Scanning

A scan collects signals, single concrete things that have already happened, from sources that would not normally sit in the same folder. Then it throws most of them away, on rules written down before the throwing started. The collecting is easy. The discarding is the skill.

Futuring: the four moves

Move 1 of 4 · Widens what counts as evidence

Structured reading across sources that disagree, then disciplined discarding.

Time: 3 hours 20 minutes, over two sessions
People: 3–8, including whoever owns the decision
You need: A shared spreadsheet or index cards, a source list, the two worksheets
You leave with: 40–80 signals, 5–8 named shifts and a log of what was discarded and why

Start here

The idea

What the move is, and why it works

What the move is

Scanning is structured reading. You choose a question and a horizon, read across sources that disagree about what matters, and record signals. A signal is one concrete thing that has happened: a product launched, a policy passed, a court case, an odd new habit. It is not a trend. A trend is what you get after you have interpreted several signals, and the scan’s job is to delay that step long enough to see something new.

What survives the narrowing is clustered into five to eight shifts. Each shift is named as a change, so it says which way things are moving. “Credentials leave the school” is a shift. “Credentials” is a topic, and a topic tells you nothing about direction.

The idea behind it

Every planning conversation reasons from evidence about the present, because that is the only evidence anyone has. Left alone, the evidence also comes from inside education and from the last two years. The same three reports get cited in every proposal.

A scan widens that base on purpose. Sources run at different speeds. An international report tells you what has already been agreed is important, and it is usually two years late. A social platform tells you what people feel before anyone has written it down, and it is mostly noise. Read only the slow sources and you get the recent past with new dates. Read only the fast ones and you mistake a fortnight for a trend. You need both.

Why the discarding is the skill

Anyone can collect 400 signals. The judgement is in what you drop, and whether you wrote down the grounds before or after you dropped it. Rules written afterwards tend to explain the choices the group had already made for other reasons.

There is a cost to name. The usual filter is fit with what the school or system has already committed to. That keeps the scan answerable, but it also filters out the signals most worth seeing: the ones that would force a commitment to be reopened. The partial remedy is a visible discard log, so the filter can be argued with later.

When

When to reach for it, and when not to

Use it when

The plan’s evidence all comes from inside education, or from the last two years.
A strategy review is due in the next term and nobody has looked outside the sector since the last one.
The group keeps describing the future as today with faster devices.
You are about to write scenarios and have nothing to build them from.

Skip it when

A good published scan already covers your question and horizon. Use it, and spend the time on scenarios.
The horizon is under a year. Below that, the future is mostly the present with known events in it.
Nobody will read the output. A scan with no decision and no reader is a reading list.

Before

Before anyone is in the room

Write the question and the horizon on one line, and check the person who owns the decision agrees with it.
Decide, and say aloud, whether the scan touches a live decision or is meant to change how the group thinks. Both are fine. They are judged differently.
Draw up a source list across at least five families: international bodies, research, journalism, specialist newsletters and social platforms.
Give each person two source families they do not normally read.
Set up one shared log with the columns from the signal log worksheet, so signals arrive in the same shape.
Book the second session two to three weeks after the first, and put a review date in the diary now.

Steps

How to run it

8 steps, about 3 hours 20 minutes, over two sessions in all. The questions in quotation marks are meant to be read out as they are.

