Diffusion of Innovation

adopter types crossing the chasm diffusion early group imitators influence majority mavens moore bass people rogers social network tipping point

The Big Picture…

Innovations diffuse over time – knowledge of them spreads – across a social system through communication channels. And this social system, or a subset, is your target market.

We can build up our view of diffusion by understanding Bass’ model (innovators and imitators), Rogers’ adopter types (innovators, early adopters, early and late majority, and laggards) and Moore’s crossing the chasm.

We’ll find that one adopter type is rarely influenced in the same way as the group to their left. So that tells us we need to change approaches to diffuse across each type. Maloney suggests we need to change the message and delivery method after 16% have adopted (where Moore’s chasm starts). Since the first 16% crave scarcity and those after look for social proof.

Moore proposes we need to build out from a niche beachhead with additional niche markets to cross his chasm. And Gladwell’s connectors, mavens and salesmen reminds us diffusion is across a social network. So can we find and leverage the influencers? Can we get the innovation to go viral?


Understanding how diffusion works lets us plan how best to speed up awareness of our innovations. Why finding influencers is useful. How do we attract those innovator types, and then switch messaging for the imitators. And being aware that, at least for technology innovations, there is a particular chasm to cross to keep diffusion happening and where that is.

The Idea

So you’ve managed to turn your idea into an innovation. It’s an innovation that that enables beneficiaries to make progress with some aspect of their life better than they can currently (see an evolved definition of innovation). You’ve addressed points that will speed up innovation adoption. And you’ve minimised innovation resistance.

All you need now is for beneficiaries to become aware of the innovation, and surely they will start using it.

Awareness comes as your innovation diffuses across a social network. That is to say, knowledge of ideas passes from person to person. And that happens both actively and passively. Active ways could be through word of mouth, advertising, social influencers, etc. And passive ways are, for example, you observing someone else using an innovation and want to find out more.

In this article, we’ll explore the background to diffusion. And in particular, I’ll take a network-first approach. You might also be interested in how understanding network topology (shape, characteristics etc) might help us. Or taking a deeper dive into the mathematics behind diffusion (go on, you know it will be fun….).

So, let’s get going with the classic definition of diffusion.

The classic definition of diffusion

Rogers gives us the standard definition of diffusion in his book “The Diffusion of Innovation“:

Diffusion is the process by which an innovation is communicated through certain channels over time among the members of a social system.

Rogers – Diffusion of Innovation

And depending on how you want to look, your target market is either that social system or a subset of it.

The social system as a network

In practice by social system we mean a collection of people, or organisations, who communicate with each other – they are connected. Such as in Figure 1.

Figure 1: A network, with nodes and connections

The people/organisation are the nodes. And how they are connected are the channels.

Knowledge of your innovation gets injected, somehow, into that network. And it spreads via various channels. People talk to people; some other people write reviews that are read by others; organisations attend conferences; people move from organisation to organisation, market to market, industry to industry. You hopefully get the picture.

Now in any network we will see a set of characteristics. For example, some nodes may be connected to very few other nodes. Diffusion risks stopping with them. Whilst some other nodes may be connected to nearly all the others. These might be influencers. You can see some typical types of networks in Figure 2 and think through how they might help/hinder diffusion.

Figure 2: Some Example Network Topologies

Knowing the structure of your social system can help in planning how to inject knowledge of your innovation. And where there might be future diffusion challenges. I look more into this over in this article.

Relationship With Adoption

There is a relationship between diffusion and innovation adoption.

Adoption – getting a new idea adopted, even when is has obvious advantages, is difficult

Rogers (2003) – Diffusion of Innovation

In fact the two are very closely related. If an innovation doesn’t get adopted then diffusion comes to a halt as no-one sees it in use or reads reviews etc.

It turns out that diffusion across members of a social system (network) happens in a surprisingly predictable manner. And our first insight is that some members are innovator types and others imitators.

Innovators and Imitators diffuse innovation

Bass, in the 1960s, observed there are two types of people/organisations in a diffusion network. Some, the innovators, are eager and actively seek out and adopt innovations. Others act more passively, and he called them the imitators.

Visually we can split out social network into the two parts in Figure 3.

Figure 3: Social network as a diffusion network

(Now, a quick confession: in this diagram, and later ones, I am simplifying things by pretending diffusion and time go from left to right across the network. In practice, the social system is more complicated. But this works well for clarity).

Bass created a diffusion model – a mathematical formula that predicts how diffusion will work. And it shows this split between innovators and imitators very nicely. His model states:

The probability of adoption at time t given that adoption has not yet occurred is equal to: p + (q * cumulative fraction of adopters at time T).

