Analogs & Precursors
Borrowing solutions that already exist — looking sideways across other industries and backward through your own industry’s past — to spark genuinely new ideas without inventing from nothing.
Every good idea already exists somewhere. The skill is knowing where to look and what to abstract from what you find.
What it is
A structured search for solutions that already exist — across other industries and back through your own.
Analogs and precursors are two directions you can search when you need a genuinely new idea but do not want to start from nothing. Both directions are based on the same insight: the solution you need almost certainly exists in a form somewhere, because most problems have already been solved — either by another industry working on a structurally similar challenge, or by a previous generation working on your exact challenge before the conditions were right.
Analogs search across space: they look sideways at other domains, industries, and contexts that have already solved a problem structurally similar to yours, even when the surface contexts look nothing alike. The connection is never obvious. A hospital and a luxury hotel share no surface similarity; what they share is the structural challenge of orchestrating consistent, dignity-centred service across a large, distributed team. The hospital finds the principle by abstracting past the surface similarity. That abstracted principle — not the imitated practice, but the structural insight — is what makes the analog productive.
Precursors search back through time: they examine the history of your own industry for earlier attempts at solving the same problem you are now facing. Something tried and stalled. A patent that was filed and never shipped. A startup that was ahead of its time. The question precursors ask is not “what happened?” but “why?” — and specifically, was the earlier attempt premature (the infrastructure, cost curve, or behavioural readiness was not yet there) or fundamentally flawed (the concept itself was wrong)? That distinction is the timing diagnosis, and it is the entire value of the precursor search.
The two directions are perpendicular on purpose. Analogs give you freshness from distance: the further the domain, the less obvious the principle, and the more genuinely new it feels when applied to your problem. Precursors give you ripe ideas hiding in history: already validated by the need, already refined through failure, waiting for the conditions to catch up. Together they create a search space that is broader and more generative than any amount of brainstorming from first principles.
Explore the search space
Click an axis to understand the search direction. Click a point to see a specific example.
The horizontal axis searches across industries for structural matches. The vertical axis searches backward through time for earlier attempts. Both originate at your current problem.
When to deploy it
For ideation that needs range. Not for optimization that needs refinement.
Use Analogs & Precursors when
- →Your team keeps generating variations on the same familiar solutions — a sign the search space is too narrow.
- →You need a genuinely new concept and first-principles thinking is producing either nothing or the obvious.
- →You are entering an established category and need to understand both the landscape of analogous domains and the history of earlier attempts in your own.
- →Ideation feels stuck inside the conventions of the industry, and you need structural distance from those conventions before anything else will work.
Do not lean on it when
- ×You are optimizing an existing concept rather than generating new ones — analog research has a high setup cost that is not justified for incremental refinement.
- ×The team treats the analog as a solution to copy rather than a source of structural principle to abstract. Imitation of surface features (the chocolates on the pillow) is not the method; abstraction of the structural insight is.
- ×There is no time or appetite for genuine abstraction. The method fails when teams list analogs but skip the step of articulating what principle from the analog is actionable in their context.
The honest limit: the abstraction step is genuinely hard, and most teams skip it. Finding the analog is easy; articulating the structural principle it contains — the transferable insight that is not domain-specific — requires careful thinking and often multiple iterations. A list of analogs that has not been abstracted produces imitation, not innovation.
How it works
Seven moves, for analogs and precursors run together or each run independently.
Frame the underlying problem structurally.
Before searching, translate the problem from its domain-specific terms into structural language. Not "how do we improve hospital patient experience?" but "how do we orchestrate consistent, dignity-centred service across a large, distributed team with many handoffs?" The structural framing is what makes the analog search work — it opens the field to every domain that has solved the same underlying structure, regardless of surface appearance.
Search for analogs across industries.
Ask who else has already solved the structural problem you framed. Cast deliberately far — the near analogs (other hospitals, other clinics) are obvious and will surface naturally. The search that pays off is the one that finds the far structural match: the hotel, the pit crew, the airport. Push past the first tier of obvious answers. The most productive analogs are rarely the ones the team names in the first five minutes.
Abstract the structural principle from each analog.
For each analog, identify the structural insight — not what they do, but why it works. This is the most important and most often skipped step. "Hotels do chocolates on pillows" is a surface observation that produces nothing actionable. "Hotels train every person in the building to the same service standard, not just the guest-facing roles" is a structural principle that transfers. Write the principle in terms your problem space can act on.
Search backward for precursors in your own industry.
Look for earlier attempts at solving the same problem: products launched and abandoned, patents filed and not used, startups that tried and did not scale, research projects that demonstrated the concept but could not ship it. The search is historical and focused on your category or close adjacencies. The question is not "what happened?" but "what are we looking at?" — is this a dead end or a ripe idea?
Diagnose the timing of each precursor.