01
Fix the question, the horizon and the decision
20 min
Agree one question and one horizon. For example: what could change what secondary students need to learn to write, by 2035? Then name the decision the scan is meant to touch. If there is none, write that down too.
Ask the room
- What decision will this scan be allowed to change?
- If there is no decision, what do we want to think differently about by the end?
- How far out is far enough that today’s arrangements stop being a constraint?
Watch for A question so broad it covers everything (“the future of education”) or so narrow it is already answered. Push for one that names a group of students and a date.
02
Hand out sources that disagree
15 min
Give each person two source families they do not normally read, across social, technological, economic, environmental, political, demographic and teaching-practice change. Agree how many signals each person brings back. Eight to twelve is enough.
Ask the room
- Which source would you never normally open, and why not?
- Whose view of schools is missing from our list?
Watch for Everyone volunteering for the institutional reports. They feel safe and they are always late.
03
Collect signals, not trends
60 min
Each person reads and logs. Each signal goes in as one sentence about something that has happened, with its source, its date and why it might matter. This can be done between sessions, alone, and usually is.
Ask the room
- What actually happened here, in one sentence?
- Is this one event, or am I already summarising several?
- Why might this matter to a student in our school in ten years?
Watch for Entries that start with “the rise of” or “increasing”. Those are trends. Send them back for the event underneath.
04
Write the discard rules before discarding
15 min
Before anyone looks at the pile as a whole, agree three to five rules a signal can fail. For example: drop anything that only extends a current trend, or that has no plausible route into a school within the horizon. Write them on the scan frame.
Ask the room
- What would make a signal not worth keeping?
- Which of these rules would have thrown out something we now take for granted?
- What do we keep even if it breaks a rule?
Watch for Rules that quietly protect current commitments. Add one rule that keeps any signal which would force a commitment to be reopened.
05
Make each signal answer a lens it does not belong to
25 min
Take the signals in turn and read each through a lens other than the one it came in under. A technology signal read as a demographic one, a social signal read as a question of teaching practice. Note anything new in the log.
Ask the room
- If this were a story about families, not technology, what would it say?
- What does this signal mean for a teacher on a wet Tuesday?
- Who loses if this spreads?
Watch for The group nodding signals through. If a lens produces nothing for ten minutes, switch lenses, not signals.
06
Narrow, and keep the discard pile
25 min
Apply the rules. Every discarded signal goes on the discard log with the rule that removed it. Every kept signal that broke a rule gets one line explaining why it stayed.
Ask the room
- Which rule removes this one?
- If we are wrong to drop it, what would we see in two years?
Watch for Discarding without naming a rule. That is a preference, and it should be logged as one.
07
Cluster what survives into named shifts
30 min
Group the kept signals into five to eight shifts. Name each as a change with a direction, in a short sentence with a verb. Then read the names aloud and cross out any that are topics.
Ask the room
- Which way is this moving, and from what to what?
- What would a school notice first if this shift were real?
- Which two of these shifts could pull against each other?
Watch for A shift that every signal fits into. It is a heading, not a shift. Split it.
08
Date it and book the review
10 min
Put the date on the scan, and a review date six to twelve months out. Name one person who will look at the discard log again at the review.
Ask the room
- When will this be out of date?
- Who will look at what we threw away, and when?
Watch for A scan with no review date. It will still be cited in three years as if it were taken yesterday.

Worksheets

Two sheets to print and fill in

Each one prints on its own sheet of A4. The grey line under each is an example from the worked situation.

Futuring · Scanning · Worksheet

Scan frame and discard rules

Fixes the question and the rules for discarding before anyone starts collecting, so the narrowing can be argued with later.

Question

Horizon

Date

Review on

The decision this scan is meant to touch

Or write “none: this is to change how we think”, and mean it.

Source families each person will read that they normally would not

Discard rules, written before any discarding

Three to five tests a signal can fail.

Signals kept that break a rule, and why

The shifts, each named as a change

“Credentials leave the school”, not “credentials”.

Example Discard rule from the Singapore scan: drop a signal with no plausible route into a Singapore classroom within the horizon, unless it would force a current commitment to be reopened.

geraldajam.com/work/toolkit

Futuring · Scanning · Worksheet

Signal log

Records each signal in the same shape, with the lens it was forced through and the rule that kept or dropped it.

Name

Source families

Date

SignalOne sentence. Something that happened.	Source and date	Lens it came in under	Lens it was made to answer	Why it might matter here	Keep or dropAnd the rule

Example An online learning platform issues verifiable credentials employers can check without the school | Think-tank report, 2025 | Technological | Pedagogical | If credentials travel without schools, who decides what counts as learning? | Keep: touches stated aims for lifelong learning

geraldajam.com/work/toolkit

Worked example

Scanning for Singapore’s schools

A horizon scan for Singapore’s schools

The group ran the scan over several months in late 2025. There was no decision to touch, so we said so at the start: this was sensitisation, a way of changing how we thought about where schooling in Singapore was heading, not a plan. That shaped every later choice.

