<- Back to tech discovery guides overview

how-to-search-better-featured-image-1600-1067

How to get better search results for tech discovery

This article is about how to get better search results, specifically in a technology, business and innovation context. I call search in this context ‘tech discovery’. When you do tech discovery, your goal could be discovering emerging technologies, finding important industry players or R&D experts, or identifying new markets for your products and solutions, for example.

In this article I will show you how you can get better results for your technology, market, R&D, and innovation searches. I will show you how the cognitive technique of associating helps you discover things that are relevant but that you didn’t even know might exist. And I will describe how Mergeflow’s way of grouping, analyzing and tagging results helps you spend less time on making complicated search queries, and more time on actually discovering things instead.

For my video, “7 tips for writing good search queries”, I use hands-on examples from various tech sectors and explore them in Mergeflow. This is part of my video course, Tech Discovery for Innovators.

Because this article is hands-on, I recommend you get a Mergeflow account if you don’t have one yet. For example, I will often link to Mergeflow searches and results in this article, so that you can look “behind the curtain” of screenshots (and you need a Mergeflow account if you want to access these links). So while it is not strictly necessary in order to understand the article, it’s a good idea to create your own Mergeflow account (we have a 7-day free trial, no commitment).

Searching is not about what you want to find. It’s about where and how other people most likely talk about what you want to find.

A very common challenge with tech discovery

Our customers have various professional backgrounds, and they work across diverse industries and tech fields. Yet despite all these differences, there is one issue that we see very often:

Highly specialized technical people in particular tend to use very specific and technical search queries.

We all do this: If you have a lot of expertise in an area, using your expertise is like riding a bicycle. You don’t think about it, you just do it. This means that very often, you end up making search queries that are more specific than you realize.

There are two reasons why too specialized and technical search queries are usually not the way to go:

1. You miss things

If you use very specific search queries, you can miss highly relevant insights because your search query uses language that only few people use. For example, when people write about venture investments or other commercial activities, they often use less specific or technical language.

2. You focus too much on methods

Very specific, technical queries usually search for methods. But what if someone uses a different method to address the same problem that you want to solve? For example, let’s say you work in machine learning, and you find that your algorithms aren’t fast enough to solve your problem. So you want to find a way to accelerate your algorithms, and you start looking for “AI chips” (= specialized computer chips that are designed to accelerate artificial intelligence or machine learning applications). But what if somebody has found a way to accelerate machine learning not with chips (which are very expensive) but with software (which can run on cheaper standard CPUs)? You will miss this opportunity if you only focus on AI chips.

Related: Software may now be eating AI chips too

If you use very specific or technical search queries, you will likely find “more of the same”. Small variations on what you already know. But this really defeats a central purpose of tech discovery: You want to discover things that are relevant but that you didn’t even know might exist.

The purpose of tech discovery is to discover new methods, companies, markets and ideas, not just ‘more of the same’.

In order to help, we always recommend a method called “associating”. Associating helps you broaden your view, which dramatically increases your chances of discovering something really new.

In order to search better we should expand and iterate This increases our chances of discovering relevant findings that we didnt even know might exist Instead we often over specify and dig ourselves into a rabbit hole

Related: What makes a good innovation analyst?

Associating helps you discover new connections across seemingly unrelated questions, problems, or ideas

The truth is that the most interesting innovations often happen at the intersection of diverse disciplines and fields (see this article from Harvard Business School, for example). Of course, this is where you want to be because it’s where the better search results are. Associating helps you get there.

Related: How to use Wikipedia to boost your discovery skills

Probably the two most important methods of associating are expanding and zooming in. I’ll start with expanding because this is what you will probably need most often.

Expanding is about finding ideas, companies, people or markets that are relevant to your interests, but that do not explicitly mention your interests.

Keep your queries short and iterate

Let’s say you are interested in “in vitro methods to recreate human immune responses”. If you search for this exact string, you will probably not get any results. Here is a better, iterative, approach:

1. Start by searching for the essential part of your topic

The essential part of your topic is the one thing that you absolutely want to be there, in order for a finding to be considered relevant at all. In our example, “in vitro” could be this one thing (because if it’s not about in-vitro, we don’t care):

When you search, start off broadly. For example, only include terms that absolutely have to be there in order for a finding to be relevant.
When you search start off broadly For example only include terms that absolutely have to be there in order for a finding to be relevant
2. Scan the results for additional terms, then iterate

Now, scan the results. In Mergeflow, the 360° view is a good place to do this. The goal is to collect additional terms that you can use to either refine your search, or to start new tech discovery searches.

