At AggKnowledge, our north star remains to improve the expert experience and knowledge network efficiency through effective use of data. As we’ve been through the evaluation process with over 50 third-party data providers, we have some findings and learnings we'd like to share:
🔍 Evaluation Process - It’s an exciting time to be in data, with new datasets coming online every day. Even with LLMs, the evaluation is still mostly a manual process. But, it’s difficult to dedicate time to this if it isn’t your full-time job. Knowing what to look for is key. It's not just about data quality as presented. It's important to understand collection methodologies, fill rates, update cadence, and how the data will actually work in practice.
🏗 Many datasets aren’t built for this case use - Most of the datasets were built for sales teams to leverage for leads. Hardly any of the datasets we’ve evaluated were even thinking about a people application such as AggKnowledge. This requires more than just a bit of transformation of the data.
⚠ Filtering/Noise - Not all data is going to be great–in fact, that’s why we have a business! Even the best datasets out there will be incomplete or contain noise. We've now processed over 45K tags to remove "low-value tags" from our clients' workflows. AggKnowledge plugs into our clients upstream and chooses the best dataset for each individual use, while simultaneously removing the low-value tags.
💰 Can be expensive without economies of scale - The costs of integrating these datasets is high, usually requiring engineers and data scientists, and sometimes requiring data to be reshaped to fit existing systems.
💾 Not all data is available for purchase - Technographics exist, but only for technology solutions. What if you want to know if someone is knowledgeable about a vendor in oil & gas? We’re developing a new kind of dataset for vendor matching called Vendorgraphics™. Stay tuned for more in the coming weeks.
If you’re an expert network, executive recruiter, panel manager, or even private equity firm (that has advisors!) and we’re not already in contact, please reach out! We’d welcome the opportunity to share how we would approach data management and enrichment for your knowledge community.
#expertnetwork #executivesearch #surveys #marketresearch #data