<- Back to tech discovery guides overview

Image by the CDC, from Unsplash

How you can discover game-changing innovations

This is not about how you can discover “the next killer app”. Instead, this article aims to help you discover game-changing innovations, or breakthrough innovations, that may have some substantial effect on your work. For example, they could help you make something that you otherwise couldn’t make. Or they could make something you do obsolete. And they are not somewhere off in the future, but they are happening here and now.

In order to discover such game-changing innovations, I will propose a divide-and-conquer approach to breakthrough tech discovery.

What do you mean by “game-changing”?

This article is not about how you can discover the next Google, Facebook, SpaceX, Spotify, etc.. Instead, it is about things that (at first) seem a lot more prosaic. But here’s the rub: typically, game-changing innovations happen gradually, then suddenly. We would like to discover them while they are still in the “gradually” phase.

First, let me explain what I mean by “game-changing” or “breakthrough”. In software engineering, there is a concept called “breaking change”. A breaking change is a change in one part of the system that potentially causes other components to fail. And this concept also captures what I mean by “game-changing or breakthrough innovation”:

A game-changing innovation does any of the following in your business: it makes something obsolete, it makes something work an order of magnitude better, or it enables something entirely new.

I added the “work an order of magnitude better” and the “something entirely new” parts.

Next, let’s take this definition apart, and look at it piece by piece. In this article, I will use the terms “game-changing innovation” and “breakthrough innovation” interchangeably.

A more detailed definition of “breakthrough innovations”

My definition above has four parts: (1) game-changing or breakthrough innovations change your business; (2) they may make something obsolete; (3) they may make something work an order of magnitude better; (4) they may enable you to do new things.

1. Breakthrough innovations change a part of your business

By “your business”, I mean your technologies, your products, services, go-to-market strategy, pricing, and so on. This includes technologies, products, or services that you use or buy in order to conduct your business. But importantly, it has to be your business, not “the world”, “humanity”, or “the way we live”. These last three concepts are too abstract, and we leave them to the pundits.

Also, game-changing innovations are not “just” about R&D. They can affect all parts of your business.

Breakthrough innovations can affect you in areas beyond R&D. This includes marketing, product design, strategy, and corporate investment, for example.

For instance, if some new tech makes a product easier to use, this will affect go-to-market and product design as well, rather than just R&D.

2. Game-changing innovations may make something obsolete

Very often, this involves some form of automation, such as software replacing manual tasks. For example, more and more manual parts of doing payroll, data analysis, consulting, or lab testing are becoming automated and hence obsolete.

But “obsolete” could also mean “become overwhelmed”. For example, let’s say you discovered a new lab automation technology that produces test results 10x faster. By itself, this is a massive improvement, of course. But this breakthrough automation tech may overwhelm your downstream R&D and business activities, if they are not also automated in some way.

3. Breakthrough innovations can make something work an order of magnitude better

Game-changing innovations can not just break things. They can also make things work better. By “work better”, I basically mean “make more from less”. In other words, in order to achieve the same or even better outcome, you have to put in a lot less resources.

A breakthrough innovation improves things by a factor of 2 or 10, not by half a percent.

For example, modern smartphones have very cheap camera lenses with lots of defects. But their software compensates for many of the lens defects. So the cheap-lens camera performs orders of magnitude better. The near-zero marginal cost of software duplication almost eliminates the very high costs of producing high-quality camera lenses that are made of expensive raw materials. Sure, making the software costs a lot of money, but not each time you ship a new smartphone.

4. Game-changing innovations may enable something entirely new

Not only may breakthrough innovations improve things. They may also enable entirely new things. For example, adding sensors or analytics to smartphones may enable new telemedicine applications.

“Enable something entirely new” could also mean “enable new business models”. For instance, if a mechanical device is enabled to collect data, this might enable data-driven business models. Further below, we will see an example of such business model innovation, in the area of shock absorbers.

A breakthrough innovation does not have to be a new technology. It could also be a new business model.

