Features

Analytics & link tracking

Per-group click tracking, clicks by audience and campaign, best send times, and delivery failures.

Open Analytics

Two things are measured, and it's worth knowing which is which. Clicks are measured per group, and they're the honest signal for interest — someone chose to tap. Delivery and read receipts are measured per recipient, so once your number has been sending for a while you can also see how many people a send actually reached and how many opened it. Neither is a vanity number, and where a figure isn't measured the page shows nothing rather than a guess.

Analytics is open from the start, but stays empty until your first send with a link.

Any http(s):// URL in an outgoing message is rewritten into a short tracking link, separately for each group. Group A and group B get different short links pointing at the same destination.

That per-group split is the whole trick: when someone clicks, we know which group it came from, and therefore which category — which is how the app can tell you that alumni clicked and second-years didn't.

You don't configure anything. Paste the real URL and it gets rewritten on the way out. (On a self-hosted deployment with no short-link host configured, URLs go out untouched and no clicks are recorded — the hosted service has this set up.)

Recipients see a promotionbot.net/r/... link. On the Pro plan, or Team with the Custom Domain add-on, you can use your own domain instead, so links carry your name.

On Team and Pro, tracked links are reported in the Billing panel (Starter doesn't rewrite links), though the figure isn't enforced (Plans & billing).

What the page shows#

The Performance tab covers your WhatsApp broadcasts. Pick a range — last 7, 30, or 90 days, or lifetime — and you get:

Four headline numbers: messages sent (it reads Messages delivered, counted in people, only when delivery receipts exist for the period), link clicks, groups reached, failed sends. Groups reached is the number of groups that received at least one send in the period. The clicks tile is a count, not a percentage — deliberately, since there's no per-person reach figure to divide by.

Each of the four tiles can carry a small vs previous chip, for example "+12% vs previous 7 days": the same range shifted back (a rolling window, so 7 days is compared with the 7 days before it). It only appears when both periods have data for that tile. If the earlier period is empty, or the number isn't measured (no delivery receipts, say), the chip is left out rather than showing 0% or a guess. Lifetime has nothing to compare with, so it never shows one. For failed sends, a drop is the good direction.

Delivery & engagement over time — messages and clicks on one chart. The useful read is the gap between them: sends going up while clicks stay flat means you're increasing volume, not attention.

Clicks by audience — clicks per category, with each category's share of clicks and a volume cue for how many sends produced them. Rows are ordered by clicks. Read the volume cue too: a big share from one small group is not a strategy. With no category data yet, this section shows clicks by campaign instead.

Top campaigns — each campaign's sends, clicks, and share of clicks.

Delivery outcome — delivered versus failed.

Best send time — the day and hour that historically get you the most clicks, in your own timezone. It needs at least 5 clicks in a time window before it names one; below that it says Not enough data rather than crowning a time window on one or two taps. Use it, then re-check it next term; it moves.

Top groups — ranked by clicks, with each group's clicks and sends alongside. Click any row for a per-group drilldown.

The Performance tab exports to CSV.

Clicks per 100 sends — the Dashboard tile and the event recap show clicks ÷ group sends × 100. It is a density, not a percentage of people (one send reaches a whole group), and it reads "—" when nothing was sent.

The Links tab lists your off-WhatsApp tracked links — the ones on posters, bios and QR codes. Create them there; their clicks are counted separately from broadcast clicks. Its list and totals are all-time, so the period pills and the Performance Export button are hidden on this tab; it has its own Export CSV.

Each link also shows Unique visitors next to its clicks: one person tapping five times is five clicks but one visitor. It's approximate — a visitor is a daily-changing, one-way code made from the IP address and browser, so the same person on two days or two devices counts more than once, and several people behind the same network with the same browser count once. A dash means it isn't measured (clicks recorded before this was added); "12+" means some of the link's clicks predate it.

Event funnel#

For events, a funnel view showing how the timeline performed message by message — which message in the ladder actually drove the clicks. This is how you find out that your reminders do the work and your save-the-dates don't.

What isn't measured#

Being straight about the limits, because they affect how you read the numbers:

  • Receipts only exist going forward. Delivery and read receipts are recorded per recipient, but only for messages sent after this was switched on. Older sends show "Messages sent · no delivery receipts yet" — that blank is deliberate. An empty cell means not measured, never zero and never an estimate.
  • Read means opened, not read. A read receipt says the message was opened on someone's phone. It does not say they took it in, and a recipient with read receipts disabled never generates one — so treat the read rate as a floor, not a headcount.
  • Reach is people where receipts exist, groups where they don't. Delivery and read rates use a recipient (people) denominator drawn from receipts. Click figures are still per group. Don't mix the two in one sentence to a board — say which you're quoting.
  • No per-person data. A group broadcast has no per-recipient identity, so there's no way to know who clicked — only which group.
  • No RSVP or attendance tracking. Sign-up links redirect to whatever you point them at; what happens there is that tool's business.

Team visibility#

Non-admins see their own number's clicks. Send volume is org-wide (it isn't recorded per number), so a member sees the organisation's send totals alongside their own click data. Admins can toggle team-wide stats.

Send failures#

Sends fail — a group the account was removed from, a media upload WhatsApp rejects, a network blip. Failures are recorded with the reason and the group, rather than disappearing into a log.

Check this page after any large send. A cluster of failures on one number is worth investigating before your next send; see Troubleshooting.

Monitor#

Monitor is shelved. It watched incoming messages on your number for keywords and competitor activity, and it was unfinished, so it is hidden from the app while we rework it. There is no Monitor tab and no keyword or competitor configuration in the product today, on any plan. Nothing you configured was deleted.

Automatic opt-out (STOP) detection is currently off. PromotionBot does not read incoming messages, so it does not notice when someone replies STOP (or unsubscribe, afmelden, …), no admin email is sent, and the "Opt-out requests" card has nothing to show. Opt-outs are yours to handle: if someone asks to stop, open that group on Groups, switch it to Inactive, and save. An inactive group is never messaged by any send. This is the same on every plan.