AutomateBook a call
KPI bot
Your most important numbers pulled together and sent to you on a schedule — no dashboard to log into.
What it is
In plain English
A KPI bot collects the numbers that tell you how your business is doing — yesterday's sales, new leads, cash in, stock running low — and sends them to you automatically on a schedule. A short, readable message lands on your phone every morning, or every Monday, with exactly the figures you care about. No logging into five different tools, no building a report by hand.
Most business owners only look at their numbers when something already feels wrong. A KPI bot flips that around: the numbers come to you, every day, so you spot the dip on Tuesday instead of at month-end. It can also watch for trouble — if stock drops below a level or sales fall off a cliff, it pings you straight away.
When you need it
Is this you?
You need this when your numbers live in several different tools and checking them is a chore — so you only really look when something has already gone wrong.
- Your data is scattered across your point-of-sale, your accounting tool, your website and a spreadsheet.
- You only check the figures when something feels off — never as a daily habit.
- Someone spends time each week copying numbers into a report by hand.
- You'd run things differently if you saw yesterday's performance over your morning coffee.
How we build it
What actually happens
- 1
Agree the numbers that matter
We pick the handful of metrics that actually drive your decisions — not a wall of charts, just the figures you'd want on a sticky note each morning.
- 2
Connect the sources
We securely link to wherever the data lives — your sales system, accounting tool, website analytics, or a spreadsheet — and pull the numbers automatically.
- 3
Build the scheduled job
A small program runs on a timer — every morning, or every Monday — gathers the latest figures and does the maths, with no one having to press a button.
- 4
Format a message you'll actually read
We shape the numbers into a short, clear message — yesterday versus the day before, this week versus last — delivered to WhatsApp, email or Slack.
- 5
Add alerts where they help
Beyond the regular update, we can set thresholds — low stock, a sales drop, a cash milestone — so the bot pings you the moment something needs attention.
Example
How it plays out in practice
The owner of a small chain of three coffee shops in Pretoria runs on gut feel during the week, then gets surprised by the figures at month-end. Sales sit in the point-of-sale, stock in a spreadsheet, and card settlements in the bank — three places she never checks until something feels off.
- 1
The bot wakes at 7am
Before the first shop opens, a scheduled job runs automatically — no one has to remember to pull a report.
- 2
It gathers yesterday's numbers
It pulls total sales per shop from the point-of-sale, checks the stock sheet, and notes which lines are running low — across all three branches at once.
- 3
It does the comparison
It works out yesterday versus the same day last week, and flags that the Menlyn branch is down 18% while Brooklyn is up.
- 4
A clean message lands on WhatsApp
“Yesterday: R24 180 across 3 shops · Menlyn down 18% · Brooklyn up 12% · Low: oat milk (Menlyn), large cups (Brooklyn).”
- 5
She acts the same morning
Over her first coffee she moves stock between branches and phones the Menlyn manager — instead of discovering the dip three weeks later in the accountant’s report.
The owner runs the business on a daily pulse instead of a month-end autopsy. Problems get caught while they’re a Tuesday, not a month — and nobody spends an hour a week building a report by hand.
What we'd need from you
To get started
- The handful of numbers you most want to see regularly.
- Where that data lives, and permission to connect to it.
- How and when you want it delivered — WhatsApp, email or Slack; daily or weekly.
- Any thresholds worth an instant alert (low stock, a big drop, a target hit).
Timeline
How long it takes
A KPI bot pulling from one or two systems is usually live in 1–2 weeks. More sources and custom alerts push it toward 3–4.
What can affect that:
- Each extra data source adds a connection to build and test.
- A tool with no clean way to export its data (no API) takes longer to wire in.
- Custom alert logic — "tell me only if X and not Y" — adds a little build time.
What affects the price
Why it costs what it costs
Every build is scoped to your situation. These are the things that move the number — so there are no surprises on the call.
- Number of data sources
- Pulling from one system is quick. Stitching together sales, accounting, web and stock into one message is more to build and maintain.
- How reachable the data is
- Modern tools with a clean export are easy. Older systems, or ones with no proper connection, take more work to read.
- Complexity of the metrics
- Counting orders is simple. Margins, cohorts or trends compared week-on-week take more logic to get right.
- Delivery channels
- Email is the simplest. WhatsApp and Slack are a little more to set up and connect.
- Alerting
- A plain scheduled summary is straightforward. Smart, conditional alerts that watch for specific situations add build time.
Ready to build something that works?
Tell us what you need. We'll tell you honestly if and how we can help.