Emerging trends: how to track them before they are obvious

The short answer

Emerging trends are changes that keep gaining momentum across independent sources before most people notice them. To track them early:

1. Define the domain, the time horizon and the decision the tracking should inform.
2. Pick signal sources that move early: preprints, patents, job postings, funding, developer activity, regulatory dockets and niche communities.
3. Record a baseline for each source so you know what normal looks like.
4. Watch for acceleration in one source and convergence across several unrelated ones.
5. Validate the signal against counter-evidence and structural drivers.
6. Decide: act, keep watching, or drop it, with the trigger written down.

What is an emerging trend, and how is it different from a fad?

An emerging trend is a change in behavior, technology, markets or rules that is still early but keeps gaining momentum, and that shows up in several sources which do not copy each other. A fad is a burst of attention that rises fast in one channel and fades, because nothing underneath it is changing.

The distinction matters because the two look identical at the start. Both produce a spike in conversation. The difference shows up in what follows the spike.

Trends attract commitments that are slow and expensive to reverse:

• Researchers start publishing on the topic and citing each other.
• Companies file patents, open roles and ship early products.
• Investors fund startups whose plans depend on the change.
• Regulators open consultations or issue guidance.

Fads attract mostly talk. If a change has plenty of mentions but no commitments behind it, treat it as a signal to watch, not a trend to plan around.

A related term is the weak signal: an early, ambiguous piece of evidence that a change might be starting. The idea comes from Igor Ansoff's work on strategic surprise, which argued that organizations should respond to partial information instead of waiting until a change is certain and the cheap options are gone. Most weak signals lead nowhere. Trend tracking is the discipline of collecting them and noticing when several begin to agree.

Which leading indicators reveal emerging trends early?

Leading indicators are sources that move before the general public notices a change. They differ in how early they move and how much noise they carry, so a useful trend tracking setup mixes several types instead of relying on one.

Roughly ordered from latest to earliest:

• Search interest. Rising searches show public curiosity. Search data is easy to access and good for confirmation, but by the time a term climbs in search, the trend is often already visible.
• Niche communities. Specialist forums, subreddits, Discord servers and practitioner newsletters adopt new tools and practices before the mainstream. They are early but noisy, and enthusiasm there does not always travel.
• Job postings. When companies start hiring for a skill or a title that did not exist before, they are committing salary budgets. A new role appearing across several unrelated employers is a strong signal.
• Funding. Startup rounds, grants and corporate venture deals show where investors expect demand. A cluster of seed rounds around one problem is more telling than one large round.
• Developer activity. For software and technical domains, new open source repositories, contributors, package downloads and questions on technical forums show what builders are actually using.
• Regulatory dockets. Consultations, proposed rules, standards drafts and public comment periods signal that a change is big enough for rule makers to act on. They also move slowly, which gives you time.
• Patents. Filings record inventions years before products, and a first filing from a large company in an unfamiliar category can reveal a strategic bet.
• Research preprints. Papers on preprint servers such as arXiv or medRxiv appear before peer review and before any product. They are the earliest view, and the hardest to interpret without domain knowledge.

No single indicator is reliable on its own. Search can be moved by a news story, job postings by one company's hiring spree, funding by investor fashion. The value comes from reading them together.

How to track emerging trends: a step-by-step method

The method below works for a solo analyst or a strategy team. It turns trend tracking from occasional browsing into a repeatable process with a clear output: a decision.

Step 1: Define the domain and the decision

Write one sentence that states what you are watching and why. "Changes in how mid-sized manufacturers manage energy costs, to decide which product lines to invest in" is a scope. "Interesting tech" is not.

Set a time horizon as well. A product team watching the next two years needs different sources from a strategy team looking ten years out. The decision you name here is the filter you will use for every later step.