For every precursor found, make a timing judgment: was this premature (the infrastructure, technology, cost curve, or behavioral readiness was not yet there) or structurally flawed (the concept itself was wrong)? These are very different findings. A premature precursor suggests the idea may now be ripe; a structurally flawed one is a warning. The diagnosis requires examining what specifically was missing and whether it is now present.
Synthesize across both axes into a working hypothesis.
After running both searches, look for convergent signals: an analog pointing in the same direction as a ripe precursor is a strong signal. The combination of "another domain already solved this" and "someone tried this before and was merely premature" is a powerful case for moving forward. Synthesize the structural principles from analogs with the timing insights from precursors into a clear hypothesis about what to try.
Carry the principles into ideation and concept development.
The output of analog and precursor research is not a solution; it is a set of structural principles and timing hypotheses that seed concept development. Each abstracted principle becomes a frame for How Might We questions or for Crazy 8s sketches. The research done here is the fuel for the next step — it must be actively fed into whatever ideation method follows, not left as a research report that gets filed away.
Best practices
What separates a productive search from a list of interesting examples that goes nowhere.
When it goes well
- ✓The structural framing is crisp before the search begins — not domain-specific but expressed in terms that could apply to multiple industries.
- ✓The analog search pushes deliberately far. The team fights the pull toward near, surface-similar examples and looks for the truly distant structural match.
- ✓Every analog is abstracted to a structural principle, not left as a domain description. The principle is expressed in terms the problem space can act on.
- ✓Precursors are diagnosed for timing, not just listed. The team distinguishes premature from flawed and records the specific conditions that were missing.
- ✓Both axes are run, and the outputs are synthesized before feeding into ideation. Convergent signals from both directions are weighted heavily.
- ✓The research is actively fed into the ideation method that follows — it does not remain a report, it becomes the input material for What Might We and sketch sessions.
The mistakes, and how to avoid them
Skipping the structural framing.
Jumping straight into "find analogies" without first translating the problem into structural language produces near, surface-similar results. The hotel shows up when you frame the problem as "how do we orchestrate consistent service across many handoffs?" — not when you frame it as "how do we improve hospital experience?"
Stopping at near analogs.
The first tier of analogs is always the obvious: other hospitals, other clinics, hotel brands already mentioned in the healthcare literature. These near analogs are too close to produce genuinely new principles. Push past them deliberately. Set a rule: if the team heard about the analog in their own industry's articles, it is too near.
Not abstracting the principle.
A list of interesting examples with no abstraction step produces nothing actionable. "Disney does X" is not a deliverable. "Disney designs every element of the physical environment to produce a specific emotional response at each moment, and rehearses every staff interaction to be consistent with that design" is a structural principle you can apply. The extra step is mandatory.
Treating precursors as a dead-end list.
Finding that something was tried before and failed is the beginning of the precursor analysis, not the end. Without the timing diagnosis — was it premature or flawed? — the list of failures is a caution sign, not a finding. The diagnosis is what makes precursor research valuable.
Filing the research and not feeding it forward.
Analog and precursor research is only as valuable as what it seeds. If the structural principles and timing hypotheses go into a slide deck and not into a How Might We session or a sketch exercise, the method has produced inputs with no outputs. Run the ideation session immediately, while the research is still fresh.
Logistics
Building an analog library and running the precursor diagnosis as a team practice.
Analogs and precursors can be run as a single focused session or as an ongoing practice that the team builds over time. A one-time sprint session produces a narrow set of inputs; a team that continuously builds an analogical library and monitors its category’s history produces richer ideation material whenever it needs it.
Assign research in advance
The productive analog session is not a blank brainstorm — it is a structured debrief of research done in advance. Assign each team member two or three domains to investigate before the session: how has the hotel industry solved service consistency? How have airports managed complex, stressful multi-stage journeys? Individuals bring structured findings; the session synthesizes and abstracts, not generates.
Use a consistent reporting format per analog
For each analog brought to the session, document: the domain, the structural problem they solved, the specific practice or mechanism they used, and the abstracted structural principle. That four-field format creates a consistent vocabulary for comparison across domains and makes the abstraction step explicit rather than implicit. Without it, sessions produce lists of interesting examples rather than actionable principles.
Run analogs and precursors in parallel, not in sequence
Split the team: half searches the analog axis (across industries), half searches the precursor axis (backward through time). The parallel search prevents either direction from being neglected when time is short. Converge after both directions have been researched, not during the research itself.
Build the analogical library as an ongoing practice
The most effective teams treat analog collection as a continuous background practice rather than a session-specific activity. A shared repository where any team member can log an interesting domain, a structural principle they noticed, or a relevant precursor they came across means that ideation sessions always start with a richer library than a single research sprint can produce.
Document the timing diagnosis explicitly
For every precursor, record the timing judgment in writing: premature or flawed, and specifically what was missing. This is the finding that is most commonly lost — teams note that something was tried before and forget to record why it failed and whether the conditions have changed. The diagnosis belongs in the team's working document, not just in the facilitator's head.
