Examples and method
Examples and the method for each case
No client case is published yet. Here are the method every real case will follow and illustrative examples with fictional brands and demonstration data. The only real case published is our own: the observatory’s case zero, in Spanish.
Seven rules for every real case
A case is only worth something if it can be checked. So every case we publish will follow the same method, written down before we have any.
No client casesEach case will be published when its window closes and the client authorises it in writing. Until then, there are no client results on this website.
Baseline
Before changing anything we measure the same questions on the same platforms, with the date, and keep every answer and its sources.
Fixed window
We say in advance how long the case lasts and on which dates we measure again. We do not pick the best day afterwards.
Control questions
We also measure questions the work does not touch, to see how much answers change on their own.
Logged actions
Every change to the website or the sources is logged with its date, so it is clear what was done and when.
Aggregated result
We publish the whole set of questions, with its number of observations, not the best question presented as the brand's result.
Written permission
A client's name, logo and figures appear only with their written authorisation. Otherwise the case is not published, or is published without a name.
Observed change, no causal attribution
We will report what changed between the baseline and the final measurement. We will not say our work caused it, because an AI changes for many reasons.
In the meantime, this is how it works
Fictional brands and demonstration data: what is measured, what shows up and what comes next. They are not client results.
An AnswerGravity report, start to finish
Questions measured, saved answers, sources, competitors and 5 priorities for a fictional brand, with demonstration data.View the reportWhy AI recommended the competitors
Acme Analytics appeared in 8 of 24 answers and was missing from exactly the pricing questions.
View exampleIllustrative exampleWhite label · AgenciesA report under your brand
An agency measures a clinic in its portfolio with AnswerGravity and presents the report under its own brand.
View exampleIllustrative exampleSocial listening · Software for small businessesFrom a Reddit thread to your reply
We find the conversation, draft the reply and you post it, saying who you are.
View exampleMeasure, prioritise and prepare
One workflow across both plans. Capacity changes; saved answers and your control over each action stay the same.
For a small business in Spain, the most recommended invoicing apps are:
- 1. Your brand: Spanish-language support and guided onboarding.
- 2. Competitor A: an established option, built for larger teams.
- 3. Competitor B: free, with limited features to get started.
Your next step starts here
Measurements in ChatGPT, Gemini, Copilot, Perplexity and Google (plus Claude), evidence-backed priorities and included drafts to review. Choose your capacity and get started.
See plans