> For the complete documentation index, see [llms.txt](https://knowledgebase.fabricdata.com/fabric-knowledge-base/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://knowledgebase.fabricdata.com/fabric-knowledge-base/resources/use-cases-and-best-practice/roi-use-case.md). # ROI Use Case ## Introduction Performance, efficiency and automation in the broadcasting industry depends heavily upon having clean, consistent and dependable data. On the consumer side, it’s what customers see and research before buying a product, and what impacts their experience on a D2C application or another channel. Product data enables, and makes or breaks, a purchase decision. As broadcasters grow and evolve over time, business-critical data becomes fragmented. There is no big picture because data is scattered across applications, including legacy on-premise applications. As all this change occurs, mission-critical data becomes inconsistent, and no one knows which application has the most up-to-date information. Poor data quality has myriad knock on effects and consequences. Some of these can be quantified, for example by the financial fines and penalties a broadcaster might be subject to if they do not comply with regulations, or by the time and cost expended on a team manually correcting or completing data. However, other costs are more complex to measure. According to Harvard Business Review, a task performed with erroneous data incurs a cost 100 times greater than if the data had been correct. This study is based on the 1-10-100 rule, developed by George Labovitz and Yu Sang Chang in 1992: applied to the problem of data quality, this rule follows a relentless logic illustrating the importance of maintaining a high level of data quality, on a continuous basis and not occasionally. It costs the company £1 to check the quality of a piece of data – this is the cost of prevention. Cleaning up the data (e.g. removing a duplicate) costs £10 – that is the cost of correction. Unverified data costs £100 – this is the cost of failure. High quality data is essential for efficient operations and underpins your organisation’s strategic outcomes. ## Types of Bad Data There is a wide variety of common data problems encountered when handling entertainment catalogues that have been managed over a large period of time: ### **Structural Hierarchies** that are incorrect or no longer fit for purpose - Over the last century, the requirements for managing structural programme metadata have changed dramatically. The hierarchy levels required for content to efficiently flow through internal supply chain systems and out across the global distribution landscape are now clearly established, and typically need to be retrospectively applied to legacy catalogues. Maintaining outdated content hierarchies requires substantially more ongoing mapping and transformation, slows down deliveries, adds significant engineering complexity layers for all future work and makes visibility for sales and licensing more opaque. ### **Duplicated Data** Whether duplicate records, attributes, or data points that are duplicated throughout supply chain systems without lineage of provenance, this creates inaccurate results, leads to unnecessary data storage costs, makes it more difficult to keep data up-to-date and consistent and leads to costly mistakes when using the data for analysis. ### **Incorrect, Malformed or Decayed Data** There are hundreds of data issue types that typically occur when catalogues have been maintained over long periods of time without strict governance, policies and control processes. These types of issues infect catalogues from the source, and permeate to every connected supply chain system. This adds viscosity to supply chains, and layers compounding cost and efficiency consequences to every connected system and processes. A summary of some common examples include: * Incorrect Sequencing (Inaccurate Season, Episode, Storyline, Running Order and Airing Order Numbering) - leads to the wrong programmes being sold, delivered and presented to consumers. * Inconsistent Character Lengths - leads to guillotined copy and degraded user experience. * Misaligned Modular Episode and Storyline Management - leads to the wrong descriptive information being available for sales, licensing and consumer presentation. * Suffixed Articles - leads to content being misindexed, misrepresented and hard or impossible to find for sales & licensing and consumer discovery. * Roman Numeral Handling & Transformation (Eg: Transformers III) - leads to duplicate records and inaccurate data points. * Cross Title Duplication - episodes stored as specials, special stored as standalone titles, etc. leads to the wrong records being available for sales, licensing and consumer presentation. * Synopsis Resumes - inclusion of actors previous credits within a synopsis leads to guillotined copy and degraded user experience on VOD & D2C applications. * Inconsistent Tone - a consistent tone across all of your editorial metadata helps to create a unified and recognisable brand identity that viewers will respond to. An integrated and cohesive tone also helps to create a more intuitive user experience, which will increase user engagement and lead to better business outcomes. This makes it easier for viewers to identify what content they’re interested in and cultivate relationships with the brand that go beyond just viewing content. A consistent tone can also help establish a trustworthy relationship between the viewer and the platform, increasing overall customer loyalty. * Unlicensed IP - with legacy catalogues it is very common that synopses, imagery and other data points that do not fall under public domain have been added to records over time. This leads to legal risk of copyright infringement and fines from regulatory bodies for unauthorised presentation of IP that is owned by other businesses. * Culturally Inappropriate Data - having outdated, culturally inappropriate data for a broadcaster can be a major risk to an organisation’s reputation. This data can lead to negative press and public backlash if the broadcast is perceived as insensitive or offensive. This could lead to scrutiny from regulatory bodies, loss of revenue, and decreased trust from viewers and customers. Additionally, using outdated or inappropriate data can lead to loss of business as potential customers may feel uncomfortable engaging with a brand seen as insensitive. * Missing or Incorrect Data - leads to significant risk of loss of viewership and revenue from customers who cannot find content, forcing them to switch to other services with better user experience. Leads to lack of visibility internally and uninformed decision making and introduces privacy and security risks. ## The Cost of Bad Data The overall cost of bad data can be broken into four main categories: macro-strategic (financial implications and reputational risk), operational efficiency (internal processes, productivity and missed opportunities), employee & team satisfaction (how easy and enjoyable is it for them to do their work), and the resulting impact to the end consumer. All data issues that make their way to the end user, ultimately end up being a material contributing factor to churn.
