Ask HN: What weird or hard problems are you trying to solve?

You know, the weird stuff ¯\_(ツ)_/¯

Not sure if this is weird so much as an area of business that not a lot of people think actively about, but get bit by.

THE PROBLEM

How do you automate and aggregate context across business departments for various forms of activity, and then map that to marketing analytics in a way that gives relevant and sufficient insights beyond just channel or user data? How do you more fully answer the question of "what happened when [$thing happened]?"

THE VALUE OPPORTUNITY

Countless people hours and marketing dollars are wasted going down fruitless rabbit holes looking for what caused some change, or thinking they found the cause in a change in performance and pursuing that when it reality it was something else. In many of these cases, this could have been easily avoided if only there were sufficient data on the business activities (internal and external) logged and aggregated with marketing data in a way that was then automatically surfaced in an appropriate manner. As the scale of the company increases, so does the impact of this.

WHY IT IS WEIRD/HARD

It's weird in the sense that only a small subset of people are immersed in analytics enough are aware they should care about it, and probably fewer geek out enough about marketing analytics and process to care about trying to solve it. It is hard because it is just as much a people challenge as a technical one. The technical side is somewhat straightforward in terms of aggregating as many data inputs as you can--it's basically a ton of data plumbing and monitoring for changes with that. Whether that's bid management platforms and DSPs or SSPs, email platforms, site analytics, etc. But then also project management tools and properly categorizing the meta data for relevant updates to be surfaced. You have challenges around walled data gardens and comparing apples to oranges around things like attribution measurement, but that is something that can be handled. Surfacing it in timely and sufficiently useful ways is an interesting design and UX challenge though, from annotations and "pull" data, to modals and callouts that are more "push" in how they inform people of context before it bites them.

The people side however, is constantly in flux in a way that the data side is not. Some aspects of this absolutely rely on consistent adherence to process to capture key data that is hard to slurp up through an API. Some of it is quite ephemeral. I've encountered team situations where people object (or struggle to due to limited training) to filling out a couple fields in a Google Sheet, or need to be hounded to fill out a given form, etc. Some companies can enforce this to levels others cannot. Things also get really interesting at large companies (think FAANG). You're dealing with many teams, many overlapping or conflicting processes such a solution would need to be embedded into, localization, internal/external vendors of varying levels of visibility needs, and also personalities who may want more control over their orgs' processes and need persuading.

At the end, this all needs to be balanced against how much utility you get out of the insights because it is easy to over-index on investing in building this tech and process out only to not get insights out of it. Unfortunately you often only learn that after the fact when you've been bitten by it.

If there's any companies trying to solve for this, please do reach out (see profile). I love chatting about it and want to help build the tools and processes that solve for this at scale and have ~15yrs experience in the space, a good chunk of which have been spent trying to solve for variations of this.

I've experienced this problem in small companies/side projects I've started. This a great perspective and great take on the problem. I'd love to help out in development/anything you'd need help with, email is in my profile.

Keeping a working Wireguard configuration with 4 peers, for both VPN-through-house, and remote LAN access.

Given the earth's population trajectory and the reduction in fertile arable unpolluted lands, there is a coming crisis in the distribution of food. How do we distribute food to high density Asian urban populations efficiently, minimizing needless motor vehicle trips, packaging and spoilage, when convenience purchasing is on the rise and average household sizes are shrinking? Our answer is a network of robotic service locations with automated stock-keeping and a shared, wholly owned logistics network plus personalized direct from fresh ingredient preparation.