Six source families, chosen for their unevenness. International bodies and think tanks, media reports, academic literature, a lighter read of social platforms, and alerts from scanning tools. The institutional sources carried a two-year lag and the social ones a two-day one. Each friend took families they did not normally read.
Seven lenses, two of them added. The usual social, technological, economic, environmental and political lenses, plus demographic and pedagogical. A scan for schools that cannot see falling enrolments or classroom practice is not a scan for schools. Almost every useful reading came from a signal forced through a lens it did not obviously belong to.
Narrowed against what the system says it is aiming for. Signals were weighed against the system’s stated aims for its graduates, its competency frameworks and the curriculum commitments already under way. This kept a global scan answerable to local priorities. It also filtered out the signals that would have required those commitments to change, which is the cost we kept arguing about.
Each societal signal read twice. Six societal signals survived, from hyper-individualised learning to geopolitical instability. Each was written once in its best light and once in its worst. A signal you can only see one way is a signal you have already made up your mind about.
A separate scan for technology. Societal signals say what people will want, not what will be possible. A parallel scan ran on its own criteria and surfaced five technologies to watch, including spatial computing and neurotechnology. The group pushed back hardest on neurotechnology. Being ready to use at scale is not the same as being right to use.
The awkward part. The social platforms gave us the most vivid material and the least defensible material, and we could not separate the two cleanly. Several of our liveliest signals came from there. None of us could say with confidence which of them were real shifts.

What came out of it

Six societal signals each read two ways, three local developments that held in any future, five technologies to watch and a discard log. The scan also produced its own doubt: the filter we used was the one most likely to hide what mattered, and only the discard log let us argue with it afterwards.

Signs

How to tell it is working, and when it has gone hollow

Working

Someone brings back a signal that makes the group uncomfortable, and it survives the narrowing.
The discard log is longer than the keep list, and every line on it names a rule.
A shift name starts an argument about direction, not a nod.
People start noticing signals in ordinary reading weeks after the session.

Gone hollow

The signals all come from the same three kinds of source the plan already cites.
Nothing is thrown away. The output is a library, not a judgement.
The shifts are topics: “AI”, “wellbeing”, “climate”.
The scan is cited a year later with no review date and nobody has opened the discard log.

Variations

Other ways to run it

In 45 minutes

Skip the collecting. Hand out a published scan, such as the OECD’s Trends Shaping Education, and spend the time on discard rules, one lens swap per person and naming three shifts. It is a reading exercise with a discipline attached, and that is most of the value.

On your own

Keep a running log for a month: two signals a week from a source you do not normally read, each in the signal log format. At the end, write your discard rules first, then narrow. Ask a colleague to read your discard log and argue with one entry.

With students

Give a class of upper secondary students the signal log and ask them to bring back signals about how people their age learn outside school. They will find things no staffroom would. Ask them to write the discard rules, and listen to what they think does not count.

What to do with what you made

Put the named shifts and the discard log in front of whoever owns the decision, with the review date on the cover. If the scan was for sensitisation, run the scenarios move within a month, while the signals are still fresh. A scan that sits for a term is already ageing.

Scenarios Crosses the two most uncertain shifts into four worlds.The Singapore signals The source families, lenses and narrowing in full.The scan as a whole How the collecting, narrowing and outputs looped back on each other.

Tools

Tools, templates and guides worth having

Published elsewhere, and worth the time.

Toolkit
The Futures Toolkit (Government Office for Science, 2017)
Free and written for people with a day, not a degree. Its horizon scanning section covers sources, signals and clustering.
Report
Trends Shaping Education (OECD, published every few years)
A ready-made scan for schools, with questions for discussion. Use it as a starting point, never as the scan.
Method
STEEP and PESTLE
The standard sets of lenses. Add demographic and pedagogical for schools, as the Singapore scan did.
Book
The Signals Are Talking (Webb, 2016)
A step-by-step method for telling signal from noise and clustering what survives.

Reading

Where it comes from

Start with the first on the list.

Government Office for Science. (2017). The Futures Toolkit. UK Government.
Practical, free and covers all four moves.
Hines, A., & Bishop, P. (2006). Thinking about the Future: Guidelines for Strategic Foresight. Social Technologies.
The practitioner’s checklist for framing and scanning.
Webb, A. (2016). The Signals Are Talking. PublicAffairs.
How to tell a signal from noise, with a method for clustering.
Saritas, O., & Smith, J. E. (2011). The big picture: trends, drivers, wild cards, discontinuities and weak signals. Futures, 43(3), 292–312.
Definitions that stop a group arguing about words.

Singapore