Further, the cost of bad data escalates depending on who discovers the issue. If bad data is caught by a data engineer, they are a hero. No harm, no foul. If bad data is caught by the general public, there could be reputational or legal repercussions depending on the situation.

The impact of bad data

It’s not just the remediation that is expensive either. As companies increasingly rely on complex data assets to help execute business decisions, the opportunity costs of bad data rise as well. As data stacks become more complex, there are also more data “handoffs” introducing the opportunity for more issues. Many industry leaders and bodies have created comprehensive analysis on the impact of bad data to media organisations. IBM and Gartner estimate this to be between 12-20% of total revenue. With the BBC revenue in 2022 at £5.33 billion, the true, combined direct and indirect cost of bad data to the business is likely to be around £1 billion annually. ## The ROI of Fixing Data High quality and trustworthy data can improve efficiency, help mitigate risk and reduce costs. Understanding the importance of data quality and having procedures to tackle the root cause of problems will ensure your business is able to use data to make effective decisions. High quality, reliable data will enable you to have confidence in your decisions, spend more time using data to drive valuable insights, and less time trying to figure out if the data is fit for purpose, or what the consequences of poor data quality are on their results. The cost of getting data quality right the first time is cheaper than the costs of having poor quality data and having to fix it further down the track. Producing high quality data requires planning and commitment. Developing a data quality plan with achievable and measurable goals that everyone can commit to can be a good place to start. Preventative measures and effective data quality management should be embedded into your organisation. It is a continual process, and everyone should be aware of the risks and have procedures in place to account for data quality. Ensuring data quality throughout all stages of the data lifecycle and being able to identify data problems proactively, and as they occur, is more beneficial and cost effective than retrospectively trying to fix poor quality data. Businesses must invest in systems and processes to ensure that their data is of high quality and is used effectively. Having good quality data management and governance processes in place can go a long way towards driving business success. ## Fabric ROI Examples Whilst the overall financial benefits of implementing a system of good data management and governance can be estimated at around 12-20% of overall revenue, at a lower level, there are a number more tangible, accurately measurable savings that Fabric has calculated with it’s customers, to highlight time and cost savings directly attributed to our software & services: * A Hollywood Studio customer saves £291,000 annually, through the automation of third party ID matching, data retrieval, normalisation and de-duplication processes. * The same Hollywood Studio customer saves £309,000 annually, through using Fabric Studio’s automation tools to generate asset & image placeholders in connected supply chain systems, and assign primary metadata attributes, compared to manual legacy tasks. * A global broadcast customer, based in the UK, saves £1.83 million annually, through the implementation of an enrichment sequence that a) Matches or Registers content with EIDR, b) Enriches from IMDB, and c) generates draft translations from Google Translate. This runs across their catalogue continually at scale, and provides savings compared with multiple legacy manual processes. * A global broadcast customer, based in the UK, saves £2.59 million annually, through the implementation of a process automation sequence that a) automatically generates asset placeholders and IDs for Master assets, b) automatically generates local language placeholders and c) automatically generates regional compliance placeholders.This runs across their catalogue and supply chain continually at scale, and provides savings compared with multiple legacy manual processes. * A South American public service broadcaster has estimated that using Fabric saves approximately £317,000 annually across their Title Registry, Curation, Enrichment and Translation tasks. * The same South American public service broadcaster also estimates that Fabric provides direct savings of approximately £495,000 annually through the use of Fabric’s translation integrations, synopsis paraphrasing tools, reporting visibility and overall catalogue preparation tasks. ## Reference #### The Cost of bad Entertainment Data Variety reported that in 2019 (pre-pandemic), the media and entertainment industry was said to have made $101 billion across theatrical and home entertainment. Fifteen percent of that is approximately $15.2 billion. Bad data causes this much damage to your bottom line? It’s not hard to believe. Everyone has bad data lurking somewhere in their organization. #### Incorrect program Data 'Reports of sharks falling from the sky are on the rise again, and no one on the Eastern Seaboard is safe.' So went the Netflix description for a listing of Stewart Lee's Comedy Vehicle, Lee's Bafta-winning BBC TV series from 2016. Rather than being corrected, the description stayed like that for two years. It is this small-but-significant mistake that puts Lee on a journey of self-discovery. #### Retention vs Acquisition The first rule of any business is to retain customers and build a loyal relationship with them, and thereby avoid customer acquisition costs. #### The Risk of Bad Data Everyone’s heard the old adage “time is money” and, in the case of metadata, it’s true. When the real money is found not in the end-product but the right to monetize it, knowing what you can and can’t deliver to the masses is key. #### Metadata Transformation Six steps to guide your business through a metadata transformation process to deliver world class entertainment experiences and become a future-proofed business. ***
Related Articles * [Case Studies](/fabric-knowledge-base/resources/case-studies.md) * [What is Fabric Studio?](/fabric-knowledge-base/studio-basics/what-is-fabric-studio.md) * [Cloud Localization Blueprint](/fabric-knowledge-base/resources/case-studies/cloud-localization-blueprint.md)