Check the ‘Markets’ section of the results

Very often, the ‘Markets’ section in Mergeflow has useful terms, particularly the “context markets” section. Context markets in Mergeflow are market estimates that came from a document that matched your query, but the market segment itself did not. Applied to our in vitro search, this would mean that Mergeflow found a document that mentioned “in vitro”, and contained, for example, a market estimate for the “bioprinting” market.

Below you can see the fastest-growing context markets for our in vitro search. Note that when you run this search, you might get different results because by that time there might be new findings.

Context markets are a good place to find additional terms for your search.
Context markets are a good place to find additional terms for your search

Depending on your interests, you can then add “bioprinting”, “organ-on-chips”, or “3D cell culture” to your search, like this:

You usually search better when you start off broad, and then narrow down, inspired by your broad search results.
You usually search better when you start off broad and then narrow down inspired by your broad search results
Look at the ’emerging technologies’ tag cloud for technology blogs

Another good place for finding additional terms is the “emerging technologies” tag cloud, for example (click on the screenshot to see it in full size):

Mergeflow discovers and tracks emerging technologies in its contents. These emerging technologies can help you make better search queries.
Mergeflow discovers and tracks emerging technologies in its contents. These emerging technologies can help you make better search queries.

Again, depending on your tech discovery goals, add one or several of these terms to your search query. If you use several, I’d recommend combining them by OR, not AND.

Starting off broadly and then iterating” has an important advantage: It helps you discover things that may seem obvious once you see them, but that you might not have thought of in advance.

Search for properties, functions and applications

You will miss out when you only search for technologies and methods

Above, I said that very specific and technical search queries often put too much emphasis on methods. If you search for methods and technologies only, this means two things:

  1. You miss out on findings that are relevant but that do not explicitly mention the method or technology that’s being used. This is often the case with venture funding events, market analyses, or more general news, for example.
  2. There might be methods and technologies that you didn’t think of. If you don’t think of them, you don’t search for them, so you don’t find them.

You can dramatically increase your chances of discovering new methods and technologies when you also search for properties, functions, and applications.

For example, assume you’re interested in low-cost water purification methods. Sure, you could tackle this topic by searching for water purification methods that you know are low-cost. But this would mean (1) a lot of upfront work, because you would have to find out which methods are low-cost; (2) if there is a new low-cost method that you don’t know, you won’t find it because you didn’t search for it.

How searching for properties speeds up your tech discovery

Rather than searching for specific methods, you will get better search results much quicker if you simply search for “water purification” AND “low-cost”:

Search better by combining what you want to do (e.g. water purification) with a desired property (e.g. low-cost).
Search better by combining what you want to do eg water purification with a desired property eg low cost

You can then zoom in very quickly on technologies, for example in the “Technology Licensing” section of Mergeflow’s 360° view (click on the screenshot to see it in full size):

Searching for broad concepts and properties helps you discover technologies that are relevant but that you didn't know might exist.
Searching for broad concepts and properties helps you discover technologies that are relevant but that you didn’t know might exist.

That’s it. Better than first spending hours on specifying an elaborate search query, right?

Better search results by zooming in on subtopics

I just did a very basic search for “smart city” in Mergeflow. This gave me ca. 2,500 science publications from the past five years. If instead I use Mergeflow’s pre-configured search for “smart city”, I’ll get even more, ca. 4,100 science publications (click on the screenshot to see it in full size):

A search for "smart city" in Mergeflow returns thousands of science publications (plus many other contents). Using the "most relevant applications" tag cloud helps you zoom in on subtopics.
A search for “smart city” in Mergeflow returns thousands of science publications (plus many other contents). Using the “most relevant applications” tag cloud helps you zoom in on subtopics.

Obviously I can’t read all these papers. So how can I make a meaningful selection of what to read, or where to start? Sure, scanning the recent publications is a reasonable place to start, but how can I do better than just that?

Related: How to use market estimates to structure a tech field (smart city)

Applications

Look at the “most relevant applications” tag cloud in the screenshot above. This tag cloud shows names of patent classes that Mergeflow assigns to its non-patent contents. We use patent classes because they are a widely known standard for categorizing tech contents, and very often they are a good proxy for applications. You can use these tags to zoom in on application subtopics. For example, when you zoom in on “traffic control systems”, results look like this:

Mergeflow automatically assigns patent classes to non-patent contents. This helps you identify and zoom in on subtopics of your general search.
Mergeflow automatically assigns patent classes to non patent contents This helps you identify and zoom in on subtopics of your general search

So even though you set out on your tech discovery endeavor very broadly, simply searching for “smart city”, Mergeflow’s metadata (patent classes in this case) help you get better search results very quickly.

