Most analysts I know make these tables and dashboards to get a job (me too), but most of the time it's low-key theatrics to score an interview. Actually useful data products usually require some specificity in addressing a use case, and they can be quite diverse. If you argue to make dashboards for the common lending user, then the UI from the protocols is adequate, tbh. If you manage to find a small use case to help the average user, then as mentioned, it's a small niche. Very little eyeballs, or maybe it's not compelling enough as a job portfolio. A good example is like AAVE revenue dashboards. If you are yield farming, then who cares? The people who do are probably financial modelers (who then need Excel to make their models, but we already know CSV exports are PITA and expensive. Again, another deterrent). Then there are the people of AAVE themselves, which probably already have done it and/or used it internally. Yet another deterrent to do duplicative work. It's like saying, "Who are you as a freelancer that can possibly have the same knowledge, problem definition, and resources to address Facebook's data problem?" Chances are you can never really tackle it because you are not in it. This makes the average solo data analyst's data product be stuck in an odd place where it's neither consumed by normies nor bigger protocols. The only alpha here, as I've realized, is that you need to truly dig hard at what some problems the protocols face are and see yourself as competing against their internal data analysts, working on problems that they know but don't have the bandwidth to address. But the chance of this is low, and it's hard because you probably don't know their idea maze well enough. Top of my mind, the only protocol where you can flex your analytic skill is Pendle. TN has been remarkably transparent in sharing his fears and concerns, so you can have a data project that addresses this :3. This, coupled by most dashboard makers using off-chain data (APIs) and closed queries, means the velocity of work getting passed around drops too. Maybe I am wrong and am ranting, but this is just the state of the space now. I think people are jaded in general, and protocols are internalizing the data for themselves because the space is maturing.

Not even gonna touch costs. Out of scope Right now for Morpho, which I am being slow now😂, I am thinking of making tables rather than dashboards. I genuinely believe, with easier data models, people can build on it and proliferate knowledge. Just like the good old times