a buyer would ask it.
- ChatGPT
- Perplexity
- Gemini
- DeepSeek
Questions
How do I find out what ChatGPT says about my brand?
Ask it the questions your buyers ask, keep every answer, and read each one for what it claims about you. Agentcites does this every week and files each claim with the sentence it came from.
Is this sentiment analysis?
It is close, but narrower. It records only claims for or against a brand on one of your topics, and it keeps the exact sentence. Plain description is left out on purpose.
Can a model invent a claim?
It can try. Every claim must quote a sentence that is really in the answer, and one that is not found is thrown away before it is stored.
Read more about getting named in AI answers
Free first scan
See it for your own site
We ask 20 of your questions to four AI assistants, and you see every answer before you pay anything. You do not need a card.