Dashboard editor upgrades so making a Dashboard is intuitive (potentially satisfying)

[laupalombi]

Problem statement

We have a huge differential value in our dashboard creation engine - it is deep, huge, incredibly flexible and allows you to do absolutely anything you want. Few or none softwares can say that in the smarthome workd.

But showing all those capabilities at once makes it hard to pick the right ones, and the process long and complex even with a UI. The opportunity is to improve this UX and create more ways to have a dashboard created according to what you want to get from it.

Community signals

Not related to feature requests themselves, since its targeted to improve life of people outside of Github community. But we have seen people complain very publicly how complex building a custom dashboard is in HA. Also the desire to "slightly customize" default built in experiences, because truth is a lot of cases don't need extra complexity.

It's damaging us as a brand for everyone that the builder that is supposed to be one of the key selling points of the product is so hard to use for creating simple things.

In Reddit, users argue one of the difficult part to get into Home Asistant is to create useful dashboards:

Home-Assistant is easy enough to get into. What will take a while is doing the dashboards correctly.

Appetite

This doesn't need to go in one big haul - we can piece this into smaller bets if we needed capacity somewhere else.

Foreseen solution

Pending research outcomes to define specifics and priority,

1. Dashboard creation has an improvement in UX so for the basics you know where to go if you needed something.
2. There are more ways to create a dashboard if you don't want to spend a lot of time but still need something specific - add sections (groups of cards), like cards that can generate things based on strategy.
3. The existing superuser way of doing dashboards is still possible - none of these improvements should affect the configs in such a significant way that makes thorough customization impossible.
4. We will be able to use strategy sections for this.

Scope & Boundaries

In Scope

• Dashboard editor changes in UX
• Sections of cards

Out of scope

• Changes outside the dashboard creation engine
• Changes in cards - unless critical
• New strategy dashboards or editor panels - any new tool outside dashboard editor.

Risks & open questions

Making improvements is easy, the big risk is projecting a series of changes on personal experiences and not really targetting what is the most important thing (which we don't know). Therefore we are conducting usability tests with non-contextual users to understand the map of existing pain points and be able to target better the solutions in terms of effort.
-- Ongoing right now OpenHomeFoundation/user-research#19

However, this risk is for UX principles - the bet still stands because we know there is a need for more templates and sections and preset solutions that can be tackled right now.

[anakinsbrna]

I think we won't see, and maybe not even achieve that creating a dashboard will be satisfying. That's a very subjective experience, for some people that's probably already the case

Let's change it instead to "Improve the usability of creating custom dashboards" or "Creating a Dashboard is Easy / Intuitive / Straightfoward"

[laupalombi]

Done, did a mix - kept the original one but added your suggestion for a more targeted approach with intuitive

[jlpouffier]

I would like more clarity on the solutions you foresee to implement to solve the problem of making dashboard creation faster and less complicated.

For example:

I see strategy sections: Which ones and how do you plan to present them

I see templates for new dashboards, what are these? And how do you plan to present them? Are they dashbaord strategies like these ones?

Please note that the last time we created a strategy (The area dashboard) it took us about 6 months to create.
Please also note that the community is already providing custom strategies on HACS with no way to use them in a usable way. (See the first gif on that repo for the current UX)

I see more ways to create a dashboard, what are these?

I see Wizards, how do you plan to present them, how do they articulate with the existing card picker, and everything else you want to deliver?

Overall, I would very much appreciate more clarity on the solution space to understand the scope of the bet.

[anakinsbrna]

@jlpouffier Laura is currently conducting a moderated usability test, holding a video call with people who are not experienced in creating dashboards and asking them to perform various tasks related to the dashboard creator engine. This study will identify where in the dashboard-creation process people struggle and what we should fix. The study we designed is of really good quality and will definitely bring a lot of insights to what exactly the pain points are - based on those, we can create better solutions.

@laupalombi correct me if I am wrong, but here the solutions are on purpose not yet identified and described because we are about to identify them with this study. Describing solutions already risks getting attached to them and appearing biased in the analysis of the interview, and, in the end, still embedding solutions we envisioned as suitable before conducting the research, instead of finding fixes to actually solve the problems.

For bets like this, it would help to add more information on what stage the work is at. It could also help to present the research that is being conducted more prominently. Currently, it's under "Risks and Open Questions" at the very end, very shortly. Since the study is already running, we can present that with much more confidence. Maybe in "signals from the community"? This section could become broader for all kinds of references/proof, not only community signals