Related: Using patent classes for concept search

People and organizations

Another way of zooming in is to first identify relevant people or organizations, and then start with reading their contents. ‘Relevant’ could be ‘publishes a lot’, or ‘collaborates with many other people or organizations’, for example. Mergeflow helps you discover both.

Let’s say I’m interested in CRISPR. Over the past five years, Mergeflow has ca. 29,000 science publications on CRISPR (click on the screenshot to see it in full size):

Over the past five years, Mergeflow has collected more than 29,000 science publications on CRISPR (only counting those publications that explicitly mention "CRISPR" or "clustered regularly interspaced short palindromic repeat").
Over the past five years, Mergeflow has collected more than 29,000 science publications on CRISPR (only counting those publications that explicitly mention “CRISPR” or “clustered regularly interspaced short palindromic repeat”).
Quickly finding researchers with the most publications…

Now, in order to find relevant people (researchers in this case), I zoom in on these science publications, and switch to a “Person” tag cloud. This shows me the researchers with the most publications:

Names of authors with the most science publications on CRISPR over the past five years.
Names of authors with the most science publications on CRISPR over the past five years

I can now simply click on one of these names, and see their publications.

…and researchers who collaborate with many others

Let’s go one step further, and discover researchers who collaborate with many others. In order to do this, I explore the authors’ co-publication network:

Co-authorship network for CRISPR science publications. Mergeflow uses co-authorship networks, not citation networks. This is because citation networks usually take years to grow, co-authorship networks are much more current.

This network gives me additional information, over “most published”. I can see who collaborates with whom. I could then zoom in on researchers who collaborate a lot (blue arrows). Or I can click on a connection between two people and see the publication(s) by these authors, for example.

Related:

Emerging technologies

As I said above, Mergeflow discovers and tracks emerging technologies, from across industries and tech sectors, in its contents. You can use this information for zooming in.

Start with the ’emerging technologies’ tag cloud for tech blogs

For example, for my search on LiDAR just now, I got 2,500 hits in technology blogs (click on the screenshot to see it in full size):

In order to search better, you can use Mergeflow's "emerging technologies" tag cloud to zoom in on likely promising findings.
In order to search better, you can use Mergeflow’s “emerging technologies” tag cloud to zoom in on likely promising findings.

I can now use the “emerging technologies” tag cloud to zoom in. On “LiDAR and machine learning”, for example:

Zooming in on tech blogs on "LiDAR and machine learning".
Zooming in on tech blogs on “LiDAR and machine learning”.
Then explore emerging tech in other Mergeflow data sets

I could then even switch to a different data set, via the “data sets” selector in the upper right:

In Mergeflow you can easily switch to results from different data sets.
In Mergeflow you can easily switch to results from different data sets

For example, I could switch to “LiDAR and machine learning” results from science publications:

Science publications on "LiDAR and machine learning".
Science publications on “LiDAR and machine learning”.

Notice what I’m doing here: I use the way Mergeflow groups and tags its results to quickly zoom in on relevant findings. But I do this as I go along, as opposed to mapping it out in detail in advance (and then probably ending up in a rabbit hole because my preconceived map does not align with what the “world of ideas, technologies, markets and companies” out there really looks like).

Related: Materials science and batteries: Emerging technologies in 2020

Mergeflow groups, analyzes and tags results, so you can get better search results with easier queries

Mergeflow organizes results by information type (venture fundings vs. markets vs. patents vs. science publications vs. news, and so on). And it analyzes and tags the results. These tags or metadata–company and people names, market segments, emerging technologies etc.–make your tech discovery searches much more effective.

Mergeflow’s result groups and metadata are a tool that helps you discover, both by broadening your view as well as helping you zoom in quickly on different facets of your topic.

In the previous sections of this article, I already used combinations of easy, almost generic search queries and Mergeflow metadata to get better search results. For example, I started a broad search and then used market segment names from the results to iterate and get more specific. Or I used names of researchers that Mergeflow identified in science publications, to prioritize the publications by the most relevant researchers.

But whereas my focus above was on the cognitive technique of associating, this section is more about ‘the mechanics of Mergeflow’, and how you can use them for effective tech discovery.