Majahan, Muller, Bass (1995) “Diffusion Of New Products: Empirical Generalizations And Managerial Uses

And as always, a picture (Figure 4) is worth more than a thousand words.

Figure 4: Bass saw that adoption of innovation started with innovator types in a network who are overtaken by imitator types as time progresses

Let’s unpack Figure 4 a little. Firstly, early in the diffusion timeline, it is mainly the innovator types that are adopting. But quickly, imitators take over. And it is drawn in such a way so that we can see the number of innovator types is typically lower than imitators. What are these two types?

Some people in your network are imitators…

Remember the last time you discovered something new and innovative. Most likely, a friend or colleague introduced it to you. You are, according to the early literature, an imitator. Or as later researchers put it, in a nicer way: you need to experience the innovation to be comfortable with it. And psychologists would say you need to have social proof.

From Bass’ perspective, you had an internal influence acting on you. And, this was probably word of mouth, or perhaps you saw the innovation in use.

Some people are innovators…

However, I’m sure you can think of examples where you were the first to find and use an innovation. Often you’re connected to the initial innovator, or you are actively searching for a solution to a problem you have.

Being an innovator means you typically have had an external influence (to the network) acting on you. Which could be advertising, you searching, or connections with other social systems (such as the innovator).

Let’s take a look at those p and q values, which are known as the bass coefficients.

Bass coefficients

We use the coefficients, p and q, in Bass’ formula to represent the size of innovators (p) and imitators (q).

Look again at Figure 4. It actually exaggerates the size of innovators to make them visible. How big are these two groups in real life? Well, studies have shown that the average value for the coefficient of innovation – p – is 0.03. And, the average coefficient of imitation – q – is 0.38. So Bass’ formula generally looks like this:

0.03 + (0.38 * cumulative fraction of adopters at time T).

It’s interesting to note that other studies have hypotheised these co-efficients are likely to be higher in the modern digital world. With imitation, q, being higher due to lower costs of communication and higher connectivity. And p being higher due to ease of discovery and trial. However, historically Bass’ model has proved remarkably accurate to real world measurements.

If you’re into the maths behind Bass’ model, then I look at the formula in some more detail over in this article. And that also includes a link to an Excel version of the model to play with.

It turns out, though, that viewing our target market as just two groups is not sufficient to fully understand the challenges of diffusion. To help us, we can turn to Rogers’ five adopter types.

Rogers’ Five Adopter Types

Rogers’ book Diffusion of Innovations is the classic reference on diffusion theory. It has several useful and valuable insights. And the one we pick up on here is his identification of five adopter types: innovators, early adopters, early majority, late majority and laggards.

We can almost see the innovator and early adopters as Bass’ innovators. And the remaining adopter types as Bass’ imitators. So we can update our view of the network to that shown in Figure 5.

Figure 5: How Bass’ innovator/imitator insight relates to Rogers’ five adopter types

And, whilst Rogers’s call these adopter types, they also affect diffusion. Let’s explore these types, starting with the innovators.


Rogers found that a group of people he called innovators are the ones most likely to have your new innovation diffused to them (and to adopt). These people are willing to take risks and are typically closely related to the industry producing the innovation and/or the innovators.  About 2.5% of your target market are these innovators. And it is through this type that knowledge of the innovation gets injected into the network.

Early Adopters

Around 13,5% of your target market are also risk-takers, like the innovator group. However, they will wait for your innovation to get some market hold before adopting.  These are your early adopters. And they are usually closely connected to the innovators. Often we can see early adopters as leaders/thought influencers. And they bridge the gap between knowledge coming into, and knowledge spreading within, the network.

The Early and Late Majority

After the early adopters, there are the two largest categories of adopters. First, the early majority who are 34% of your target market. And second, the late majority, a further 34%. 

The average customer is in the early majority

Those in the early majority often have contact with the early adopters but are choosing to adopt much later. They are needing some proof of the innovation before adopting. And this group includes your average customer. The late majority need to overcome a degree of scepticism before adopting. As such, they are doing so after the average customer has.


Finally, we find the hardest group to persuade. Making up 16% of customers, these are the laggards. These customers are typically resistant to change and hard to get on board. Whilst laggards are in your social network, they might not really be your target market.

One observation Moore makes (which we look at lower down) is important to note here:

…any of the adopter groups will have difficulty accepting an innovation if it is presented to them in the same way as the group to the immediate left.

