Backward design: planning learning from the outcome

What is backward design?

Backward design is an approach to planning learning that starts with what learners should be able to understand and do, then works backwards to decide how they will demonstrate it and how they will learn it.

It has three stages:

Identify the desired results: what should learners know, understand and be able to do?
Determine acceptable evidence: what would demonstrate that they have achieved those results?
Plan learning experiences and instruction: what teaching, practice and support will help them succeed?

The central question is not “What shall I cover?” but “What should learners be able to do with their learning?”

For example, a workplace trainer might replace “Deliver a presentation on complaints handling” with “Enable colleagues to resolve routine complaints fairly and recognise when escalation is necessary”.

The presentation could still be useful. However, it becomes a resource chosen to support learning, rather than the starting point that determines everything else.

Where backward design comes from

Grant Wiggins and Jay McTighe popularised backward design through Understanding by Design, first published in 1998 and expanded in a second edition in 2005. Their framework connects desired learning, assessment evidence and teaching.

It builds on earlier traditions of planning education around purposes and evaluating whether those purposes have been achieved. Wiggins and McTighe did not invent the general idea of starting with an end in mind.

Their distinctive emphasis is on understanding and transfer: learners should make sense of important ideas and use their learning in situations they have not simply rehearsed.

Understanding by Design, often shortened to UbD, is a published, commercially supported framework. The core planning approach can be used without buying a template, course or consultancy service.

Backward design is a planning framework, not a guarantee of improved results. Its rationale is alignment between outcomes, evidence and teaching. Whether a particular course succeeds also depends on subject expertise, teaching, accessibility, practice and feedback.

Why backward design matters

Starting with topics or slides can produce a course that looks complete but leaves an important question unanswered: can learners actually use what was taught?

A learner might:

Recall a complaints policy but struggle with an upset customer.
Identify a calculation method but choose the wrong one in a workplace problem.
Describe inclusive supervision but fail to adapt their communication.

Backward design helps expose these gaps before delivery.

It is particularly useful in FE, Skills and workplace learning because outcomes often involve applying knowledge under realistic conditions. Attendance, completed slides and a successful recall quiz may provide some information, but they do not necessarily demonstrate practical capability.

The approach also helps teachers and trainers prioritise. When time is limited, each activity should have a clear purpose: developing essential knowledge, building understanding, practising application or revealing a learning gap.

The three stages explained

Stage 1: identify the desired results

Start with the intended destination. In an FE programme, this may be shaped by qualification specifications, occupational requirements and learners’ starting points. In workplace training, it may come from a genuine performance need.

Distinguish between three kinds of result:

Type of result	Planning question	Example
Knowledge	What must learners know?	The complaints process and escalation routes
Understanding	What must learners make sense of?	Why a fair response depends on evidence, authority and context
Transfer	What should learners do independently in a new situation?	Handle an unfamiliar routine complaint and justify their response

Avoid stopping at vague outcomes such as “Understand complaints handling”. Understanding matters, but you need to clarify how it will become visible.

A more useful outcome is:

Given an unfamiliar customer complaint and access to the current policy, the learner can identify the issue, choose an authorised response, explain their reasoning and record the action appropriately.

Notice the conditions. Learners have access to the policy because that reflects the job. Requiring memorisation would test something different.

An overarching question can also help organise the learning:

What makes a response fair when the customer’s preferred solution is not available?

This creates room for reasoning, rather than reducing the session to a script.

Stage 2: determine acceptable evidence

Before choosing activities, decide what would count as convincing evidence of learning.

Ask:

What would learners need to say, produce or do?
Which criteria distinguish a successful response from an unsuccessful one?
Does the task test the intended capability?
What support or resources should be available?
Will one task be sufficient, or is evidence needed across several situations?

Assessment does not have to mean a formal examination. Evidence might include a worked solution, practical demonstration, explanation, role-play, portfolio item or observation of workplace performance.

Choose methods that match the outcome. A multiple-choice question may check knowledge of an escalation rule. It cannot, by itself, demonstrate that someone can manage a difficult conversation.

Use both:

Formative evidence, gathered during learning to guide feedback and teaching.
End-point evidence, gathered to judge whether the intended outcome has been achieved.

For assessed qualifications, check awarding organisation requirements before designing your own assessment. Backward design helps you prepare learners for valid assessment; it does not authorise changes to prescribed conditions or criteria.

Stage 3: plan learning experiences and instruction

Now select the teaching and practice learners need.

Work backwards from the assessment:

What knowledge is needed to attempt the task?
Which decisions are likely to be difficult?
What misconceptions could lead to errors?
What should be modelled explicitly?
Where will learners practise with support?
When will that support be reduced?

A useful sequence is:

Check starting points.
Explain and model the essential knowledge and reasoning.
Provide guided practice.
Give feedback and opportunities to improve.
Introduce varied situations.
Check independent application.

This is a practical sequence, not a compulsory UbD formula. Some learners may need additional instruction or practice. Others may already possess the foundational knowledge.

Planning backwards does not mean teaching backwards. Learners still need a sensible route from foundations to increasingly demanding application.

Worked workplace example: handling customer complaints

The following is an illustrative training design, not a prescribed industry standard.

A customer service manager reports that newer colleagues can find the complaints policy but make inconsistent decisions about resolving or escalating cases.

Stage 1: set the outcome

By the end of the initial training and supported practice, colleagues should be able to:

Clarify the customer’s concern.
Use the policy to identify available options.
Act within their authority.
Explain the next step respectfully.
Record relevant information without unnecessary personal detail.