Now, with these definitions in mind, let’s move on to the next question. How can you discover breakthrough innovations in the outside world? I talk about “outside world” here, as opposed to “within your own organization”. “Within your organization” is interesting as well. But it is beyond the scope of this article.

A divide-and-conquer approach for discovering game-changing innovations

Where and how should you look if you want to discover game-changing battery tech innovations, for example? Should you look for the “wonder battery cell” that is 10x better than what we have now?

Or, to use a different example, what should you look for if your goal were “make machine learning 10x less energy-intensive”? By the way, there is strong motivation for looking for such solutions. Current machine learning approaches have a sizeable carbon footprint.

And then, when you find something, how can you make sure that this really is a breakthrough innovation? And not just an incremental improvement?

Let’s start with this last aspect, “incremental vs. breakthrough innovation”.

“Incremental vs. breakthrough innovation” may not really be the point

A lot has been written about “incremental vs. breakthrough innovation”. For example, here and here. I don’t argue with these definitions. But what if “incremental vs. breakthrough” simply isn’t the right contrast set?

For example, in his book, “How innovation works”, Matt Ridley argues that innovation is typically bottom-up and incremental. Yet his examples are all what I’d call breakthrough innovations; antibiotics, for instance.

How can this be? How can “bottom-up and incremental” result in “breakthrough”? And how does this impact how to best discover game-changing innovations?

Now, this is where “divide-and-conquer innovations” come in.

Divide-and-conquer innovations

Divide-and-conquer innovations combine “breakthrough” and “incremental”. The “breakthrough” part is that you set an ambitious goal. For example, your goal could be, “I want to make batteries that are twice as good as the ones available now”.

The “incremental” part says that you break down your goal into parts, rather than hoping or aiming for the one big breakthrough. In other words, you don’t try to discover the “wonder battery cell”. Instead, you split up the overall goal into parts. Ideally, these parts complement and compound each other. For a breakthrough battery improvement, these parts could be battery manufacturing, battery cell design, and material choice and production, for example.

“Divide and conquer” is how I would describe Tesla’s “Battery Day 2020”. In their presentation, Elon Musk and Drew Baglino (SVP of Powertrain and Energy Engineering at Tesla) presented how they plan to dramatically improve battery design and manufacturing. All the innovations they presented compound each other. If they all work out, the result will be a twice-as-good battery. And if only some work out, they will still achieve massive improvements over the current status quo. You can watch the presentation in the video below.

Tesla’s 2020 Battery Day presentation. Elon Musk and Drew Baglino present their steps toward dramatically improving battery design and manufacturing. This part of the presentation starts at ca. 59:24 in the video.

Now, what does all of this mean for breakthrough tech discovery?

Divide-and-conquer breakthrough tech discovery

When you do “divide-and conquer tech discovery”, you don’t look for the “wonder battery cell”. Instead, you look for several compounding innovations. In our battery example, you’d look for tech from across manufacturing, materials, cell design, and so on.

Here is another way to put this. You start off with the big goal. Next, you make a list of ingredients that are required to reach the goal. We talk about this “list of ingredients approach” in one of our other articles as well. In that article, we describe the “mind and hands” approach that good innovation analysts use.

Divide-and-conquer tech discovery looks for compounding innovations from across several areas, rather than for the one big breakthrough.

Advantages of the divide-and-conquer approach

At Mergeflow, we like the divide-and-conquer approach for the following reasons:

Searching becomes easier

If you tried discovering the “wonder battery cell”, what would you even search for? It is hard to come up with terms for something that does not exist yet. In contrast, when you break down the big goal into sub-topics, things become much more tangible and easier to verbalize.

Related: Wikipedia helps you make better search queries.

You discover things earlier

Above, I argued that game-changing innovations are the result of compounding innovations. This means that incremental innovations show up first. Then they compound each other. So if you set out to discover these incremental innovations, you can discover things earlier and more often. “More often” because you are looking for several things, not just one. Plus, these “several things” have a better chance of actually happening.