Moore, Crossing the Chasm

That is to say, we can’t present the innovation to the early majority in the same way we have to the early adopters. And Maloney further uses this in his 16% rule (which we’ll also look at later).

We often look at Rogers’ adopter types through his adoption curve.

Rogers’ Adoption Curve

We can plot the adoption rate through the social system over time (see Figure 6). This curve is known as Rogers’ Adoption Curve. And it shows the above percentages of the target market predicted for each adopter type.

Figure 6: Rogers’ five adopter types, when they start adopting in time, and what percentage of your target market are in each adopter type.

Rogers’ set the size of each adopter type based on standard deviations from his empirical observations of the literature. And as a rule of thumb, they are good to use. What is really interesting is Bass continued work shows results that are very close to Roger’s curve.

Relation of Rogers’ adoption curve to Bass’ Diffusion model

Rogers’ adoption curve allows us to understand adopter types in a broad manner. What I find quite fascinating is Bass’ 1969 mathematical treatment of diffusion. It builds on his 1962 imitation model, and, I give it an article all to its self.

For now, we can note that Bass’ model produces curves remarkably close to Rogers’ adoption curve. And the impressive part is that by varying the coefficients, you get remarkable mappings to real data.

Just look at Figure 7, which shows real sales volumes of various technologies with the adoption curve overlaid.

Figure 7: Bass’ diffusion model compared to real-world data (non-animated)

One outcome of Bass’ work is we obtain better size values for Rogers’ adopter types than standard deviations. Though I have to say, Rogers approach of using standard deviations was pretty close. For example, we find innovators make up 0.8%-2.8% where Rogers said 2.4%.

But, before we get too happy, there is a bias problem when it comes to innovation.

Pro-innovation Bias

A challenge with diffusion is that we believe innovations are always good. There is a pro-innovation bias.

…an innovation should be diffused and adopted by all members of a social system, that it should be diffused more rapidly, and that the innovation should be neither re-invented nor rejected.

The Diffusion of Innovation, Rogers.

What we tend to believe is that once a node in the network hear’s about an innovation they will adopt and they will diffuse the innovation further. This is what we see with viral infections, such as the spread of COVID-19 across the globe in 2019/21.

The bias means research focuses on adoption rather than innovation resistance. But, diffusion can fail. And Moore points out it risks doing so more at the boundaries between adopter types than elsewhere (especially for high tech innovations).

And, he identifies that the most significant instance of this is between the early adopters and the early majority (for high tech innovations). Welcome to Moore’s chasm.

Crossing the Chasm – a diffusion problem

One reason that diffusion can fail at the boundaries between adopter types is that internal influences work better between people of the same kind. Moore, with his marketing background, highlights that:

…any of the adopter groups will have difficulty accepting an innovation if it is presented to them in the same way as the group to the immediate left.

Moore, Crossing the Chasm

It turns out for high-tech innovations the gap between the early adopters and the early majority is considerable. It is here where the balance between external and internal influences shifts. And where more pragmatic consumers replace risk-takers.

Moore saw this gap as so considerable he called it a chasm. And his book “Crossing the Chasm” describes why this chasm occurs and ways to cross it. He also updates Rogers’ adopter type names with his own. For example, the early adopters become the visionaries. See Figure 8.

Figure 8: There is a gap we must cross between each of Rogers’ adopter types. Moore noticed that for technological innovations, the gap between early adopters and the early majority was quite considerable. He named this gap, the chasm.

Moore sees the pragmatic early majority are not keen on listening to the visionary early adopters. Pragmatists see visionaries as:

  1. lacking respect for colleagues’ experiences.
  2. taking a greater interest in technology than in their industry.
  3. failing to recognize the importance of existing product infrastructure.

Additionally, he sees pragmatists as being concerned with the overall disruptiveness of high-tech innovations.

From a network perspective, the information about the innovation is not efficiently passing across the communication channels (Figure 9). And so, diffusion is at considerable risk of stopping.

Figure 9: The chasm as seen in a social network

Moore suggests several approaches to help cross the chasm using analogies of bowling alleys and tornadoes. We will come to this shortly. But first, let’s look at the insight from another marketer.

Maloney’s 16% Rule

Whereas Moore attributes the chasm to the early majority not trusting the early adopters. Maloney thinks the chasm is due to the early adopters not wanting to share what they have discovered.

Malony takes the six principles of persuasion introduced in the book “Influence: The Psychology of Persuasion“. From there, he identifies that innovators and the early adopters crave scarcity. They want things others don’t have. The early majority, on the other hand, desires social proof. They want what they see others have and use. If this social proof idea sounds familiar, we talked about this earlier when looking at imitators.