Step 2: Set your signal sources

For each leading indicator type, pick specific sources rather than categories:

• Two or three preprint categories or journals for research
• A patent search saved on the relevant classes or keywords
• Job boards filtered by the titles and skills that would signal adoption
• A funding database or newsletter covering your sector
• The repositories, packages or forums where builders in your domain gather
• The regulators and standards bodies that govern your market
• A short list of specialist communities and practitioners

Aim for breadth across types and depth within each. Ten excellent sources spread across six indicator types beat fifty news sites that all cover the same stories.

Step 3: Establish a baseline

Before you can spot acceleration, you need to know what normal looks like. For each source, note the current level: how many papers a month mention the topic, how many open roles carry the title, how often the regulator publishes on it.

A baseline can be rough. The point is to replace "this feels like it is everywhere" with "this has gone from occasional to weekly." Without a baseline, every new mention feels like a trend.

Step 4: Watch for acceleration and convergence

Two patterns separate real emerging trends from noise:

1. Acceleration. The rate of change increases, not just the level. Mentions that grow faster each period matter more than mentions that stay high.
2. Convergence. The same change appears in sources that do not influence each other. A research result, a patent filing, three job postings and a regulatory consultation pointing the same way is far stronger than fifty articles quoting one press release.

Convergence is the most reliable signal in the whole method, because independent sources rarely agree by accident. When they do, something structural is usually moving.

Step 5: Validate the signal

Before you bring a trend to a meeting, test it:

• Look for counter-evidence. Search for failed pilots, critical papers and companies quietly dropping the idea.
• Find the driver. Name the cost curve, regulation, demographic shift or technical breakthrough that would make the change keep going. A trend without a driver is often a fad.
• Check the source chain. Make sure your "independent" sources are not all citing the same original report.
• Ask a practitioner. One conversation with someone who works in the field can save weeks of misreading.

Step 6: Decide, and write down the trigger

Every tracked trend should end in one of three states: act now, keep watching, or drop. For "keep watching," write the specific evidence that would move it to "act," such as a second major company entering or a rule moving from consultation to final.

Writing the trigger in advance protects you from two opposite errors: acting on excitement, and ignoring a trend because it arrived slowly.

What is horizon scanning, and how do strategy teams use it?

Horizon scanning is the structured, team-based version of trend tracking. The UK Government Office for Science's Futures Toolkit defines it as "the systematic collection of insights on emerging trends and weak signals of change to identify potential threats, risks and opportunities."

The same toolkit describes five steps: recruit a scanning group, identify sources, gather scan data in an organized structure, analyze the scans together, and write up the results. It notes that scanning can run for a defined period or as a continuous function that informs strategy, and it recommends sources beyond the obvious, including expert consultation, professional journals and sources not published in English.

Three frameworks help keep a horizon scan honest:

• PESTLE and STEEP. These are checklists of categories for drivers of change. PESTLE covers political, economic, social, technological, legal and environmental drivers. STEEP covers social, technological, economic, environmental and political ones, and variants such as STEEPV add values. Their job is to stop a scan from becoming a technology scan by default, which is the most common drift.
• The futures cone. This model starts at the present and widens as it moves forward in time. The center line is what happens if current trends continue; the widening edges are the range of plausible and possible futures. Foresight practitioners often divide the cone into probable, plausible and possible futures, with a preferable future as the one an organization wants to steer toward. The cone is useful for labeling signals: a weak signal at the edge of the cone deserves watching, not a budget.
• Three Horizons. This framework, also in the GO-Science toolkit, looks at the dominant way of doing things today (horizon one), the emerging pattern that may replace it (horizon three), and the transition between them (horizon two). It helps teams see that an emerging trend often sits alongside the incumbent model for a long time.

For a deeper review of methods, GO-Science also published a practical review of horizon scanning approaches covering tools and techniques used across government.

Which tools help with trend tracking?

Trend tracking tools fall into three groups: tools that measure attention, tools that cover early structural indicators, and tools that monitor your chosen sources over time.