AI and this method
AI retrieves near analogs fluently and lists precursors readily. It struggles with the far jump and the timing diagnosis.
Toggle between modes to see how the search space changes when AI leads the search — what it clusters on, and what it misses.
In-depth example
Cleveland Clinic: a far analog and a precursor diagnosis that changed the category.
Cleveland Clinic needed to transform patient experience. The breakthrough came from a single far structural analog — luxury hotels — and a precise precursor diagnosis about what earlier attempts had borrowed wrong. Toggle to see what changed with a hypothetical AI-first search.
The analog search
The team was explicit about starting far from healthcare. They did not ask “what do other hospitals do?” That would have surfaced near analogs — similar settings with similar constraints, confirming what was already known. Instead, they asked: “who has already solved the problem of making people feel cared for across many staff, many touchpoints, and a complex choreography they do not control?”
That question pointed to luxury hotels. Not because hospitals are like hotels in any surface sense, but because the structural problem was identical: consistent, dignity-centred service across a large, distributed team where the guest or patient is always at the mercy of the next handoff.
What they abstracted — and what they refused to borrow
What they noticed in hotels
Every person in the building — not just the concierge, not just the room service team — was trained to see themselves as part of the guest experience. A hotel housekeeper who passed a guest in the hallway greeted them by name. A bellman who overheard a complaint addressed it. The experience was choreographed across the entire staff, not delegated to a guest relations department.
The abstracted principle
Patient experience is the sum of every handoff — from orderly to nurse, from receptionist to physician — not a single moment. Every person in the building is responsible for it. Dignity-centred care cannot be delegated to a department; it must be the standard for the entire organisation.
What they refused to borrow
The surface details: chocolates on pillows, turndown service, concierge desks. Those are implementation details, not the principle. Importing hotel aesthetics into a hospital would have been imitation. Importing the organisational model for consistent service across all staff was the analog that mattered.
The precursor check
The team also looked backward. Had hospitals tried to redesign patient experience before? Yes — hotel-inspired “patient amenity” programmes from the 1990s had largely failed, and the team diagnosed why: they had borrowed the surface (aesthetics, amenities) rather than the structure (organisation-wide service training). The earlier attempts were not premature — they were misdirected.
That precursor diagnosis sharpened the approach: the principle to borrow was organisational, not cosmetic. The team created the first Chief Patient Experience Officer role in a major hospital system — a structural change modelled on the hotel’s Director of Guest Services, not its interior design.
What the search produced
The combination of one far analog (luxury hotels) and one precursor diagnosis (earlier patient amenity programmes had borrowed the wrong thing) gave the team a sharp, actionable principle. Cleveland Clinic went from the bottom quartile of patient experience scores in their peer group to the top. The insight — that experience is an organisational model, not a service layer — came entirely from looking sideways and backward, not from studying what other hospitals were doing.
Two directions, one method
The two axes are perpendicular on purpose. Each direction requires a different search skill.
Analogs search outward across other industries in the present. Precursors search backward through your own industry’s history. The skill each requires — abstraction versus timing diagnosis — is different enough that they benefit from being run separately before being synthesized.
Frameworks
Where Analogs & Precursors shows up.
An ideation-phase method designed to expand the solution space before convergence begins. It appears at the moments in each framework when conceptual range matters most and incremental thinking is the main risk.
Related methods
What to combine with Analogs & Precursors.
Sources & further reading
The work behind this method.
The Medici Effect
Frans Johansson (2004)
The defining account of intersection thinking — the idea that breakthrough innovations occur at the intersection of concepts from different fields, not within any single field. Johansson's central argument is that the further you search from your starting domain, the higher the density of potential ideas, because you encounter concepts and principles that have never been combined with your field's problems before. This is the core intuition behind the analogs direction: the value of the far search grows with distance.
Creative Confidence
Tom Kelley and David Kelley (2013)
The Kelleys' treatment of how design thinkers build the habit of looking to other domains for structural inspiration. Their concept of the analogical library — a team's personal collection of insights, principles, and mechanisms from adjacent and distant domains, built continuously as a practice rather than a one-time exercise — maps directly onto the analog axis of this method. The book's cases show repeatedly how far structural matches produced solutions that near matches could not.
Where Good Ideas Come From
Steven Johnson (2010)
Johnson's analysis of innovation patterns across centuries provides the best conceptual frame for why precursors matter. His concepts of the "slow hunch" (ideas that circulate and develop over years before becoming viable) and "the adjacent possible" (the space of ideas that are one step beyond what currently exists) explain why historical precursors are such a rich source of ripe innovations. An idea that failed in 1990 because the cost curve had not moved is a slow hunch waiting for the adjacent possible to catch up — exactly what the precursor timing diagnosis is designed to identify.