You reduce your risk of coming back empty-handed

In divide-and-conquer tech discovery, you don’t put all your eggs in one basket, that one basket being “the one big breakthrough”. If this one big breakthrough doesn’t happen, parts of it still might. And then you can at least discover these parts. And, as we argued in our article on how to beat analysis paralysis, 20% of something is still a lot better than 100% of nothing.

Now, let’s leave the “definitional phase”. Let’s look at some examples of breakthrough innovations discovery.

Examples of breakthrough innovations, and their potential impact

The game-changing innovation examples I describe below are not necessarily the stuff of front page headlines. But they may have profound impact within and adjacent to their tech fields and industries. This combination, not front page headline but potentially profound impact, means that they are easy to miss. This in turn means that a systematic approach to discovering them is even more important. We will get to this part, discovering them systematically, further below.

1. Making machine learning less energy-intensive

Let’s revisit our “make machine learning 10x less energy-intensive” goal from above. Divide-and-conquer suggests that we split up this overall goal. For example, we could look for more-efficient algorithms, advances in computing hardware, and so on.

Neural Magic is a potentially game-changing innovation here. Their software lets you train and run machine learning models on standard CPUs. Standard CPUs are a lot cheaper and need less energy than GPUs, which are typically used. You can read more about how we discovered Neural Magic here.

Screenshot of the Weekly360 tech discovery update from Mergeflow on "deep learning". Neural Magic, a potentially game-changing innovation in machine learning, was one of the featured  venture capital investments.
Screenshot of the Weekly360 tech discovery update from Mergeflow on deep learning Neural Magic a potentially game changing innovation in machine learning was one of the featured venture capital investments

Now, Neural Magic alone probably won’t get you all the way to 10x less energy. But it could help you achieve substantial performance benefits. This is the “make something work better” part. On the other hand, there might also be a “fail” part here. For example, if your business is “making GPUs for machine learning applications”, then Neural Magic could make what you do obsolete.

2. A potentially game-changing innovation in power electronics

“I want to improve the setup and operations effectiveness of my power grid by an order of magnitude.”

Let’s divide and conquer again. Here, our sub-topics could be innovations in power electronics components, power grid management software, IoT, and so on.

In one of these areas, power electronics components, Mergeflow helped us discover digital circuit breakers.

Related: Software is eating circuit breakers

A circuit breaker protects electrical circuits from damage (from short circuits, for example). Unlike traditional circuit breakers, a digital circuit breaker uses software and semiconductors. This gives them speed and flexibility. It means that they could replace meters, load controllers, demand management systems, and so on.

Atom Power digital circuit breaker, a game-changing innovation in the field of power systems. Mergeflow's tech discovery software helped us discover Atom Power. Image from atompower.com.
Atom Power digital circuit breaker a game changing innovation in the field of power systems Mergeflows <a href=httpsmergeflowcomproduct data type=page data id=28 target= blank rel=noreferrer noopener>tech discovery software<a> helped us discover Atom Power Image from <a rel=noreferrer noopener href=httpswwwatompowercom target= blank>atompowercom<a>

By thus putting multiple devices into one, digital circuit breakers could make the setup and operations of a power grid a lot more effective. This is the “make something work better” part. But if you make traditional load controllers or demand management systems, digital circuit breakers could make these products obsolete. There could also be a “new” part. This could be new, more flexible on-demand power supply pricing, for instance.

3. A new, digitalized, type of car suspension technology

“I want to read a book or watch a movie while riding in my car.”

If you want this, you will need autonomous driving because you can’t read and drive at the same time. Of course, “autonomous driving” is a whole group of topics in and of itself.

Furthermore, you will need something against motion sickness. After all, unless you are lucky, reading while driving will probably make you motion sick.

Regarding the “motion sickness” part, Mergeflow helped us discover something unexpected. Not a side-effect-free anti-nausea drug, but a car suspension technology.

ClearMotion makes a system that proactively counters the undesired motion caused by bumps in the road. If you suffer from motion sickness, this is likely a game-changing technology (“make something work an order of magnitude better”). You can watch the video below to see ClearMotion’s system in action.

