Featured

How AI changes industries — from efficiencies to disruption

An organization is nothing more than a string of components held together by transaction costs (1). When technology and especially AI influences these components and costs then companies and industries change, irreparably (2).

Let’s take a common example. What are the components needed to produce and sell bound encyclopedias?

The components in the value chain that make up the encyclopedia industry (simplified and illustrative).

Transaction costs are costs related to the coordination between two components (e.g. finding and negotiating with suppliers, coordination and communication, monitoring, technology etc.), or coordination and communication inside components (e.g. scheduling, synchronization, compliance, handling delays etc.).

A company exists because it can reduce transaction costs compared to society in general (e.g. Encyclopedia Britannica would have a more efficient system for producing an encyclopedia than most of us), and it can excel at certain components in order to achieve a competitive advantage.

Darker components indicates importance to achieve competitive advantage.

Customer value

But, that is not all. A company does not achieve success just because it excels at manufacturing, distribution, marketing or sales. It also needs to be preferred amongst customers by being better at delivering to their need.

If we take the same encyclopedia example, what do people value when it comes to their encyclopedias?

Customers don’t buy encyclopedias because of an encyclopedia company’s distribution, sales and marketing setup. They buy it because it offers them relevant quality information.

What we can see from this model (above) is that what determined success in the encyclopedia industry was different from what was important to the customer. Which is an indication of a product centric industry which has turned its attention away from its customers and towards its own transaction costs.

This makes a company ripe for disruption (7).

What happened to the encyclopedia industry?

Wikipedia happened. And the big problem for the industry as with many other industries is that since they understand their industry through the lens of their chain of components they can’t see that they are being disrupted by a company that offers the same customer value through completely different means.

“Uber, the world’s largest taxi company, owns no vehicles. Facebook, the world’s most popular media owner, created no content. Alibaba, the most valuable retailer, has no inventory, and Airbnb owns no real estate” — Tom Goodwin, 2015 (3)

So what did Wikipedia do?

First of all Wikipedia is free and online. which means there is no distribution or sales & marketing component in the Wikipedia chain. These were the two most important components in the old industry value chain and when you remove them completely there is very little competition left (2).

Secondly Wikipedia delivered better customer value. People care about the content, and with Wikipedia the content was first proven to be of satisfactory quality (4), but it was also continuously updated and easily searchable serving additional behaviors and needs of its readers and writers (Wikipedians).

The wikipedia value chain upended the old industry by making former core components irrelevant and over-delivering on what customers cared about.

Wikipedia outperformed the traditional industry in almost every way that mattered, while making less relevant what the industry had got stuck on.

What is the lesson here?

The lesson is that as companies and industries grow up they turn their attention away from what is important to their customers and towards whats important to themselves (5).

Their customers are perceived as too messy, fussy, human, complex .. , while transaction costs become easy to track, measure and manage. What’s easy to measure becomes important while what can’t be reduced to a simple number gets ignored.

Summarizing Clayton Christensen at Startup Grind Global.

Focusing on your own components and transaction costs works as long as customers have no other choice. But once there is choice industries are in trouble.

e.g. the financial technology [Fintech] startup scene didn’t disrupt the financial industry back in the 2010s because they didn’t offer any meaningful choice to customers. They made some improvements to the transaction costs, but customers didn’t care. Until someone can come up with something meaningfully different people will be using their old banks.

Compare this to the entertainment industry where companies like Netflix, HBO, Hulu, Prime threw a holy hand grenade (6) [maximum damage] into the old TV business model offering people the opportunity to find high quality content at their own leisure and to their own taste.

Part of an old slide made by the author reflecting on how industries won’t “disrupt” unless customers are offered a real choice. Innovation is not about technology, but the customer outcome.

This can also be said for the formerly mentioned examples: Uber, Facebook, Alibaba, Airbnb etc. They are all ignoring the components and transaction costs of the legacy business models, but improving on delivering customer choice in what customers value.

What has this got to do with AI?

AI is promising a more fundamental change to our world, business, personal and social life than any technology before it. The Internet, social media, mobile, data, they were all big, but AI promises to be bigger.

So then it’s not only a question of: how can I use AI to reduce my transaction costs (e.g. produce content, schedule meetings, automate decision making or make data correlations).

The question is when will a company who can see the customer value and is willing to ignore the value chain throw a holy hand grenade into a legacy industry and make that industry’s business model redundant?

That is the promise of AI, and that is why AI-projects need to come with a more ambitious approach than just being a technology, efficiency or data project. It’s not about the tool, but about the outcome.