Attention tools

• Google Trends. Free, and the standard way to see how search interest in a term changes over time and by region. According to Google's own documentation, each data point is divided by the total searches in its geography and time range, then scaled from 0 to 100. The numbers show relative interest, not search volume, so compare terms against each other rather than reading the values as counts.
• Exploding Topics. A paid trend discovery service owned by Semrush. It uses machine learning to detect growth across search, social, forums, news, ecommerce and podcasts, and human analysts review the results to filter out fads. It is most useful for consumer, product and software trends, and for finding topics you did not know to search for.

Structural indicator tools

• Preprint servers and research search engines for papers
• Google Patents or national patent office search for filings
• Job boards and hiring data services for new roles and skills
• Startup funding databases for rounds and new entrants
• GitHub and package registries for developer adoption
• Official registers, such as the Federal Register in the US, for proposed rules and comment periods

Monitoring tools

The structural sources are where early signals live, but checking them by hand every week is what makes most trend programs stall. A monitoring layer reads a fixed set of sources continuously and reports what changed. Teams that already track competitors often extend the same setup to trends; our comparison of market intelligence options for small teams covers the main categories. Kindal is one way to do this: you choose the sources, including research papers, patents, filings and specialist accounts, and it writes a short brief only when something changes, with every line linked to its source.

Worked example: tracking a hypothetical domain

Consider a hypothetical mid-sized company that makes industrial cleaning equipment. The strategy team wants to know whether robotic floor cleaning in warehouses is becoming an emerging trend worth a product line.

Define. Scope: autonomous cleaning in warehouses and logistics sites. Horizon: three years. Decision: whether to fund a prototype next planning cycle.

Sources. Robotics preprints, patent filings on autonomous cleaning, job postings for robotics technicians at logistics firms, funding news for warehouse robotics startups, safety standards bodies for mobile robots, and two operator communities where warehouse managers discuss equipment.

Baseline. The team records the current monthly level for each source: a handful of relevant papers, occasional patents, few job postings, rare discussion among operators.

What acceleration and convergence would look like. Over several months, the team would look for papers moving from navigation research to cleaning-specific tasks, a large equipment maker filing its first patents in the category, logistics companies posting roles that mention maintaining cleaning robots, and operators asking each other about vendors. If three or four of these move together, the signal has converged.

Validation. The team would check for pilots that were cancelled, identify the driver (labor availability and the falling cost of sensors, if the sources support it), and confirm that the job postings are not all from one company.

Decision. If the signals converge and the driver holds, the team funds the prototype. If only the funding news moves, the trend stays on "keep watching," with a written trigger: a standards body publishing safety guidance for the category.

The example is invented, but the structure transfers to any domain: name the decision, pick early sources, set a baseline, wait for convergence, and write the trigger.

What are the most common mistakes in tracking emerging trends?

The most common mistake is treating volume as momentum. A topic that dominates headlines for a week has attention, not necessarily a trend behind it. The other frequent errors:

• Watching only one source type. Teams that rely on search data or social media see trends late and confuse fads with change.
• No baseline. Without knowing the starting level, every mention looks like growth.
• Counting echoes as confirmation. Fifty articles rewriting one report are one signal, not fifty.
• Scanning only technology. Regulatory, social and economic shifts often decide whether a technology trend becomes real. PESTLE exists to prevent exactly this blind spot.
• Falling in love with a trend. Once a team has presented a trend, it tends to read new evidence as confirmation. Assign someone to argue the opposite case.
• No decision attached. A trend report that does not change a plan is reading, not intelligence.

How to keep trend tracking going

Trend tracking pays off over months, not in a single session, so the process has to survive busy weeks. Three habits keep it alive:

• A fixed review cadence. A short weekly scan and a deeper quarterly review, where each tracked trend gets moved to act, watch or drop.
• A shared log. One place where every signal is recorded with its source, date and the trend it relates to, so patterns become visible over time.
• Pruning. Remove sources that never produce useful signals and trends that have stalled. A smaller watchlist gets read; a large one gets ignored.

The goal is not to predict the future. It is to notice change early enough that your options are still cheap, and to know, each quarter, which trends moved and why.