See how ClearMotion’s car suspension technology can help you get less, or not at all, motion-sick. ClearMotion also makes RoadMotion. RoadMotion is a software platform that fingerprints road surfaces and stores the data in the cloud. This might be a game-changing business model innovation.

On top of their car suspension technology, ClearMotion also builds a new software platform. This software platform lets them collect very detailed road condition data. This makes ClearMotion a data company. So their innovation also enables something new. Now, regarding the “fail” part of breakthrough innovations: If you are a traditional shock absorber maker, or an anti-nausea drug maker, ClearMotion’s technology may make what you do obsolete.

Now we have seen some examples of potentially breakthrough innovations. Next, let’s look at where and how you can discover them.

Where can you discover game-changing innovations?

By “where”, I basically mean “where on the web”. This is not to say that other sources such as meetings or conferences are not relevant. But these other sources are outside the scope of my article.

Divide-and-conquer tech discovery has implications on where to look for game-changing innovations. When you divide a big goal into subtopics, these subtopics usually differ in nature. They span across different technology fields, and some are more “technical”, whereas others may be more business or application oriented. For instance, our battery tech example above included materials science, manufacturing methods, and others, thus spanning different fields.

Now, because your topics span across tech fields, and from R&D to business, you need to look at various data types. This includes data from R&D, venture investments, patents, markets, and others. All these data sets serve different purposes and complement each other.

So as a next step, let’s look at how you can use all these different data for breakthrough tech discovery. The data sets I discuss below are all available in our software.

Venture capital investments

Venture investments are great for discovering market-relevant innovations early on. This includes technologies as well as business model innovations.

It’s also interesting to see who the investors are. For example, they might be the corporate venturing arm of one of your competitors. Or they might be unexpected investors. For instance, ClearMotion also received an investment from Qualcomm Ventures, not just from automotive-related companies. Of course, given ClearMotion’s data platform activities, this makes sense. But you actually have to dig a little bit to find out about ClearMotion’s software platform. So the Qualcomm investment is a useful signal here. It triggers you to do this digging in the first place.

Market research

By “market research”, I mean publications written by analysts about market segments and important players in these segments (companies usually). This research also includes estimates of market segment sizes and growth rates.

At Mergeflow, we differentiate between direct markets and context markets. We describe this in more detail in our technical knowledge base, how to do a 360° search with Mergeflow. For example, if you were interested in edge computing, the “edge computing market” would be a direct market. In contrast, the “collaborative robotics market” would be a context market. A context market is not about your technology or topic per se. Rather, your topic plays some role in the market. In our example, “collaborative robotics” is an application of “edge computing”.

Context markets are particularly useful for discovering new applications (in another article, we wrote about how Wikipedia can help you identify applications). For example, let’s say that you monitor “edge computing”. If then the first market report on the “collaborative robotics market” comes up in this context, this could indicate that “collaborative robotics” is now a new, market-relevant application of “edge computing”.


Patents do not necessarily have a reputation as an early indicator. It can take up to 18 months or so for a patent to be published after it has been filed. How long it really takes depends on many factors, such as the jurisdiction where a patent is being filed. But you cannot count for something to be filed as a patent and then be published right away.

Still, patents can be a valuable resource for breakthrough or disruptive innovation discovery. For instance, if a company wants to ensure freedom to operate in an area, they might file patents. But they will probably not publish anything else in this area because they do not want to alert their competitors. We describe an example of this scenario in another article, “Discovering strategies in additive manufacturing”.

Scientific publications

Scientific publications provide very early signals of new technological developments. Some might even say “too early”. If your topic shows up only in scientific publications but nowhere else, this may indicate that your topic is in a very early stage of technology maturity.

So how much effort should you put into analyzing and monitoring scientific publications for breakthrough innovations? I don’t want to give a one-size-fits-all answer to this. But I would probably spend more effort if it is relatively easy to verify the new technology. For example, if the topic is a new algorithm, it might be relatively easy to implement and verify this new tech. In contrast, if the topic is a new type of fuel cell cathode, verifying this might require a lot more effort.

Technology blogs

Scientific publications can be difficult to assess if you are not an expert in a field. But technology blogs can be a good bridge between “smart non-expert” and “subject matter expert”.

