A market dashboard on company signals in Data Studio (Looker Studio)

Google's welcome page says that in April 2026 the Looker Studio product name returned to its original name, Data Studio. This guide uses the current name. It shows how to build a dashboard on the Fokals Market Series dataset (market_series): a BigQuery view for each cut of the data, rate, index and growth charts, an industry filter, and the settings that stop a weekly series from being summed into nonsense. Fokals is delivered direct, by REST API and as bulk files, which you load into BigQuery first, with BigQuery's own loader, as the guide to loading company data into BigQuery describes, and then connect Data Studio to your own table.

What one row of the series is

The market series aggregate the company tables once a week, as of each Sunday. A row is one metric, one dimension, one window and one as-of date, and it is written once. The table lists the columns that matter for charting, and the data dictionary lists them all.

Three properties shape the charts. Rates are taken over a same-store cohort, the companies covered since before the window began, so a rising line does not mean the company index grew. A cut is written only where its cohort is large enough to support a rate, so some cuts have gaps. And index is 100 at the series' first as-of date, so two indexes with different base dates are not comparable.

Shape the data in BigQuery

Build one view for each cut instead of one wide table, so field values are never mixed. This view keeps the industry cut and renames dimension for the filter.

CREATE OR REPLACE VIEW fokals.market_by_industry AS
SELECT
  as_of,
  metric,
  metric_name,
  unit,
  dimension AS industry,
  window_days,
  cohort,
  rate,
  growth,
  index
FROM fokals.market_series
WHERE dimension_kind = 'industry';

The industry values are the ids of the industry label. Take the list from the data with SELECT DISTINCT industry, and keep readable names in a small table that you maintain.

Partition the underlying table by as_of. Google's page on the BigQuery connector says that for a table partitioned by DATE, DATETIME or TIMESTAMP you can set the partition column as the main date filter, and that partitioned tables help charts render faster and cost less. The same page says a data source can read a table, a view or a custom SQL query, that you need a Google Cloud project with a billing account, and that reports can incur BigQuery query and storage costs.

A custom query lets the report's date range reach the database. Data Studio passes @DS_START_DATE and @DS_END_DATE as text, so convert them as Google's parameters page shows. The project name here is illustrative.

SELECT as_of, industry, rate, growth, index
FROM `acme-robotics-prod.fokals.market_by_industry`
WHERE metric = 'hiring_new'
  AND window_days = 7
  AND as_of BETWEEN PARSE_DATE('%Y%m%d', @DS_START_DATE)
                AND PARSE_DATE('%Y%m%d', @DS_END_DATE)

Build the charts

Four charts and three controls cover the dashboard.

• Rate over time. A time series chart with as_of as the dimension, industry as the breakdown dimension and rate as the metric, and as_of as the date range dimension. Google's time series reference says the Group the rest as Others setting is on by default and merges the series beyond the Top N limit into one. Turn it off, because a merged series of rates is not a rate.
• Index. The same chart with index as the metric. A line at 112 is 12 percent above that series' own starting level, whatever the unit of its rate.
• Growth. A bar chart of growth for the latest as_of, sorted from highest to lowest. growth is empty where there was nothing before the window, so the bar for such an industry is missing, not zero.
• Cohort table. The latest as_of by industry with cohort, rate and growth, so a viewer can see how many companies or postings sit behind each figure.
• Controls. A date range control, a drop-down on industry, and a single-select drop-down on metric.

Google's drop-down control page says a single-select list needs a default selected value unless Allow Select all is on. Give metric a default, so a viewer never sees every metric at once. Filters apply at the data source level, not the chart level, as the controls page states, so a control on industry filters every chart that shares the source. The same page says to use a filter property, not a control, to restrict data before viewers see it: fix window_days that way.

Lay the pages out by source, because the families from one source share a cadence. A hiring page can use hiring_new, hiring_closed, hiring_net and hiring_open. A page on roles can use ai_role_share and work_mode_share. A technology page can use tech_added, tech_removed and tech_prevalence with a drop-down on the subject, and a page on announcements and intent can use news_event and intent_surge. The dictionary names GET /api/v1/series/metrics as the live catalogue with the subjects of each family, so seed the metric list from it instead of typing values.

Traps in this data

• Sums. Google's aggregation page says a metric left as None in the data source defaults to Sum in reports. Rates, growth and the index must never be added across rows. Keep one row for each point: one metric, one window_days, one dimension_kind and one industry per series.
• Overlap. Seven-day windows that end on consecutive Sundays do not overlap, but 30-day and 180-day windows do. A line of 30-day points is smooth by construction, and neighbouring points are not independent readings.
• Monthly and yearly sources. Traffic tiers are monthly and headcount is yearly, so the weekly line for those metrics steps instead of sloping.
• Window length. The 7-day window is the most responsive and the 30-day and 180-day windows the steadiest, so choose one window for each chart to match the question.
• Units. rate is in the unit the row names: per 100 companies, a share of postings from 0 to 1, or a median in US dollars. Put one unit on each axis and format shares as percentages.
• Gaps. A missing point is unknown, not zero. Do not fill it.

Because the series change once a week, the connector's freshness setting hardly matters. Google's data freshness page gives 12 hours as the default for BigQuery, and a new Sunday appears within that interval after you load it.

Who may see the dashboard

A report can reach people who have no access to your BigQuery project. Google's data credentials page says Owner's credentials let viewers see a report without their own access to the dataset, and its page on sharing says link sharing can let anyone on the web view a report and that reports can be embedded in other sites. Fokals is licensed by written agreement for internal use, embedding in a product or redistribution. A dashboard open to your own team is internal use, and one open to customers or the web may be another use, so settle which applies before you share a link.

Reading the series

Market Series describes industries, countries, size bands and markets, so it answers which cohort moved. To see the companies behind a movement, go to the company-level datasets: Hiring Activity, Technology Changes and Company News. The weekly tables are written once and never revised, so the dashboard shows what was recorded as of each Sunday. The use cases on market research from weekly series and sector hiring trends show how to read the charts, and the methodology defines every window, rate and index.

Frequently asked questions

The queries and code on this page are examples to adapt. Test them in your own environment before you rely on them.

What this page says about the products it names was checked against their public documentation on 4 October 2026. Product and company names are trademarks of their owners. Fokals is not affiliated with them or endorsed by them.