← All work Case study · Agency
Seat 1 · Agency · Merkle Sokrati (dentsu), for Fox Star Studios

Measuring what film marketing couldn’t see

Business Analyst, Merkle SokratiMar 2018 – Feb 2019 · 7–8 film campaigns3 min read
93%reach above plan
11.7xdigital-attributed ticket bookings vs the studio’s benchmark
#1most overperforming video from India, per Meta
02 · The situation

A film gets one weekend, and nobody could see what sold tickets

ClientFox Star Studios, 21st Century Fox’s Indian film studio (now Disney)
My roleRan the Meta and Google campaigns; built the attribution logic
Worked withFox Star’s data science team
StakesUp to $86K of spend in four hours before release

Film is a binary business. Opening weekend largely decides a film’s fate, so there is no second chance to fix a campaign. Yet in 2018 no off-the-shelf tool could tell the studio which digital activity actually sold tickets. Platforms reported views and clicks; ticket sales happened somewhere else entirely.

The missing link was the join between online and offline data. A viewer might watch a trailer on YouTube, see a Meta ad two days later and buy a ticket through a booking partner. Each system saw one fragment, so every platform claimed credit and none of the claims could be tested.

03 · The call

Release is weeks away and platforms only report views and clicks. How do you decide where the money goes?

Make your call to see mine.

04 · What I built

Joining what people saw to what they bought

Step 1 · Before

Every system saw one fragment

A trailer on YouTube, a Meta ad two days later, a ticket through a booking partner. Each claimed credit; none could be tested.

Step 2 · Matching

Online joined to offline

Ad exposure was matched to ticket transactions using transaction codes and cinema check-ins: the journey from first impression to seat.

Step 3 · Attribution

Custom weighting, tested windows

My main contribution, with the data science team: custom weighting rules and tests of 1, 7 and 28-day windows to see how film audiences actually decided.

Step 4 · Live reallocation

Budget moved during release week

I used the outputs to shift budget between channels and creatives in real time. For Ek Ladki Ko Dekha Toh Aisa Laga, reach came in 93% above plan.

Exhibit · step 1 of 4
Three disconnected touchpoints (trailer, ad, ticket) joined into one journey, attribution windows of 1, 7 and 28 days, and reach 93% above plan YouTube trailersees views Meta adsees clicks Ticket bookingsees sales each system sees one fragment and claims the credit matched by transaction codes and cinema check-ins attribution windows tested 1 day 7 days 28 days reach: plan vs actual plan +93%
Now showingEvery system saw one fragment
05 · What happened

The campaigns beat plan, and Meta noticed

93%higher reach than plan for Ek Ladki Ko Dekha Toh Aisa Laga
11.7xticket bookings attributed to digital, vs the studio’s benchmark
#1most overperforming video from India, per Meta
  • The plan was built from historical data, consumer trends and benchmarks, so the 93% was measured against a realistic target.
  • Meta featured the work at its Partnership Forum in January 2019.
  • The account was among the first in India whitelisted for Meta’s ThruPlay optimization.
  • exchange4media featured Sokrati’s work for Fox Star in “This is how Fox Star Studios is upping its digital game.”
06 · What I'd change

I’d pair attribution with lift studies

Attribution tells you who gets credit; a lift study tells you what actually caused the sale. Together they would have made the model much harder to argue with.

When outcomes are binary, measurement is not reporting. It is how you decide.

Measurement and attribution · Entertainment · India · Meta, Google, YouTube

Hiring for a role like this?astitvasawhney@gmail.comMore ways to reach me