Standard Innovation

A few weeks ago I wrote about trust in data, which is something I keep hearing about from people we work with. A simple example: ask for “last 12 months revenue” and you can get three different numbers. Finance counts to the end of last month, the CEO counts to today, and the board pack counts to the latest statement date. Each answer is reasonable, but they don’t match. Add AI tools that each pick their own approach, and the gap grows. So we recorded a 20-minute webinar on how we fix it with clients. Tom Mullen and Sandy (Alasdair) Wilson cover the two-part approach. Key takeaways: 1. Agree what your numbers mean. Most inconsistencies come from definitions, not technology. A business glossary sets one meaning, one calculation, one source and one owner for each core metric. 2. Make every tool use those definitions. Map the glossary to your data, turn it into fixed queries and point your AI tools at them. Don’t let each tool work the number out its own way. 3. Avoid the common traps. Don’t ask an LLM to write SQL on the fly, and don’t let your AI tools work around the queries you’ve already defined. 4. You can start this week. List your top five metrics, write down how each is calculated, get everyone to agree, then put those queries in one place and test your AI tools against them. None of this is new. But the businesses doing the hard yards on foundations are the ones getting real value from AI. Watch the recording here 👇 https://lnkd.in/eApkBFeT

Trust in data is a key topic I’m seeing at the moment. If we wind back 12 months, most businesses were struggling to move AI initiatives from proof of concept into production. In reality, not much has changed. Today, the tools are more accessible and powerful than ever, and proof of concepts are much easier to spin up. But, adding guardrails can only get you so far. When two different numbers land on a board report, trust is eroded quickly. Those doing the hard yards to put the infrastructure in place to ensure that questions get consistent answers, are the ones that are leveraging the tools to their advantage. If you want to get more out of your data, but don't know where to start, we've put the framework we use into a workbook. It explains how to choose a first use case, scope it, and lay foundations the next one can stand on. Link below. https://lnkd.in/eX-TzcQs

AI tools can be transformative for music rights investors looking to evaluate more catalogs and scale with less manual work. BUT... without laying the correct data foundations, you might find that the same questions start producing very different answers. And that becomes a big problem when you're making investment decisions or putting numbers in front of the board. So how do you make sure the data is trustworthy??? In our new on-demand webinar, our CEO Tom Mullen and Lead Data Engineer Sandy (Alasdair) Wilson run through the two-part approach we use to get confidence in numbers and get the most out of AI. They cover: - Why the same question can produce different answers - How to get your team and AI tools speaking the same language - Three steps you can take this week Watch it here ---> https://lnkd.in/eUHkmdgq Standard Innovation

The choices your fund makes around ingesting data can be more strategic than you think. In our latest Catalog Value Chain episode, our CEO Tom Mullen talks through the options available and how to find the balance between: ⏰ The job you need to get done today 🚀 The data advantage you want to build over time Watch it here 👉🏻 https://lnkd.in/ghVvZ8ix

In any data project, once you are through the initial evaluation/proof-of-concept phase, you need to put a flag 🚩 in the ground and lay the foundations for future initiatives. The foundational work usually includes setting up the core technologies for your project (ideally this one and others), agreeing on how the systems fit together along with the various layers of access and control that you need to keep it safe. These foundational steps needn't be overbearing, but they are important to get right and they need to be seen as an investment and an accelerant for future use cases. The foundations do require effort and in themselves often don't return direct value to the business, but they do create latent value to be extracted at all future stages. Make sure you take your time to get them right for now and for the future, but don't spend too long on trying model the unknown or over engineer things on day 1. You're aiming for directionally correct, a good balance between planning for the future and getting started today.

Getting data from different sources into one place is a foundational step for any music rights investor. 🏗️ In the second part of our Catalog Value Chain series, we’re covering data ingestion and how to choose the right tools for your business. Some tools will get you up and running quickly 🏃♂️➡️. Others give you more flexibility, ownership and control 🦾 . It’s worth thinking about how data could give your business an advantage over time, as well as the job you need to get done today. 💡 We cover: - How ingestion supports diligence, collections and reporting - ETL vs ELT and the different types of ingestion tools (spoiler alert, no sandwiches involved) - Choosing an approach that fits your objectives, team and budget Watch the full episode here: https://lnkd.in/ejjiSSUf Subscribe for a full breakdown of how data and infrastructure can support the music investment journey: https://lnkd.in/ghVvZ8ix