Figure 10: Maloney’s 16% Rule – messaging and media needs to change between the first 16% adopters and the remaining (non-animated)

Given the different cravings, and reflecting Moore’s view that no group is persuaded in the same way the group to the left is, Maloney came up with his 16% rule. And this states:

Once you’ve reached 16% adoption of any innovation, you must change your messaging and media strategy from one based on scarcity to one based on social proof, in order to accelerate through the chasm and the tipping point.

Maloney’s 16% rule

Maloney effectively suggests marketing to the first 16% (the innovators and the early adopters) should emphasise the scarcity and be through PR or other more exclusive means. To cross the chasm, and reach after the 16%, the marketing needs to change to emphasise social proof and to be through mass media to drive that “I need it too” feeling.

If we can get over the chasm, and avoid the 16% issue, then the next aspect we are looking at is hitting the tipping point.

The Tipping Point and diffusion

At a point in time when you are in the early majority, diffusion (and adoption) of your innovation accelerates. This point is the tipping point.

Figure 11: The tipping point is where your innovation really starts to take off

Both Moore and Gladwell consider this tipping point. Moore’s “Inside the Tornado“, talks about high-tech innovations being in the bowling alley, the tornado and eventually on Main Street. Gladwell’s “The Tipping Point” considers this from an epidemic perspective and concludes the need to find specific types of people to accelerate through the tipping point: connectors, mavens, and salespersons.

Bowling alleys, Tornados and Main Street

Moore suggests creating a complete niche product that fulfils the needs of an initial beachhead of customers in the early majority. At this point, you have the first taste of the mass market. You are now into the bowling alley. Here you need to keep knocking down more and more niche markets by extending your product. Now you are establishing yourself as having a product suitable for the mainstream. You are building a reputation and forging alliances.

At the next point in time, you hope to be pulled into a tornado of a host of pragmatists buying your product. Now you are in Moore’s tornado.

But, there is a strange thing with the tornado. Your strategies need to be opposite to those applied in the bowling alley. Now you have to focus on generic product, mass marketing, commoditization etc.

Moore’s approach is product/market focused. Gladwell takes a more network-first view.

Connectors, Mavens and Salesmen

Gladwell’s “The Tipping Point” circles us back to network structures and metrics. He identifies, amongst other things, a law of the few. That in any social network there are three key actor types:

  • Connectors – are connected to many people (i.e. have high degrees of centrality) and excel at linking people together, including across different social networks (they cross structural holes and often have weak ties)
  • Mavens – we rely on these people to connect us to new information. Gladwell sees then as “starting word of mouth epidemics”. A modern interpretation could see them as social media influencers.
  • Salesmen – these are the networks persuaders.

Ideally, you should identify the connectors, mavens and salesmen in the network of your target market. And then determine the best way to persuade then to diffuse. Doing so should accelerate diffusion.

Adding weight to this, Keller and Berry’s book, “The Influentials“, suggests that “One American in ten tells the other nine where to shop and what to buy”.

Is Gladwell correct?

However, one researcher pushed back against this in 2008. He revisited the early experiments that informed Gladwell’s book and found different results. Finding that in most cases a normal person, rather than influencer, was responsible for the tipping point.

A trend’s success depends not on the person who starts it, but on how susceptible the society is overall to the trend – not how persuasive the early adopter is, but whether everyone else is easily persuaded.

Is the tipping point toast?

Despite this, spending on social media influencers keeps growing.

I’ll leave it to you to decide if this means influencer marketing is actually getting better results. Or if we as a population are becoming more easily persuaded.

Wrapping Up

We’ve seen that diffusion is spreading the message of an innovation over channels between entities in a social system over time. That social system is a network of your target market. A small part of the network contains innovator types. They are searching for solutions to problems and are influenced from outside the group. A more significant number observe the innovation in use and start using it themselves. These are imitators or known as those requiring social proof. They are influenced internal to the group through, for example, word of mouth.

The whole network can be better divided into adopter types: innovators, early adopters, early and late majority, and laggards. Remarkably, we can predict the size of these types. But there is a gap to cross between each type as they have different values. And we saw that we need to alter our communication approach to each group since any particular type will not be persuaded by the techniques used by the group to their left.

In high tech, there is a big gap between the early adopters and early majority. The latter do not trust the former. And the former does not want to give up the scarcity that they value to the masses. We have to change our message and channel between the first 16% of adopters and the rest.

Co-create value by leaving your views

This site uses Akismet to reduce spam. Learn how your comment data is processed.