[alexpetralia]

* Design matters a lot - if it looks bad, people won't look at it.

* Layout for dashboards is almost completely formulaic. A panel for selected high-level stats (user growth % increase from last year, user % increase from last month, # new users added), a panel for breakdowns (user growth by marketing channel, user growth by registration cohort), a panel at the top for filters ("let's filter the entire dashboard by just this marketing channel, or just this registration cohort") identical to all breakdowns provided, and finally a row-level drill-down ("show me the users in this particular aggregation"). It took me a very long time to learn that this design is entirely cookie-cutter for good reason. Users always want the same things: stats, breakdowns, filters and drill-downs.

* Padding matters, font matters, color palette matters, no typos matter, visual hierarchy matters (i.e. big bold numbers versus smaller grey numbers).

* Always define the key metrics first (based on fact tables). All dimensions and drill-downs in the dashboard will derive from these front-and-center stats.

* Reconcile to existing metrics before broadcasting widely - almost always, people have the same stats in extant technologies (i.e. Excel, Mixpanel, Salesforce) and will instantly find inconsistencies between your figures and the extant ones.

* The vast majority of users will be passive viewers. Very few users will be "power" EDA (exploratory data analysis) users. EDA views will look different from the view that passive viewers want - keep them separate

* Obviously, the more things done in code, which promotes modularity and composability, the fewer data integrity issues you will have

[gpm]

> * Layout for dashboards is almost completely formulaic. A panel for selected high-level stats (user growth % increase from last year, user % increase from last month, # new users added), a panel for breakdowns (user growth by marketing channel, user growth by registration cohort), a panel at the top for filters ("let's filter the entire dashboard by just this marketing channel, or just this registration cohort") identical to all breakdowns provided, and finally a row-level drill-down ("show me the users in this particular aggregation"). It took me a very long time to learn that this design is entirely cookie-cutter for good reason. Users always want the same things: stats, breakdowns, filters and drill-downs.

Is there any chance you could link an image of what a good version of this looks like?

[alexpetralia]

Sure, I could almost choose any one on Google as they all follow the same template - here is one: https://coderthemes.com/hyper/saas/index.html

1. Six top-level stats jump out at you: customers, orders, revenue, growth %, current week revenue, previous week revenue. All of these stats are adorned with a few substats (smaller text), almost always a % up/down from last period

2. A few large panels with breakdowns: revenue over time, revenue vs projections, revenue by referral source, revenue by location

3. The top right has your filter buttons, and generally it includes every breakdown dimension on the page. For example, "let's look at this dashboard by just the Google referral source" or "let's look at our stats from the U.S. geography only" or "let's filter this for last 2 years only"

4. Drill-down is "top selling products." This isn't truly a drill-down, as it is still an aggregation, so you really want to drill-down to the record-level. If you filter the dashboard for "U.S. sales by the Google referral source for the last 2 years only", people invariably want to see what the actual row-by-row sales were, and that is the drill-down. They can easily export this and reconcile to source systems. As an example, for some of the work I do, sales reps don't just want aggregations about their sales leads, they want the actual names of actual sales leads (as rows) so they can contact them.

So again, four major parts to a dashboard, which really drive from two simpler (likely familiar to most data analysts): metrics and dimensions.

[gpm]

Thanks :)