What we observe and what we measure
We begin by defining the question. An observation records what appeared under particular conditions. A measurement adds a sample and clear counting rules. Our interpretation explains why the finding may matter to a business. Keeping these steps visible helps readers follow the reasoning.
From a query to a useful page
We read search estimates alongside their source, market and date. Then we examine the task behind the query and the content already available. Related questions can belong on one substantial page; a separate resource needs a distinct purpose and something useful to add.
Recording Search and AI results
We record the query, platform, market, date and checking method. Traditional results, AI Overviews and AI Mode are examined separately. A result feature reported by one tool describes that observation; comparisons need a consistent scope.
Reading answers and citations
We preserve the answer and follow its source links. A mention, a citation and a recommendation describe different things. Wording, language and conversation context can affect the next answer. Unavailable cases stay unknown so they do not distort the sample.
Working with dated datasets
Every data-backed study identifies its dataset, collection period and sample. Imports are stored as dated versions. Earlier evidence remains available when a new version arrives. Change analysis needs at least two comparable snapshots.
Keeping the market clear
An English presentation of Bulgarian research still describes Bulgaria. A translated keyword cannot transfer demand to another market. We identify the market behind each dataset and use separate evidence for cross-market comparisons.
Public sources and permitted use
These editorial resources combine public technical guidance with Z-Consult interpretation. A future imported dataset needs its actual source, date and publication approval. Raw exports and exact provider metrics also need confirmed rights for the intended public use.
Who is responsible for the material
We identify the author and separate sourced facts from observations and professional interpretation. Numbers need evidence readers can trace. Publication and material-update dates describe changes to the resource; an expert-review date is added only after that review takes place.
How far a finding applies
We explain the limits that affect the conclusion. For example, an AI snapshot records a moment, while answers can change later. When traffic and sales move together, we also check campaigns, prices and seasonality before attributing an outcome. Rankings and AI citations remain platform decisions.
Corrections and new evidence
We correct factual errors and record a new update date when sources, methodology or interpretation change materially. Study evidence remains versioned, so a later import does not rewrite an earlier result. For corrections or feedback: office@z-consult.bg.