The aim is not to eliminate every escalation. Appropriate escalation is part of successful performance.

Stage 2: decide what evidence is needed

Learners complete an unfamiliar simulated complaint and create the accompanying case note.

The trainer uses criteria covering:

Accurate identification of the issue.
A response consistent with policy and delegated authority.
Clear, respectful communication.
Appropriate escalation where necessary.
An accurate and relevant record.

A learner who sounds confident but promises an unauthorised refund has not met the intended outcome.

A learner who refers to the policy before deciding may have demonstrated sound practice, not weakness.

The simulation provides initial evidence. A later review of suitable workplace cases, with appropriate privacy controls, checks whether learning transfers into day-to-day work.

Stage 3: build the learning

An initial workshop could include:

A diagnostic case: learners explain what they would do and why.
A worked example: the trainer models how to interpret the policy and weigh options.
Contrasting cases: learners compare situations requiring resolution, referral or escalation.
Guided rehearsal: pairs practise with feedback.
An unfamiliar scenario: learners demonstrate their decisions more independently.
A follow-up plan: managers identify opportunities for supported workplace practice.

Slides might introduce the process or display a worked example. They support the design rather than define it.

How this differs from a slides-first approach

Slides-first starting point	Backward design starting point
“We need a session on the complaints policy.”	“What decisions must colleagues make correctly?”
Organise slides around policy headings.	Define outcomes and evidence before selecting content.
End with a recall quiz.	Use a task that requires application, supported by knowledge checks.
Treat completion as success.	Check what learners can demonstrate and whether it transfers.

A topic list is not inherently unhelpful. It becomes a problem when coverage substitutes for a clear learning purpose.

FE example: using maths to plan catering quantities

An FE tutor wants learners to apply ratio and proportion in a vocational catering context.

Desired result: learners can scale a recipe for a changed number of portions, calculate ingredient quantities and explain whether their answers are reasonable.

Acceptable evidence: learners complete an unfamiliar catering order, show their calculations and explain how they checked one result. Where relevant, the task includes a unit conversion.

Learning activities: the tutor checks prerequisite knowledge, models a scaling calculation, addresses common conversion errors and provides practice with different quantities before independent application.

Two checks are important here.

First, the catering context should make the maths meaningful without introducing unnecessary reading barriers or assuming specialist knowledge that has not been taught.

Second, completing several nearly identical examples may show that learners can follow a pattern. An unfamiliar but appropriately challenging task provides better evidence that they can select and use the method themselves.

Any formal qualification assessment must still follow the applicable specification and assessment arrangements.

A practical backward design checklist

Use this before building your next session or module.

Define the destination

What real need is this learning addressing?
What should learners know, understand and do afterwards?
What should they be able to apply in a new situation?
Which requirements are fixed, and where do I have flexibility?

Specify the evidence

What task would make the intended learning visible?
What criteria will I use?
Could someone succeed through guessing, copying or presentation skills rather than the intended learning?
Are the assessment conditions appropriate and accessible?

Build the route

What prerequisite knowledge should I check?
What explanation, modelling and practice are needed?
Where will feedback change the learner’s next attempt?
How will I support learners without removing the intended challenge?

Finally, check alignment in both directions: every outcome should have appropriate evidence, and every substantial activity should support an outcome.

Limitations and common misunderstandings

It can become too assessment-driven

Designing around a narrow test can produce narrow learning. Start with worthwhile outcomes, then choose evidence that represents them. Do not define the whole curriculum around what is easiest to mark.

Not every outcome fits a simple checklist

Professional judgement, creativity and ethical reasoning may require portfolios, discussion, observation over time and several examples of performance. Clear criteria can support judgement without prescribing one identical response.

A training problem may actually be a workplace problem

People may already know what to do but lack time, authority, functioning systems or managerial support. Backward design cannot fix those conditions through better activities alone.

A plan must remain adaptable

Planning the destination in advance does not mean ignoring learners’ questions or emerging needs. Formative assessment should influence pace, examples and support.

It does not replace teaching expertise

Alignment is useful, but aligned activities can still be poorly explained or insufficiently practised. The framework helps organise decisions; it does not make those decisions automatically.

Related approaches

Constructive alignment, associated with John Biggs, also connects intended outcomes, teaching activities and assessment. It emphasises learners constructing meaning through activities aligned with those outcomes.

Competency-based learning focuses on demonstrating specified capabilities. Backward design can help plan the route towards those demonstrations, while recognising that competence may require evidence over time.

Problem-based and project-based learning can sit within a backward-designed course. Choose the intended learning and evidence first, then select a problem or project that genuinely develops them.

Summary and your next step

Backward design starts with meaningful results, identifies evidence of learning and then builds the teaching and practice required.

Its value lies in shifting attention from what the trainer presents to what learners can understand and use.

Your next step: choose one upcoming session and complete this sentence before opening your slides:

“After this learning, people should be able to ___, and I will know because they can ___.”

Use that answer to decide what to keep, change or remove.

Sources and further reading

Wiggins and McTighe: Understanding by Design, expanded second edition. The original authors’ detailed account of backward design, understanding and transfer.
Vanderbilt University: Understanding by Design. An accessible university teaching guide to the three stages and their application.
Biggs: Enhancing teaching through constructive alignment. The original 1996 journal article explaining the related constructive alignment approach.
Carnegie Mellon University: Why should assessments, learning objectives and instructional strategies be aligned?. Practical guidance on matching assessment and teaching to intended learning.