Related articles:

This “bridge property” of tech blogs makes them interesting for divide-and-conquer discovery of breakthrough innovations. After all, it is quite likely that you have more expertise on some sub-topics than on others. Consider our battery tech example. If you are the materials but not the manufacturing expert, tech blogs could help you get insights into manufacturing innovations.

Technology transfer from R&D organizations and universities

Many universities and research institutes have technology transfer offices. These offices are responsible for commercializing (some of) the research from their organizations. Typically, this is done via some kind of licensing agreement between the tech transfer office and a company that commercializes the technology.

Ideally, offerings from tech transfer offices are not just very early-stage technologies. They also have been vetted, to some extent, with respect to their practicability. Both aspects make tech licensing offerings attractive for breakthrough tech discovery.

Mergeflow aggregates tech offerings from across several hundred research organizations worldwide. This saves you the trouble of having to search all these databases individually.

Public R&D funding

Public R&D funding often goes to innovative, early-stage companies. For example, it is not uncommon that a company receives public R&D funding before they receive venture capital funding. This is because VCs typically invest a bit later in the life cycle of a technology. They usually require some proof of commercial scalability of the new technology.

There are several organizations worldwide that provide such funding. In Europe, this includes EU CORDIS. In the US, these organizations include NIH, NSF, and SBIR.

Some of these funding organizations make breakthrough innovation discovery easier than others. For example, SBIR typically funds individual companies. This makes it a lot easier to pinpoint where an innovation comes from. In contrast, EU CORDIS typically funds large consortia of companies and research organizations. This often makes it difficult to discern who really does the innovative work in the consortium.

Clinical trials

Clinical trials are a good place for discovering game-changing innovations that in some way serve a medical purpose. Increasingly, these innovations include not just new medical treatments or highly specialized devices. They also include more and more consumer devices such as wearables.

Related: Clinical trials on nutrition, wearables, and lifestyle.

Clinical trials are classified into phases. This makes it easy to assess the technological maturity of a candidate breakthrough innovation: clinical trial phases very closely correspond to Technology Readiness Levels (TRLs).

A method for discovering game-changing innovations

Now let’s put this all to practice. Below I describe a tech discovery method that can dramatically improve your chances of identifying breakthrough innovations early on.

Please bear in mind that you may have to adjust the method to your specific needs. What’s below should not be seen as axioms, but rather as guidelines.

Mergeflow's tech discovery software includes semantic models of more than 200 topics from across various industry and technology fields. These topics can provide input for the "divide your goal into sub-topics" step described below. The models for these emerging technology topics are built into our software. You can also preview them in our website.
Mergeflow’s tech discovery software includes semantic models of more than 200 topics from across various industry and technology fields. These topics can provide input for the “divide your goal into sub-topics” step described below. The models for these emerging technology topics are built into our software. You can also preview them in our website.

So here is our method, which we also cover in Mergeflow’s knowledge base:

1. Write down your big goal
   - Writing down the goal enforces clarity.
   - Easier to share with and get feedback from colleagues.
   - Try to keep this short (2-3 sentences).

2. Divide your goal into sub-topics
   - A good tactic is using first principles thinking here.
   - Refer to Mergeflow's set of emerging technologies for additional input.

3. Do 360° searches for each of your sub-topics
   - Explore the data sets described in this article.
   - Refer to our 360° search checklist for guidance.
   - If you work in a team, use collaborative discovery.

4. Evaluate your findings
   - If you find a company:
      - What can their technology do?
      - What is their business model and their pricing?
      - Can you find anyone there to contact?
   - If you find a technology (e.g. a scientific publication):
      - How could you implement the technology in your context?
      - What can the technology do in your context that you cannot do?
   - If you find a business model innovation:
      - Would services have to be "productized" in order to make the model work?

5. Follow new developments within and across your topics
   - Include topics that did not return much in your 360° search because...
      - ...they may become more active in the future.
      - ...they could shift to a higher technology maturity level.
   - You can use Mergeflow's Weekly360 email update service to automate this.