Technology & AI for Business
You already have the data. The question is which decision it should change.
Sales in a billing app, enquiries in a chat, expenses in a ledger, footfall in a notebook. Most owners are sitting on useful information and are not sure what to do with it. This session is about starting with a decision, choosing a few numbers, checking they are right, and reading them without fooling yourself. No dashboards or special software are needed.
Explore this page6 sections
More numbers do not mean better decisions. The right few, read carefully, do.
Many owners decide by feel, and experience is valuable. But feel can miss slow changes: a product that sells less each month, a group of customers who never return, a cost creeping up. A few simple numbers, reviewed on a fixed day, can show what memory does not.
The danger runs the other way too. Data can be incomplete, defined differently in different places, or read as if it proved more than it does. Good practice is plain: be clear about the question, be suspicious of the data, compare like with like, and treat the number as a prompt for judgement and not a replacement for it.
What the session covers
35 topics across 7 areas. Seven parts, from the question to the routine.
Start With a DecisionData answers questions. Pick one.5
- Which product or service to push, drop or reprice
- Whether to hire, buy stock or extend hours
- Where customers come from and which sources are worth the money
- Why repeat business is falling or rising
- What would you do differently if the number came out high or low?
If no result would change your action, you do not need that number.
Choosing a Few NumbersFewer, and understood.5
- Sales, margin and costs, by product or service
- Enquiries, conversions and repeat customers
- Cash in hand, receivables and what is overdue
- Time taken, returns, complaints
- Defining each number in one sentence so everyone means the same thing
Where the Data LivesFind it before you trust it.5
Hypothetical example: a shop owner finds that sales appear in the billing app, but many cash sales were never entered. Any average from that app is wrong until the gap is fixed.
- Billing and accounting software, payment apps and bank statements
- Messaging apps, forms and enquiry sheets
- Notebooks, receipts and memory
- Getting data into one place with an export or a simple sheet
- Who is responsible for keeping each source up to date
Checking Data QualityGarbage in, confident garbage out.5
- Missing entries, duplicates and inconsistent spellings
- Totals that should match another source
- Changes in how something was recorded over time
- Outliers: an error, or a real event?
- Writing down what you assumed
Reading Numbers SensiblySimple comparisons beat clever ones.5
- Comparing the same period: this month with last month and with the same month last year
- Ratios and averages instead of raw totals
- Splitting by product, area or customer type to see what is hidden in the total
- Small samples and one-off events
- Correlation is not cause: two things moving together do not prove one drives the other
When a number surprises you, check the data before you change the business.
Showing and Sharing ItSo that people can use it.5
- One simple table or chart for each question
- A baseline and a target, not just a figure
- A short written note on what changed and what you will do
- Sharing with the team in a way that is fair, not punitive
- Not exposing customers' personal details in a report
A Review Routine, and Where AI FitsHabit matters more than tools.5
- A fixed weekly or monthly review, short enough to keep
- One decision made or one thing changed each time
- Using AI tools to tidy or summarise data, only with data you are allowed to share
- Checking any AI-produced figure against the source
- Looking back: did the earlier decision work?
What participants leave with
- A decision-first worksheet applied to a real question
- Their own list of three to five numbers, each defined in a sentence
- A data-quality checklist
- A simple review template for weekly or monthly use
- A list of the common traps in reading numbers
What this session is not
- A statistics or data science course
- A tour of dashboard or analytics software
- A guarantee that data will produce better results
- A replacement for experience and judgement
How the session runs
A practitioner who reviews numbers regularly in a business shows how, using hypothetical figures that are labelled as made up. Participants choose one decision, list the numbers that bear on it and note where each lives. They then practise checking a small sample dataset with planted errors, and draft their own review routine. Real business or customer data is not needed, and nobody is asked to reveal their figures.
What your students leave with
- A clear decision they want data to help with
- Three to five numbers worth tracking, and a reason for each
- A simple routine for checking that their data is complete and consistent
- Comparisons that mean something: over time, by customer type, per unit
- An awareness of the common ways numbers mislead
- A short weekly or monthly review they can actually keep up
Scheduled sessions
Nothing scheduled yet
Sessions are arranged with a college once a date is agreed. Ask us and we will find the right person for it.
A student rather than a college? See what is coming up, or ask your placement team to host this.
Pick the decision, choose a few numbers, check them, and look at them on the same day every month.
Tell us who your students are and what stage they are at. Sessions are free for participants.