Skip to content
Syntasa

Customer story · Media & Streaming

500× faster clickstream processing, on NOW's own Google Cloud.

NOW, formerly NOW TV, used Syntasa inside its own Google Cloud to turn complex Adobe clickstream into data every analyst can query, cutting a typical query from 1,700 seconds to 3.4.

Talk with our teamDownload the case study

Results

Queries in seconds, not half an hour.

  • Clickstream processing500×faster than the previous SAS and Hadoop environment
  • Query time3.4 secfor a query that used to take 1,700 seconds
  • Analytics capacity48%of the analytics team's capacity recaptured, by NOW's estimate
  • New dashboards3 monthsto build more product dashboards than in the previous two years

Context

What gets watched, on which device.

NOW, known as NOW TV until 2021, offers video on demand and broadband in the UK and Ireland.

As its streaming services grew, the product team wanted to know which content customers watch, on which devices, and whether they click on the content recommended to them.

The answers were in the Adobe Analytics clickstream from the website, apps and devices. The team wanted that data ready for every analyst, in a platform that could scale as NOW moved into new markets.

The constraint

Split data, a slow platform and 2,000-line queries.

  1. One Report Suite per platformAdobe Analytics captured each platform and device separately, so one customer's behavior across devices was hard to combine.
  2. A slow legacy platformLarge volumes of clickstream sat in a SAS and Hadoop environment, and analysts waited for data to be loaded and reshaped before they could use it.
  3. Raw data analysts couldn't readSimple questions, like how many customers added something to their watchlist and then watched it, took SQL up to 2,000 lines long and up to half an hour to run.
  4. The data had to stay in placeAny new platform had to run natively in NOW's existing Google Cloud, not copy the data into another tool.

Deployment

From Report Suites to answers in seconds.

Syntasa replaced custom SQL scripts and manually run jobs with one automated pipeline, running inside NOW's Google Cloud.

  1. Step 1: Ingest every Report Suite

    The Adobe Analytics app configures itself to parse NOW's custom fields (eVars), takes in the newest behavior continuously, and checks it against Adobe's Reporting APIs.

  2. Step 2: Stitch each customer together

    Clickstream from every Report Suite is enriched with custom variables and lookups, and a device graph joins each customer's sessions across devices.

  3. Step 3: Open it to every analyst

    A unified customer schema in BigQuery hides the complicated clickstream logic, so analysts run their own queries and build Tableau dashboards without specialist knowledge.

“The fact that we can deploy Syntasa on our pre-established environments, and exploit the outputs there, means that it adds no extra analytical effort.”

Tom Slade, Insight Manager, NOW TV

Outcome

What changed for the business.

  • Queries in seconds

    A query that took 1,700 seconds on SAS and Hadoop now runs in 3.4 seconds, so the team experiments and iterates over 99% faster.

  • Capacity back for analysis

    NOW estimates the efficiencies recaptured 48% of its analytics capacity, time now spent investigating more interesting questions.

  • Dashboards in months, not years

    With all its metrics in consolidated Tableau dashboards, the team built more product dashboards in three months than in the previous two years.

  • Ready for new markets

    The automated pipeline scales with NOW's data as the service grows, without adding analytical effort.

“The decision to invest in Syntasa was quite easy as it was loaded with benefits that we could leverage just straight out of the box, without the need to sacrifice or replicate any technology investment we had already made.”

Kevin Mithani, Senior Product Owner, Digital, NOW TV

FAQ

Questions teams ask first.

Resources

Hear it from the NOW team.

Case study · PDF · 4 pages

Read the full NOW story.

The full story, with the architecture, how it works, a timeline and the detailed results.

Download the case study

Related.

Your constraint won't be our first.

Tell us what you are working under. We will bring the closest story and the engineers who delivered it.

Talk with our teamDownload the case studyA working session with our engineers, on your data and constraints.