AI needs to come with a more ambitious approach

How do we solve it?

Start with the customer. Go back to why we are doing what we are doing in the first place: what are the desired customer outcomes that drive our own desired business outcomes?

A company produces impact by delivering value to the customer enabling a behavior driving value back to the business.

With the outcomes in place map the chain of components that are needed to offer the customer the value they desire (I’m going to use a Strategy Consulting example for reference below).

Simplification of value chain for strategy consulting deliverable.

For each of these components simply start with these questions:

1. Which components are more important to the customer? Why?
2. Which components represent our competitive advantage? Why?

These questions will help the team discuss if their current deliverables align with what the customers wants to achieve and value.

In this example the same chain of components would lead to the same client desired outcome where strategy development and implementation support is most important to the client. The visual shows three different types of consultancies with three different strengths (darker components).

Next ask: what are the transaction costs and how could any of the components be improved with AI?
This is the efficiency question.

Then finally ask: keeping only the components important to the customer how can we reimagine the entire chain with AI reaching the same outcome to the client?

This is a simple approach to start a conversation looking for how and where industries are changing, the impact of AI and where to focus interests and investments.

An alternative approach:

There is an alternative approach using the Wardley Map (6) that helps the organization better discuss the impact of AI on its current chain, but it’s not as simple in terms of identifying new chains and components.

The Wardley map has an y-axis indicating closeness to the customer (higher) and infrastructure (lower). And an x-axis indicating innovative components to the left while commodity components are put further to the right.

In this map (above) I’ve added the desired client outcome: Competitive Advantage at the top as it is the outcome an imagined client desires.

The purple arrows indicate the discussion in the team regarding the position of each node. Do we think it will become more or less important to the customer? Will it need to be more innovative or will it drop towards commodity?

Then I’ve mapped the consultancy component value chain to the board including importance to the client for each component (more important means higher on the board) while our uniqueness in the offering is moved to the left (innovative) or right (commodity).

With a map of the current state the team can discuss if components need to be moved up or down in terms of how important they will be to clients in the future, or left or right depending on how AI (or other influences) will change their needed degree of innovative or commodity.

Different consultancies could use the map to demonstrate and discuss their different strengths and assumptions about the market (what their type of clients find valuable).

With the Wardley Map it is also possible to discuss and draw completely new chains keeping the desired customer/client outcome at the top and mapping new ways to accomplish the outcome improving our own ability to deliver differentiating value to the customer.

Only for illustrative purposes I’ve sketched the Wikipedia value chain on the left and the old encyclopedia value chain on the right. You will see they are very different and the new one won’t emerge from the old one. Starting with the old one therefore might anchor the team in a limited view of what’s possible. The Wardley Map is therefore better to use after the first new components are identified in order to qualify the thinking in terms of ability to deliver value to the customer and map out the rest of the components of the chain that are needed.

In summary

As with all new technologies organizations are keen to figure out how it will impact what they are doing and what changes need to be made (if any). Either to keep up (FOMO) or to get ahead (competitive advantage).

As with previous hypes it is easy to jump on the technology narrative and turn opportunities into tech. data or efficiency upgrades, but the promise these technologies introduce is always bigger which is also why we care (doing the same only faster hardly inspires anyone).

But to get there we need to ask new questions, we need to step back from the technology myopia and see these changes as bigger than getting a few new cool toys in our toolbox. We need to see it from a customer / client desired outcome perspective and identify the value, importance or maybe irrelevance of parts of the value chain we are relying on for value today.

Sources:

(1). Ardon Iton, Theories of Value Chains, http://ns1.crfm.net/~uwohxjxf/images/Theories_of_Value_Chains.pdf

(2). Phillip Evans, How data will transform business, https://www.ted.com/talks/philip_evans_how_data_will_transform_business

(3). Tom Goodwin, The battle is for the customer interface, https://techcrunch.com/2015/03/03/in-the-age-of-disintermediation-the-battle-is-all-for-the-customer-interface/

(4). The reliability of Wikipedia, https://en.wikipedia.org/wiki/Reliability_of_Wikipedia#:~:text=The%20study%20concluded%20that%20%22Wikipedia,documentation%20of%20its%20survey%20method.

(5). Clayton Christensen (Innovator’s Dilemma) & Marc Andreessen (a16z) | Startup Grind Global, https://www.youtube.com/watch?v=IkBp1ntD3Zc&t=1812s

(6). Wardley maps, https://medium.com/wardleymaps

(7). Tony Ulwick, When Is A Market Ripe For Disruption?, https://strategyn.com/when-is-a-market-ripe-for-disruption/