Filters and date ranges
Last updated
Filters narrow what you see by date range, engine and tag. They apply to the whole page: the KPI row, the chart, the prompts table, the Sources panel and the CSV files you export. Change a filter and every number changes with it.
How to: Filter your dashboard
The filters
- Date range. Days are UTC dates. Pick 7, 30, 90 or 180 days, or choose Custom and set a start and an end date. The default is 7 days.
- Compare. Set a second period to compare with. See Compare periods below.
- Engine. One or more engines. An engine that is off or not offered has no data. See Engines and collection.
- Tags. Auto tags (branded, intent) and your own tags. See Tags.
- On the Sources panel: Listicles only and Gaps only. They can be combined. See Listicles and Gaps.
Compare periods
The Compare button next to the date range sets a second period and shows how each figure moved between the two. Choose:
- Previous period. The period of the same length that ends the day before your date range starts. For a 30-day range, that is the 30 days before it.
- Custom period. Any start and end date you pick, up to 366 days.
- Don't compare. Turns it off.
What changes on the page:
- Headline boxes. Each box shows its change with an arrow, the size of the change and a word: better, worse or no change. Visibility and share of voice change in percentage points (pts). Average position changes in places, and a lower position is better. Sentiment changes in points on its -100 to +100 scale.
- Chart. Your brand's line for the compare period is drawn dashed behind the current line, lined up day by day, so you can see both shapes at once. Hover a day to read both values.
- Report. The report states the compare period and shows the changes. See Reports and white label.
When a change is shown
A change is shown only when both periods have at least 10 answers behind the figure. For share of voice that means at least 10 mentions of tracked brands in each period. For average position it means at least 10 answers that name your brand, and for sentiment at least 10 answers with a tone label. Below that, the box says "Not enough answers to compare" and shows how many each period has. A small sample moves a lot, so a change from two small samples can be noise. See Sample size (n).
Example
Brindlepath (fictional) compares the last 30 days with the 30 days before. Visibility is 33% against 28%, with 40 answers in each period, so the box shows "+5 pts, better". In the earlier 30 days Pelliwick was not yet a tracked competitor, so share of voice for that period counts Pelliwick only in answers collected after it was added. Read a change in share of voice with that in mind. See Projects.
Rules
- Filters narrow the answers first, then each metric is computed on what remains, so n changes too.
- Active prompts only, always.
- Rows add up: the sum of the rows you see equals the total under the same filters.
- A CSV export follows the filters on screen. See CSV import and export.
Example
Brindlepath (fictional) filters to one engine, the last 7 days and the tag "non-branded". n drops from 120 to 38 and visibility is computed on those 38 answers only. Remove the tag and it is computed on 120 again.