Combining simple queries with Mergeflow's result groups and metadata helps you get better search results more quickly.
Combining simple queries with Mergeflows result groups and metadata helps you get better search results more quickly

How to discover emerging technologies

“How can you discover emerging technologies and new companies in an established tech sector?”

A while back, we wrote about how we discovered new innovations in the area of shock absorbers. Now, when you look at mainstream industry news on shock absorbers, you will find, well, mainstream industry companies (click on the screenshot to see it in full size):

Mainstream industry companies in the area of shock absorbers.
Mainstream industry companies in the area of shock absorbers.

Of course, there is nothing wrong with this. But if you only had ten minutes, where would you rather look to find new innovations and companies: mainstream industry news or venture fundings?

Venture capital investments

If you go for option (2), looking at venture fundings, you increase your chances of quickly discovering new companies doing new things:

Venture-backed shock absorber companies. Notice that ClearMotion got two rounds of funding.
Venture-backed shock absorber companies. Notice that ClearMotion got two rounds of funding.

Related: The surprisingly interesting business of shock absorbers

Other Mergeflow data sets

Of course, venture fundings are not the only good place for discovering emerging technologies. For example, technology licensing opportunities are new proof-of-concept-level technologies coming out of universities and R&D organizations. Another good place are publicly funded research projects. In particular, I recommend checking out SBIR (Small Business Innovation Research) grants. SBIR funds individual companies, unlike for example EU CORDIS projects, which usually go to large consortia (which makes it difficult to discern who really did the innovative work).

Here is, as a diagram, Mergeflow’s heuristic for discovering emerging technologies:

A simple heuristic that dramatically improves your chances of discovering emerging technologies in Mergeflow.
A simple heuristic that dramatically improves your chances of discovering emerging technologies in Mergeflow

Just remember, it’s a heuristic. All of this is a question of probabilities, not mathematical axioms.

Related: How you can discover game-changing innovations

Find important industry players

“What companies are the important industry players in my topic?”

A great place to start looking for this is the ‘Markets’ section of Mergeflow’s 360° view.

For example, if you’re interested in agriculture drones, here is what Mergeflow extracts from market estimates:

Market information for the topic of 'agriculture drones'. Mergeflow extracts company names from market reports. 'Bigger font' means 'company is mentioned in more market estimates'.
Market information for the topic of ‘agriculture drones’. Mergeflow extracts company names from market reports. ‘Bigger font’ means ‘company is mentioned in more market estimates’.

‘Most relevant companies’ are those companies that market analysts mention in their reports, and that Mergeflow extracts from these texts.

Other good places to explore are patents (companies filing for patents) and industry news (companies mentioned there and extracted by Mergeflow). If your topic has clinical relevance, clinical trials are also a good place to look.

Here is, in graphical form, the heuristic for finding important industry players:

Our heuristic for easily finding important industry players in Mergeflow

How to find R&D experts

“Who are the most relevant R&D experts in my topic?”

I already talked about this above, in the section on zooming in on people and organizations. So we can keep things short here.

But besides scientific publications, there are two other data sets that are particularly good places for finding researchers: (1) technology licensing and (2) research projects. And, similar to industry players above, if your topic is medical, also look at clinical trials.

Mergeflow's heuristic for finding researchers in a topic.
Mergeflows heuristic for finding researchers in a topic

Identifying new markets for your products and solutions

“How can I discover new markets for my products or solutions?”

Let’s say you want to find out what relevant markets are for neuromorphic computing (= computing hardware, e.g. chips, that mimic aspects of how the brain works).

Markets relevant to 'neuromorphic computing'. 'Context markets' are a good place for finding markets that you might not have been aware of yet.
Markets relevant to ‘neuromorphic computing’. ‘Context markets’ are a good place for finding markets that you might not have been aware of yet.

I already talked about context markets in Mergeflow above, in the section on how you can iterate and expand your search. There, I used context markets to get ideas for additional search queries. But context markets are also very useful for identifying new markets for your products and solutions.

I highlighted in red some markets that I found interesting. For example, ‘digital utility’, ‘automotive V2X’, ‘AI in IoT’ and ‘AI infrastructure’ could all be applications, or markets, for neuromorphic hardware. And the ‘memristors’ market should be interesting as well because neuromorphic computing often uses memristors.

Of course, not all context markets are automatically new markets for your products and solutions. But once you see the list, you’ll be able to identify the interesting markets quite easily. The hard part is digging them out from the constant stream of news and other contents. But Mergeflow does this for you.

Related: How to find niche markets for new technologies