Film campaign measurement coverage and its limits

What is included

This analysis begins with generated movie dashboards on SouthSide Studio's public site. A dashboard can contain client delivery from our collaborations or our network's own film coverage. A post mentioning a film is not automatically proof that its producers worked with us. The case studies label the two kinds of work separately.

Posts can belong to more than one delivered campaign. The dashboard total answers what each campaign's list contains; a research sample needs each shortcode once. Talent PR and brand work are excluded here, while the homepage's account-archive totals cover a broader set of posts.

Coverage of the public movie-dashboard sample

Download the coverage table as CSV.

99.9% of the unique owned posts have a numeric views record. The 4 other owned posts are unmeasured for this analysis, rather than treated as posts with zero views. Estimated-only partner posts stay outside all research distributions.

Three totals that answer different questions

• Account archive: the homepage's measured post statistics cover the stored account-insights records, including posts outside this movie-dashboard scope. They are not all film collaborations.
• Campaign dashboard: its delivery list can include a crossover post also claimed by another campaign, and explicitly marked partner-post estimates. Summing campaign dashboards can count a post more than once.
• This research sample: 6,465 unique movie-related posts with actual numeric views, with no Estimated values and no repeated shortcodes. Their accumulated recorded views sum to 82,75,94,682; views can include repeat viewing and are not unique people.

A film's public page shows its larger audience metric, either views or accounts reached. Reach summed across posts can also repeat the same account. Keep that page's metric label when quoting a campaign and do not combine views and reach into one audience figure.

Why an average can hide the shape of a campaign

In this deduplicated sample, the 65 highest-viewed posts (the top 1% of measured posts, rounded up) contribute 20.9% of recorded views. The pooled median is 63,615 views per post; the middle half falls between 14,223 and 1,25,444. This is a description of this archive, not a typical result promised to a buyer.

For a format comparison, use the post-format benchmark, which applies a common July 2024 publication cutoff. Our historical archive has different views availability by format, so this common cutoff avoids pooling incompatible older periods.

Dates, estimates and observation gaps

The dated measured posts were first published from 2020-10-04 to 2026-10-26. A dashboard's “Captured At” field is the post publication timestamp. Monthly charts group the current recorded total by that timestamp; they do not show views that occurred during a particular month.

The stored insights do not consistently record when each metric was observed. Publication time cannot fill that gap. These readings are cumulative snapshots, and the data guard can preserve an earlier higher reading if a later API response drops. A fair elapsed-time comparison needs fresh, dated readings at the same age after publication.

Partner-published collaboration posts that the API does not expose can appear on a campaign dashboard at that campaign's measured per-post average. They are explicitly labelled Estimated. Their delivery links remain useful evidence of publication, while their estimated metrics are excluded from this research.

What the public figures can establish

They establish linked publication records and the stored post-level metrics available to us. They do not separate paid and organic distribution, identify individual viewers, measure brand lift or attribute ticket sales. No causal claim about a creative choice follows from a higher post total alone.

When planning a campaign, agree on the delivered-post list, reporting date, chosen audience metric and treatment of missing or partner-post data. If the business goal is enquiries or ticket conversion, define the tracking and reporting for that goal separately. See film marketing in Bengaluru, Kannada film promotion and documented delivery examples.

Reproduction and corrections

The build joins the committed post-insights and account-archive snapshots to the generated dashboard CSVs. Instagram shortcodes identify posts, numeric zero values are retained, absent or malformed values are excluded, and percentiles use linear interpolation (R-7). The tables are regenerated from those inputs when the site builds. The fixed editorial review date documents the methodology review; the source snapshot commit date changes when committed evidence changes. Neither is a measurement observation date.

Corrections can be sent to contact@southsidestudio.in with the post or dashboard link and the figure being checked.