Guide
Your dashboard has four kinds of chart, and they answer four different questions. Reading them in the wrong order is how people reach confident conclusions from thin evidence. Here is what each one is built from.
Four figures, shown for today and for the last 7 days:
That last one is the most overlooked figure on the dashboard. 200 messages across 3 conversations and 200 messages across 40 conversations describe completely different days. Volume alone cannot tell them apart; this number can.
Alongside these sit two highlights — the busiest hour and the top contact, both computed over 7 days rather than today. That window is deliberate: a single unusual evening should not be able to redefine what "normal" looks like.
Every hour of the day gets a bar, split into sent and received, aggregated across the selected window (7 days by default). Hours with no activity are shown as zero rather than skipped, so the silent parts of the day stay visible — the gaps carry as much information as the peaks.
A tall bar made mostly of sent messages and a tall bar made mostly of received messages mean opposite things. The first is someone initiating; the second is someone being contacted. Same height, different story.
One point per day, sent and received tracked separately, over 30 days by default.
Use this chart for one thing only: direction of travel. Day-to-day noise in messaging is enormous — weekends, holidays, travel, a single busy group chat. Nothing meaningful can be read from one day against the next. A four-week slope is worth something; a two-day dip is worth nothing at all.
Conversations ranked over the last 30 days, showing for each: name or number, whether it is a group, total messages, the sent and received split, and when the last message occurred.
Default sorting is by total volume, and that includes groups. A busy family or work group will usually sit at the top and tell you very little. Re-sort by sent to see where someone is actually spending effort — that ranking is far more informative and looks quite different on most accounts.
Each conversation also breaks down into six four-hour blocks, with sent and received counted separately in each. This answers a narrower question than the global hourly chart: not "when is this number active" but "when does it talk to this particular person".
It is the most suggestive view on the dashboard, and therefore the one to be most careful with. A conversation concentrated in one block is a routine — two colleagues who talk during work hours, two friends who catch up after dinner. Routines are extremely common and almost always boring. Treat a striking pattern as a question to ask, never as an answer you already have.
Different charts default to different time spans — 7 days for the hourly view, 30 days for the daily trend and contacts. When two charts seem to disagree, check the window before concluding anything; most apparent contradictions are simply different periods.
Sessions, the 30-second gap, and why the number is deliberately conservative.